A Practical Guide to Free-Energy Devices Author: Patrick J.

Kelly

Chapter 10: Automotive Systems

Note: If you are not at all familiar with basic electronics, you might find it easier to understand this chapter if you
read chapter 12 first.

There are two or three main objectives for people who create automotive devices – increasing the mpg
performance and reducing the harmful emissions are the top two priorities, while running the vehicle on water
alone is the aim of a few people.

The first two objectives are readily achievable, but running a vehicle on water alone is not going to happen for
almost everybody. This idea is peddled by con artists who sell worthless “plans”, claiming that these will run a
vehicle on water for anybody who wants to construct these simple devices. This is just not true. You are
welcome to download the "HydroStar" and "HydroGen" plans free from http://www.free-energy-info.com/P61.pdf
and http://www.free-energy-info.com/P62.pdf. However, most experienced people looking at these plans are
convinced that they could not possibly produce enough hydrogen/oxygen gas mix to run an engine. While I have
never heard of anyone, anywhere, ever getting an engine to run on these plans, the present day science of water
is so inadequate, that I am not in a position to be certain that they could not work, and so I am just highly doubtful
as to them being viable devices.

Just before getting on to explain the construction details of practical systems, let me put the running of an engine
on water alone in its proper context. The internal combustion engine which you own has an efficiency less than
50%. This means that at least half of the energy available from the fuel which you use is wasted and does not
produce any useful mechanical output power. In many cases, that percentage can be as high as 90% wasted, but
let’s be generous and assume that your particular engine is especially good and manages 50% efficiency.

The main way of running an engine with water as the only fuel, involves splitting water into hydrogen and oxygen
and then burning those gases to power the engine. To be self-sustaining, the splitting of the water has to be done
by the electrics of the vehicle and that means that the efficiency of the water splitting has to be more than 200%
efficient. That just doesn’t happen with simple systems, so please forget the notion of building some device in
your garage with a couple of hours work and waving goodbye to filling stations forever – it ain’t going to happen.

Just to set the record straight, it is possible to appear to run a car on water alone, but the difficulty level is about
the same as building a rocket capable of going into orbit, something well beyond the capabilities of most people,
including me. This document does tell you how it can be done, but please understand that it calls for exceptional
skills, very considerable expenditure and a great deal of patience, so for the time being, please forget about it.

What can be done quite readily and at low cost, is to construct a device which will raise the efficiency of your
engine. This is done by feeding a hydrogen/oxygen gas mix (called “HHO” gas) into your engine along with the
air which is drawn in to make the engine run. A device of this type is called a “booster” as it boosts the fuel burn,
extracting a greater percentage of the fuel’s available energy. An important side effect of this improvement in the
burn quality of the fuel is the fact that unburnt fuel no longer gets pushed out of the exhaust as harmful emissions.

Another effect is that the engine has greater pulling power and runs smoother. Inside your engine, carbon
deposits will have built up from previous un-boosted running and these deposits get burnt away when you use a
booster and that internal cleaning extends the engine life.

Some people worry about the fact that burning HHO gas produces water and they imagine this water causing
rusting inside the engine. What they don’t realise is that the ordinary fuel used in the engine is a “hydrocarbon”
which is a compound of hydrogen and carbon and that fuel actually splits up to form hydrogen which the engine
burns. Actually, it is the carbon part of the hydrocarbon fuel which is the problem, producing Carbon Dioxide,
Carbon Monoxide, and physical carbon deposits inside the engine. A normal fuel burn produces water anyway,
but you don’t get rusting inside the engine as the temperature there is so high that any water is in the form of
steam or vapour which dry out completely when the engine is switched off. Adding a small amount of HHO gas
has no adverse effects at all.

This document describes different types of booster. Let me stress that each engine is different and it depends on
how inefficient the engine is to begin with, what sort of mpg improvement is likely to be produced by a booster.
Just to make sure that you understand what is involved, a booster is a simple container which holds a set of
plates submerged in water which probably has an additive to make the water conduct electrical current better. A
pipe from the top of the container feeds the gas into the air filter of the vehicle, via one or two simple safety
devices. Adding this gas causes a major improvement in the quality of the fuel burn inside the engine and cuts
harmful emission to near zero.

10 - 1
As a consequence of this, it is possible to reduce the amount of fossil fuel being sent to the engine, which is not
something which should be done if HHO gas is not being added, as the engine is liable to overheat and some
valve damage could occur. It is a completely different matter if HHO gas is being added. However, all recent
engine designs have an Electronic Control Unit (“ECU”) which controls the amount of fuel being sent to the
engine. The ECU accepts input signals from an “oxygen sensor” placed in the exhaust stream, and often a
second sensor after the catalytic converter to make sure that the catalytic converter has not failed.

Unfortunately, the much improved exhaust caused by the better fuel burn caused by the HHO gas, causes the
ECU to think that the engine fuel-air mix must be too low, and so it pumps in more fuel in an effort to compensate.
Ideally, this can be dealt with by adding a circuit board which adjusts the signal coming from the oxygen sensor so
that it is correct for the improved fuel burn. Details of how to do this are in a companion document.

So, to recap, the only practical device which you can build yourself and use to improve automotive performance is
a ‘booster’. Using a booster improves the efficiency of the fuel burn inside your engine and that results in more
power, better torque, smoother running and vastly improved exhaust emissions. If the ECU is not adjusted or its
input signal not controlled, the mpg figures may actually get slightly lower due to unwanted excess fuel being
pumped into the engine. If a control circuit is used to correct this ECU error, then mpg gains will be produced.

So, what mpg gains can be expected? The worst I have ever heard of was 8% which is very rare. The lowest
likely gain is 20%. Typical gains are in the 25% to 35% bracket. Not particularly unusual is 35% to 60%, while
gains up to 100% and over have been achieved but they are rare. A realistic expectation would be a 33% gain.

This chapter is divided up into the following sections:

1. Simple DC boosters, using a 12-volt electrical input.

2. Advanced DC boosters using much higher DC voltages.
3. Water-splitters which use pulsed electrical signals to change water into "HHO" gas.
4. Running engines without fossil fuels.
5. Other useful devices.

One thing which needs to be understood:

Caution: A booster is not a toy. If you make and use one of these, you do so entirely at your own risk.
Neither the designer of the booster, the author of this document or the provider of the internet display are
in any way liable should you suffer any loss or damage through your own actions. While it is believed to
be entirely safe to make and use a properly built booster, provided that the safety instructions shown in
this document are followed, it is stressed that the responsibility for doing this is yours and yours alone.

Simple DC Boosters.
It is important that you understand the basic principles of electrolysis if you are to be successful in building and
operating a booster, or alternatively, buying and operating a booster. A "DC booster" operates on "Direct Current"
which is the sort of electrical power delivered by a car battery.

The method is very simple in basic outline. Two metal plates are placed in water and an electric current is passed
between the plates. This causes the water to break down into a mixture of hydrogen gas and oxygen gas (The
two components used in the Space Shuttle). The greater the flow of current, the larger the volume of gas which
will be produced. The arrangement is like this:

Remembering that the result of doing this is to produce fuel for the Space Shuttle, you should avoid doing this
indoors and letting the gas produced by the process collect on the ceiling. There are many videos on the web
where people act in a dangerous manner and perform electrolysis indoors using a container which is open at the
top as shown above. Please, please don't do that as it is highly dangerous - it is not a party popper which pushes

10 - 2
the Space Shuttle into space! If you were to collect a cupful of HHO gas and light it, the resulting ignition would
probably damage your hearing permanently, so don't do it under any circumstances. Just like the fact that a very
useful chain saw is a dangerous device which needs to be treated with respect, so too, please understand that the
very useful HHO gas mix contains a lot of energy and so needs to be treated with respect.

This style of electrolysis of water was investigated by the very talented and meticulous experimenter Michael
Faraday. He presented his results in a very technical and scientific format which are not understood by most
ordinary people. But in simple terms, he tells us that the amount of HHO gas produced is proportional to the
current flowing through the water, so to increase the rate of gas production, you need to increase the current flow.
Also, he found that the optimum voltage between the two "electrode" plates is 1.24 volts.

This sounds a bit technical, but it is a highly useful piece of information. In the arrangement shown above, twelve
volts is being connected across two plates in water. Faraday tells us that only 1.24 volts of that twelve volts will
go to make HHO gas and the remaining 10.76 volts will act as an electric kettle and just heat the water, eventually
producing steam. As we want to make HHO gas and not steam, this is bad news for us. What it does tell us is
that if you choose to do it that way, then only 10% of the power taken by the booster actually makes HHO gas and
a massive 90% is wasted as heat.

We really don't want a low electrical efficiency like that. One way around the problem is to use two cells like this:

This arrangement uses our 1.24 volts twice while the twelve volts stays unchanged and so the electrical efficiency
goes up to 20% and the heat loss drops to 80%. That is quite an improvement but even more important is the fact
that twice as much HHO gas is now produced, so we have doubled the electrical efficiency and doubled the gas
output, giving a result which is four times better than before.

We could go one step further and use three cells like this:

This time we are using three of our 1.24 volt sections and this gives us an electrical efficiency of 30% and three
times the amount of gas, making the system nine times more effective.

This is definitely going in the right direction, so how far can we take it when using a twelve volt battery? When we
use the construction materials which years of testing has shown to be particularly effective, there is a small
voltage drop across the metal plates, which means that the very best voltage for each cell is about 2 volts and so
with a twelve volt battery, six cells is about the best combination, and that gives us an electrical efficiency of 62%
and six times as much gas, which is 37 times better than using a single cell, and the wasted electrical power
drops down from 90% to 38%, which is about as good as we can get.

10 - 3
Of course, it would not be practical to have six boxes each as large as a car battery as we would never manage to
fit them into most vehicles. Perhaps we could just put all the plates inside a single box. Unfortunately, if we do
that, a good deal of the electric current would flow around the plates and not make much gas at all. A top view of
this arrangement is shown here:

This is a disaster for us as now we will not get your six times the gas production or our massively reduced
heating. Thankfully, there is a very simple fix for this problem, and that is to divide the box up into six watertight
compartments using thin partitions like this:

This gives us back our high efficiency by blocking the current flow past the plates and forcing the current to flow
through the plates, producing gas between every pair of plates.

In passing, if this booster were to be powered by the electrics of a vehicle, then the voltage although called
"twelve volts" will actually be almost fourteen volts when the engine is running so that the "twelve volt" battery will
get charged. This would allow us to use seven cells inside our electrolyser, rather than the six cells shown above
and that would give us seven times the gas volume that a single pair of plates would give. Some people prefer six
cells, and others, seven cells - the choice is up to the person constructing the unit.

We have been discussing the methods of increasing the gas production and reducing the wasted energy, but
please don't assume that the objective is to make large volumes of HHO gas. It has been found that with many
vehicle engines, very good performance gains can be had with a HHO gas production rate of less than 1 litre per
minute ("lpm"). Flow rates of as little as 0.5 to 0.7 lpm are frequently very effective. Remember, the HHO gas
from a booster is being used as an igniter for the regular fuel used by the engine and not as an additional fuel.

The big advantage of an efficient booster design is that you can produce the wanted volume of gas using a much
lower current, and so, a lesser extra load on the engine. Admittedly, there is not much additional engine load
needed by a booster, but we should reduce the extra amount by intelligent design.

In the discussion above, the battery has been shown connected directly across the booster or "electrolyser". This
should never be done as there is no protection against a short-circuit caused by a loose wire or whatever. There
should be a fuse or a circuit-breaker as the first thing connected to the battery. Circuit breakers are available from
any electrician's supply outlet as they are used in the "fuse box" in homes, to provide protection for each lighting
10 - 4
circuit and each power socket circuit. They are not expensive as they are manufactured in very large volumes.
They are also available on eBay. The circuit breaker is wired like this:

a common design (rated at 32 amps) looks like this:

Some would-be constructors feel that some aspects of the construction are too difficult for them. Here are some
suggestions which might make construction more straightforward.

Constructing a seven-cell housing is not difficult. Pieces are cut out for two sides, one base, one lid and six
absolutely identical partitions. These partitions must be exactly the same so that there is no tendency for leaks to
develop. If you decide to use the bent-plate system of electrodes shown on the next few pages, then drill the bolt
holes in the partitions before assembling them:

The bottom piece is the same length as the sides, and it is the width of the partitions plus twice the thickness of
the material being used to build the housing. If acrylic plastic is being used for the construction, then the supplier
can also provide an “adhesive” which effectively “welds” the pieces together making the different pieces appear to
have been made from a single piece. The case would be assembled like this:

10 - 5
Here, the partitions are fixed in place one at a time, and finally, the second side is attached and will mate exactly
as the partitions and ends are all exactly the same width. A simple construction for the lid is to glue and screw a
strip all the way around the top of the unit and have the lid overlap the sides as shown here:

A gasket, perhaps of flexible PVC, placed between the sides and the lid would assist in making a good seal when
the lid is bolted down. The gas outlet pipe is located in the centre of the lid which is a position which is not
affected if the unit is tilted when the vehicle is on a steep hill.

Years of testing have shown that a really good choice of material for the electrode plates is 316-L grade stainless
steel. However, it is very difficult to connect those plates electrically inside the cells as you need to use stainless
steel wire to make the connections and bolted connections are really not suitable. That leaves welding the wires
to the plates and welding stainless steel is not something which a beginner can do properly as it is much more
difficult than welding mild steel. There is a good alternative, and that is to arrange the plate material so that no
wire connections are needed:

While this six-cell design may look a little complicated to a quick glance, it is really a very simple construction.
Each of the plates used in the central cells is just this shape:

10 - 6
The plate shapes shown above are arranged so that there is access to the bolts from above and they can be
reached by a spanner and held steady while the other nut is being tightened.

Unless you are skilled in bending plates, I suggest that you use stainless steel mesh for the plates. It works very
well, can be readily cut using tin snips or any similar tool and it can be bent into shape by the home constructor
using simple tools - a vice, a piece of angle iron, a small piece of mild steel sheet, a hammer, etc.

You will find a skip outside any metal fabrication shop where scrap pieces are tossed for recycling. There will be
off-cuts of various sizes of angle iron and all sorts of other small sections of sheet and strip. They are in the skip
mainly to get rid of them as the fabrication business gets paid almost nothing for them. You can use some of
these pieces to shape your booster plates, and if you feel bad about costing the business about a penny, then by
all means put them back in the skip afterwards.

If you clamp your plate between two angle irons in a vice, then careful, repeated gently tapping with a hammer
close to the bend location, will produce a very clean and neat bend in the plate:

The bent sheet can then be clamped between two steel strips and a sharp U-shaped bend produced by tapping
with a hammer, again, along the line of the required bend:

The thickness of the steel bar on the inside of the bend has to be the exact width of the required gap between the
finished plate faces. This is not particularly difficult to arrange as 3 mm, 3.5 mm, 4 mm, 5 mm and 6 mm are
common thicknesses used in steel fabrication, and they can be combined to give almost any required gap.

10 - 7
There are many varieties of stainless steel mesh. The style and thickness are not at all critical but you need to
choose a type which is reasonably stiff and which will hold its shape well after it is bent. This style might be a
good choice:

Your local steel supplier probably has some types on hand and can let you see how flexible a particular variety is.
The shape shown above is for a "three plate per cell" design where there are two active plate faces. Ideally, you
want two to four square inches of plate area per amp of current flowing through the cell, because that gives very
long electrode life and minimum heating due to the plates.

This style of construction is reasonably easy to assemble as the two bolts which pass through the partitions and
which hold the plates rigidly in place, can be accessed from above, two spanners being used to lock them tight.
Lock nuts are optional. If you feel that your particular mesh might be a little too flexible or if you think that the
bolts might eventually loosen, then you can attach two, or more, separator insulating pieces - plastic washers,
plastic bolts, cable ties or whatever to one of the plate faces.

These will hold the plates apart even if they were to become loose. They also help to maintain the gap between
the plates. This gap has to be a compromise because the closer the plates are together, the better the gas
production but the more difficult it is for the bubbles to break away from the plates and float to the surface and if
they don't do that, then they block off some of the plate area and prevent further gas production from that part of
the plate as the electrolyte no longer touches the plate there. A popular choice of gap is 1/8 inch which is 3 mm
as that is a good compromise spacing. Circular spacers would look like this:

If the current is low enough, an even more simple shape which has just a single pair of active plate surfaces per
cell, can be used as shown here:

10 - 8
Any of these designs can be 6-cell or 7-cell and the plates can be constructed without outside help. You will
notice that the electrical connections at each end of the booster are submerged to make sure that a loose
connection can't cause a spark and ignite the HHO gas in the top of the housing. There should be a gasket
washer on the inside to prevent any leakage of the electrolyte past the clamping bolt.

If you want to use three active plate pairs in each cell, then the plate shape could be like this:

The electrolyte is a mix of water and an additive to allows more current to flow through the liquid. Most of the
substances which people think of to use to make an electrolyte are most unsuitable, producing dangerous
gasses, damaging the surfaces of the plates and giving uneven electrolysis and currents which are difficult to
control. These include salt, battery acid and baking soda and I strongly recommend that you do not use any of
these.

What is needed is a substance which does not get used up during electrolysis and which does not damage the
plates even after years of use. There are two very suitable substances for this: sodium hydroxide, also called
"lye" or "caustic soda". In the USA, this is available in Lowes stores, being sold as "Roebic ‘Heavy Duty’ Crystal
Drain Opener". The chemical formula for it is NaOH.

One other substance which is even better is potassium hydroxide or "caustic potash" (chemical formula KOH)
which can be got from soap-making supply shops found on the web. Both NaOH and KOH are very caustic
materials and they need to be handled with considerable care.

Bob Boyce of the USA is one of the most experienced people in the construction and use of boosters of different
designs. He has kindly shared the following information on how to stay safe when mixing and using these
chemicals. He says:

10 - 9
These materials are highly caustic and so they need to be handled carefully and kept away from contact with skin,
and even more importantly, eyes. If any splashes come in contact with you, it is very important indeed that the
affected area be rinsed off immediately with large amounts of running water and if necessary, the use of vinegar
which is acidic and so will neutralise the caustic liquid.

When making up a solution, you add small amounts of the hydroxide to distilled
water held in a container. The container must not be glass as most glass is not
high enough quality to be a suitable material in which to mix the electrolyte. The
hydroxide itself should always be stored in a sturdy, air-tight container which is
clearly labelled "DANGER! - Potassium (or Sodium) Hydroxide". Keep the
container in a safe place, where it can’t be reached by children, pets or people who
won't take any notice of the label. If your supply of hydroxide is delivered in a
strong plastic bag, then once you open the bag, you should transfer all of its
contents to sturdy, air-tight, plastic storage containers, which you can open and
close without any risk of spilling the contents. Hardware stores sell large plastic
buckets with air tight lids that can be used for this purpose.

When working with dry hydroxide flakes or granules, wear safety goggles, rubber gloves, a long sleeved shirt,
socks and long trousers. Also, don’t wear your favourite clothes when handling hydroxide solution as it is not the
best thing to get on clothes. It is also no harm to wear a face mask which covers your mouth and nose. If you are
mixing solid hydroxide with water, always add the hydroxide to the water, and not the other way round, and use a
plastic container for the mixing, preferably one which has twice the capacity of the finished mixture. The mixing
should be done in a well-ventilated area which is not draughty as air currents can blow the dry hydroxide around.

When mixing the electrolyte, never use warm water. The water should be cool because the chemical reaction
between the water and the hydroxide generates a good deal of heat. If possible, place the mixing container in a
larger container filled with cold water, as that will help to keep the temperature down, and if your mixture should
“boil over” it will contain the spillage. Add only a small amount of hydroxide at a time, stirring continuously, and if
you stop stirring for any reason, put the lids back on all containers.

If, in spite of all precautions, you get some hydroxide solution on your skin, wash it off with plenty of cold running
water and apply some vinegar to the skin. Vinegar is acidic, and will help balance out the alkalinity of the
hydroxide. You can use lemon juice if you don't have vinegar to hand - but it is always a good idea to have a
bottle of vinegar handy.

The concentration of the electrolyte is a very important factor. Generally speaking, the more concentrated the
electrolyte, the greater the current and the larger the volume of HHO gas produced. However, there are three
major factors to consider:

1. The resistance to current flow through the metal electrode plates.

2. The resistance to current flow between the metal plates and the electrolyte.
3. The resistance to current flow through the electrolyte itself.

1. In a good electrolyser design like those shown above, the design itself is about as good as a DC booster can
get, but understanding each of these areas of power loss is important for the best possible performance. We
were taught in school that metals conduct electricity, but what was probably not mentioned was the fact that
some metals such as stainless steel are quite poor conductors of electricity and that is why electrical cables are
made with copper wires and not steel wires. This is how the current flow occurs with our electrolyser plates:

The fact that we have folds and bends in our plates has no significant effect on the current flow. Resistance to
current flow through the metal electrode plates is something which can’t be overcome easily and economically,

10 - 10
 and so has to be accepted as an overhead. Generally speaking, the heating from this source is low and not a
matter of major concern, but we provide a large amount of plate area to reduce this component of power loss
as much as is practical.

2. Resistance to flow between the electrode and the electrolyte is an entirely different matter, and major
improvements can be made in this area. After extensive testing, Bob Boyce discovered that a very
considerable improvement can be made if a catalytic layer is developed on the active plate surface. Details of
how this can be done are given later in the companion "D9.pdf" document as part of the description of Bob’s
electrolyser.

3. Resistance to flow through the electrolyte itself can be minimised by using the best catalyst at its optimum
concentration. When using sodium hydroxide, the optimum concentration is 20% by weight. As 1 cc of water
weighs one gram, one litre of water weighs one kilogram. But, if 20% (200 grams) of this kilogram is to be
made up of sodium hydroxide, then the remaining water can only weigh 800 grams and so will be only 800 cc
in volume. So, to make up a 20% "by weight" mix of sodium hydroxide and distilled water, the 200 grams of
sodium hydroxide are added (very slowly and carefully, as explained above by Bob) to just 800 cc of cool
distilled water and the volume of electrolyte produced will be about 800 cc.

When potassium hydroxide is being used, the optimum concentration is 28% by weight and so, 280 grams of
potassium hydroxide are added (very slowly and carefully, as explained above by Bob) to just 720 cc of cold
distilled water. Both of these electrolytes have a freezing point well below that of water and this can be a very
useful feature for people who live in places which have very cold winters.

Another factor which affects current flow through the electrolyte is the distance which the current has to flow
through the electrolyte - the greater the distance, the greater the resistance. Reducing the gap between the
plates to a minimum improves the efficiency. However, practical factors come into play here as bubbles need
sufficient space to escape between the plates, and a good working compromise is a spacing of 3 mm. which is
one eighth of an inch.

However, there is a problem with using the optimum concentration of electrolyte and that is the current flow
caused by the greatly improved electrolyte is likely to be far more than we want. To deal with this we can use an
electronic circuit called a "Pulse-Width Modulator" (or “PWM”) circuit. These are often sold as "DC Motor Speed
Controllers" and if you buy one, then pick one which can handle 30 amps of current.

A PWM circuit operates in a very simple way. It switches the current to the electrolyser On and Off many times
every second. The current is controlled by how long (in any one second) the current is On, compared to how long
it is Off. For example, if the On time is twice as long as the Off time (66%), then the average current flow will be
much greater than if the On time were only half as long as the Off time(33%).

10 - 11
When using a PWM controller, it is normal to place its control knob on or near the dashboard and to mount a
simple low-cost ammeter beside it so that the driver can raise or lower the current flow as is considered
necessary. The arrangement is like this:

There is a more sophisticated circuit controller called a "Constant-current Circuit" and that allows you to select the
current you want and the circuit then holds the current at your set value at all times. However, this type of circuit
is not readily available for sale although some outlets are preparing to offer them.

Some of the most simple boosters don't use a PWM circuit because they control the current flow through the
booster by making the concentration of the electrolyte very low so that the resistance to current flow through the
electrolyte chokes off the current and holds it down to the desired level. This, of course, is far less efficient and
the resistance in the electrolyte causes heating, which in turn, is an operational problem which needs careful
handling by the user. The advantage is that the system appears to be more simple.

Feeding HHO Gas to an Engine.

When using a booster of any design you need to realise that HHO gas is highly energetic. If it wasn’t, it would not
be able to do it’s job of improving the explosions inside your engine. HHO gas needs to be treated with respect
and caution. It is important to make sure that it goes into the engine and nowhere else. It is also important that it
gets ignited inside the engine and nowhere else.

To make these things happen, a number of common-sense steps need to be taken. Firstly, the booster must not
make HHO gas when the engine is not running. The best way to arrange this is to switch off the current going to
the booster when the engine is not running. It is not sufficient to just have a manually-operated On/Off switch as
it is almost certain that switching off will be forgotten one day. Instead, the electrical supply to the booster is
routed through the ignition switch of the vehicle. That way, when the engine is turned off and the ignition key
removed, it is certain that the booster is turned off as well.

So as not to put too much current load on the ignition switch, and to allow for the possibility of the ignition switch
being on when the engine is not running, instead of wiring the booster directly to the switch, it is better to wire a
standard automotive relay across the oil pressure unit and let the relay carry the booster current. The oil pressure
drops when the engine stops running, and so this will also power down the booster.

An extra safety feature is to allow for the (very unlikely) possibility of an electrical short-circuit occurring in the
booster or its wiring. This is done by putting a fuse or contact-breaker between the battery and the new circuitry
as shown in this diagram:

If you choose to use a contact-breaker, then a light-emitting diode (“LED”) with a current limiting resistor of say,
680 ohms in series with it, can be wired directly across the contacts of the circuit breaker. The LED can be
mounted on the dashboard. As the contacts are normally closed, they short-circuit the LED and so no light
shows. If the circuit-breaker is tripped, then the LED will light up to show that the circuit-breaker has operated.
The current through the LED is so low that the electrolyser is effectively switched off when the contact breaker
opens. This is not a necessary feature, merely an optional extra:

10 - 12
A good source for general components needed in building boosters is The Hydrogen Garage in the USA, website:
http://stores.homestead.com/hydrogengarage/StoreFront.bok A very important safety item for any booster is the
“bubbler” which is just a simple container with some water in it. The bubbler has the gas coming in at the bottom
and bubbling up through the water. The gas collects above the water surface and is then drawn into the engine
through an outlet pipe above the water surface. To prevent water being drawn into the booster when the booster
is off for any length of time and the pressure inside it reduces, a one-way valve is placed in the pipe between the
booster and the bubbler.

If the engine happens to backfire, then the bubbler blocks the flame from passing back through the pipe and
igniting the gas being produced in the booster. A bubbler is a very simple, very cheap and very sensible thing to
install. It also removes any traces of electrolyte fumes from the gas before it is drawn into the engine. In practice,
it is a very good idea to have two bubblers, one close to the booster and one close to the engine. The second
bubbler makes sure that every last trace of electrolyte fumes are washed out of the HHO gas before it enters the
engine.

There are various ways to make a good bubbler. In general, you are aimed at having a five-inch (125 mm) depth
of water through which the HHO gas must pass before it leaves the bubbler. It is recommended that a bubbler is
built inside a strong container such as this one:

These strong containers are generally sold as water filters. They can be adapted to become bubblers without any
major work being done on them. At this point, we need to consider the mechanism for moving the HHO gas out of
the booster and into the engine.

It is generally a good idea to position the gas take-off pipe in the centre of the lid so that if the booster gets tilted
due to the vehicle operating on a sloped surface, then the surface level of the liquid remains unchanged
underneath the gas pipe. A common mistake is to use a gas pipe which has a small diameter. If you take a
length of plastic pipe of a quarter inch diameter (6 mm) and try blowing through it, you will be surprised at how
difficult it is to blow through. There is no need to give your booster that problem, so I suggest that you select a
gas pipe of half an inch (12 mm) or so. If in doubt as to how suitable a pipe is, then try blowing through a sample
length of it. If you can blow through it without the slightest difficulty, then it is good enough for your booster.

10 - 13
One other thing is how to deal with splashes and the spray from bubbles bursting at the surface of the electrolyte.
You want some device which will prevent any spray or splashes caused by the vehicle going over a very rough
road, from entering the gas pipe and being drawn out of the booster along with the HHO gas.

Various methods have been used and it is very much a matter of personal choice as to how you decide to deal
with the issue. One method is to use a piece of suitable material across the end of the pipe. This is generally
called anti-slosh material because of the job which it does. The material needs to let the gas pass freely through
it but prevent any liquid getting through it. Plastic pot-scrubbers as a possible material as they have an
interlocking mesh of small flat strands. The gas can flow around and through the many strands, but splashes
which go in a straight line will hit the strands and drip back into the booster again. Another possible device is one
or more baffles which will catch the liquid but let the gas pass freely by:

OR

OR

The HHO gas produced by a DC booster of this type contains about 30% monatomic hydrogen, which means that
30% of the hydrogen is in the form of single atoms of hydrogen and not combined hydrogen pairs of atoms. The
monatomic form is about four times more energetic than the combined form and so it takes up a greater volume
inside the booster housing.

If the booster is left turned off for a long period of time, then these single hydrogen atoms will eventually bump into
each other and combine to form the less energetic diatomic form of the gas. As this takes up less space inside
the booster, the pressure inside the booster drops and this has been known to suck water out of the bubbler back
into the booster. We don't want this to happen as it dilutes our carefully measured electrolyte concentration and it
can make the bubbler ineffective due to lack of water.

10 - 14
To deal with this, a one-way valve is put between the booster and the bubbler, positioned so that it does not allow
flow back into the booster. In very cold climates, a 28% by weight potassium hydroxide electrolyte will not freeze
until -40O C, it is more difficult to stop the bubblers freezing. While it is possible to have equipment which unplugs
and can be taken indoors overnight, an alternative is to use alcohol or paraffin (kerosene) instead of water and
they generally do not freeze and their fumes are not harmful to an engine.

The bubbler design is not difficult. Ideally, you want a very large number of small bubbles to be formed and float
upwards through the water. This is because it gives the best connection between the gas and the water and so
can do a really good job of washing any traces of hydroxide vapour out of the HHO gas before it gets fed to the
engine. Small bubbles are also better separated from each other and so there is no real chance of a flame
passing through the water where large bubbles might merge together and form a column of gas as they rise to the
surface.

In this good bubbler design, the pipe which feeds the HHO gas into the bubbler is bent into an L-shape. The end
of the pipe is blocked off, and many small holes are drilled in the horizontal section of the pipe. Only a few holes
are seen in this diagram, but there will be a large number in the actual construction. Like the booster itself, the
gas outlet pipe needs to be protected from splashes of water caused by the vehicle going over a bump. It is very
important to make sure that water is not drawn into the engine along with the gas, so anti-slosh material or one or
more baffles are used to prevent this happening. So the overall protection for the gas flow is:

10 - 15
Where the first bubbler is close to the booster and the second one is placed close to the engine. Once in a while,
the water from the first bubbler can be used to top up the water inside the booster so that any traces of hydroxide
which may have reached the bubbler are returned to the booster, keeping its electrolyte concentration exactly
right and making sure that the water in the bubbler is always fresh.

There is one final item which is an optional extra. Some people like to add a gas-pressure switch. If, for any
reason, the pressure starts to rise - say that the outlet pipe became blocked - then the pressure switch would
disconnect the electrical supply and stop the pressure rising any further:

One decision which has to be made is the rate of HHO gas production which is the best for you. Most people
seem to think that the larger the volume of HHO gas the better. That is not necessarily true because a very
effective use of the gas is to make it act as an igniter for the engine's normal fuel and very satisfactory results
have been achieved with HHO gas flow rates in the range of 0.4 to 0.7 litres per minute. You control the rate of
gas production by controlling the current, either by the concentration of the electrolyte or by adjusting the current
flow using an electronic circuit.

Each litre of water produces about 1,750 litres of HHO gas, so you can estimate the length of time the booster can
operate on one litre of water. If, for example, your booster is producing 0.7 litres of gas per minute. Then, it will
produce 1,750 litres in 1,750 / 0.7 minutes and that is 2,500 minutes or 41 hours 40 minutes. As the booster only
operates when you are driving, you are looking at 41 hours of driving time and if you drive about two hours per
day, it would take three weeks to use one litre of water. The internal dimensions of your booster allow you to
calculate how far the electrolyte level will drop if one litre of water is taken out of it.

Generally speaking, it is normally considered that topping up the booster with water by hand every so often, is a
perfectly good method of operation. The booster design described above has a good electrolyte capacity in each
cell and so topping up with water should not be a major task. As tap water and well water have a good deal of
dissolved solids in them, when the water is taken away by electrolysis, these solids drop out of solution and fall to
the bottom of the housing, and/or coat the plates with an layer of unwanted material. For this reason, life is so
much easier if distilled water is used for making electrolyte and for topping up the booster after use.

It is possible to have an automatic water supply for your booster even though that is probably over-kill for such a
simple device. If you decide to do that, then you need a water supply nozzle for each of your six or seven cells. It
is not necessary for the electrolyte level to be exactly the same in each cell, but you would normally have them at
roughly the same height. Your automated water supply could be like this:

10 - 16
A point which might not be immediately obvious is that because the gas pressure inside the booster is probably
about 5 pounds per square inch ("psi"), once the water pump stops pumping, it is possible for the gas pressure to
push out the remaining water in the inlet pipes and escape through the body of the pump. To prevent this, an
ordinary one-way valve is put in the water supply pipe to prevent flow back towards the pump.

Up to now, the HHO gas feed to the engine has just been indicated in a vague way in spite of the connection point
being important. With most engines, the HHO gas should be fed into the air filter where it mixes well and is fully
dispersed inside the air being drawn into the engine. You sometimes see diagrams which show the connection
point being close to the engine intake manifold. This is not a good idea because the lowered pressure there
causes reduced pressure inside the booster which in turn produces more unwanted hot water vapour, so stick
with feeding the gas into the air filter. If there is a supercharger on the engine, then feed the HHO gas into the
low-pressure side of the supercharger.

The “Smack’s” Booster.

The style of booster described above has the advantages of high electrical efficiency, easy construction, very few
specialist parts and a large electrolyte volume per cell. There are many other very successful booster designs
which have very different forms of construction. One of these is the "Smack's Booster" where electrical cover
plates are clamped together and placed inside a length of plastic pipe:

10 - 17
The advantages of this design are the very simple construction, compact size, reasonable performance and the
fact that you can buy one ready-made if you want to. You can download a copy of the construction details free
from http://www.free-energy-info.com/Smack.pdf The electrical efficiency of this design is lowered a bit because
only a single body of electrolyte is used and so current can bypass the plates. The overall performance is a
respectable 1.3 lpm for 20 amps, though you may wish to lower the current and settle for about half that rate of
HHO gas production. The construction of a 5 lpm version is at http://www.youtube.com/watch?v=cqjn3mup1So

The “Hotsabi” Booster.

Another design which is very easy to build is the "HotSabi" booster, which is a single threaded rod inside a length
of plastic pipe with a stainless steel inner lining. It has the lowest possible electrical efficiency, being just a single
cell with the full vehicle voltage connected directly across it, but in spite of that, it's performance in actual on the
road use has been remarkable, with a reported 50% improvement on a 5 litre capacity engine. This excellent
performance is probably due to the design having a steam trap which removes the hot water vapour produced by
the excessive heating caused by having only a single cell with so much voltage across it (remember, 90% of the
power supplied to this booster design goes in heating the electrolyte).

As the designer of this booster has freely shared his design, the free construction plans can be downloaded from
http://www.free-energy-devices.com/Hotsabi.pdf

Controlling the Oxygen Sensor

When an mpg. improving device such as an electrolyser is fitted to a vehicle, the result does not always produce
better mpg. figures. Older vehicles which are fitted with a carburettor will see an immediate improvement. This is
not the case for more recent vehicles which come with computer control of the fuel sent to the engine.

When an electrolyser is attached to the engine, it causes the fuel burn inside the cylinders to be greatly improved,
with a corresponding improvement in engine performance. Unfortunately, the fuel computer is expecting the
same amount of unburnt oxygen to come out of the engine, and when it doesn’t detect it, the computer increases
the fuel flow rate into the engine in an attempt to get back to it’s normal, inefficient method of running. That action

10 - 18
cancels the mpg improvement produced by the electrolyser unless something is done to adjust the operation of
the computer.

In the most simple terms, most vehicles which have an Electronic Control Unit (“ECU”) to control the fuel flow into
the engine, are fitted with one of two types of exhaust sensor. The majority have a “narrowband” sensor while the
remainder have a “wideband” sensor. The ideal mix of air to fuel is considered to be 14.7 to 1. A narrowband
sensor only responds to mixtures from about 14.2 to 1 through 14.9 to 1. The sensor operates by comparing the
amount of oxygen in the exhaust gas to the amount of oxygen in the air outside the vehicle and it generates an
output voltage which moves rapidly between 0.2 volts where the mixture is too lean, and 0.8 volts when it passes
below the 14.7 to 1 air/fuel mix point where the mixture is too rich (as indicated by the graph shown below). The
ECU increases the fuel feed when the signal level is 0.2 volts and decreases it when the signal voltage is 0.8
volts. This causes the signal voltage to switch regularly from high to low and back to high again as the computer
attempts to match the amount of “too lean” time to the amount of “too rich” time.

A simple control circuit board can be added to alter the sensor signal and nudge the fuel computer into producing
slightly better air/fuel mixes. Unfortunately, there is a severe downside to doing this. If, for any reason, the fuel
mix is set too high for an extended period, then the excess fuel being burnt in the catalytic converter can raise the
temperature there high enough to melt the internal components of the converter. On the other hand, if the circuit
board is switched to a mix which is too lean, then the engine temperature can be pushed high enough to damage
the valves, which is an expensive mistake.

Over-lean running can occur at different speeds and loads. Joe Hanson recommends that if any device for
making the mix leaner is fitted to the vehicle, then the following procedure should be carried out. Buy a “type K”
thermocouple with a 3-inch stainless steel threaded shank, custom built by ThermX Southwest of San Diego.
This temperature sensor can measure temperatures up to 1,800 degrees Fahrenheit (980 degrees Centigrade).
Mount the thermocouple on the exhaust pipe by drilling and tapping the pipe close to the exhaust manifold, just
next to the flange gasket. Take a cable from the thermocouple into the driver’s area and use a multimeter to show
the temperature.

Drive the vehicle long enough to reach normal running temperature and then drive at full speed on a highway.
Note the temperature reading at this speed. When a leaner mix is used, make sure that the temperature reading
under exactly the same conditions does not exceed 180 degrees Fahrenheit (100 degrees Centigrade) above the
pre-modification temperature.

David Andruczyk recommends an alternative method of avoiding engine damage through over-lean fuel/air
mixtures, namely, replacing the narrowband oxygen sensor with a wideband sensor and controller. A wideband
oxygen sensor reads a very wide range of Air/Fuel ratios, from about 9 to 1 through 28 to 1. A normal car engine
can run from about 10 to 1 (very rich) to about 17.5 to 1 (pretty lean). Maximum engine power is developed at a
mix ratio of about 12.5 to 1. Complete fuel combustion takes place with a mix of about 14.7 to 1, while the mix
which gives minimum exhaust emissions is slightly leaner than that.
10 - 19
Unlike narrowband sensors, wideband sensors need their own controller in order to function. There are many of
these units being offered for sale for retro-fitting to existing vehicles which have just narrowband oxygen sensor
systems. David’s personal recommendation is the Innovate Motorsports LC-1 which is small, and uses the very
reasonably priced LSU-4 sensor. This wideband controller can be programmed. Most controllers have the ability
to output two signals, the wideband signal suitable for running to a gauge or new ECU, plus a synthesised
narrowband signal which can feed an existing ECU. The trick is to install a wideband sensor, with the LC-1
controller and then reprogram it to shift the narrowband output to achieve a leaner mix as shown here:

Actual Air/Fuel Mix Wideband Output Original Narrowband Shifted Narrowband

Output Output
9 to 1 9 to 1 Mix is too Rich Mix is too Rich
10 to 1 10 to 1 Mix is too Rich Mix is too Rich
11 to 1 11 to 1 Mix is too Rich Mix is too Rich
12 to 1 12 to 1 Mix is too Rich Mix is too Rich
13 to 1 13 to 1 Mix is too Rich Mix is too Rich
14 to 1 14 to 1 Mix is too Rich Mix is too Rich
14.6 to 1 14.6 to 1 Mix is too Rich Mix is too Rich
14.8 to 1 14.8 to 1 Mix is too Lean Mix is too Rich
15 to 1 15 to 1 Mix is too Lean Mix is too Rich
15.5 to 1 15.5 to 1 Mix is too Lean Mix is too Lean
16 to 1 16 to 1 Mix is too Lean Mix is too Lean
18 to 1 18 to 1 Mix is too Lean Mix is too Lean

This system allows you to set the narrowband “toggle point” very precisely on an exact chosen air/fuel ratio. This
is something which it is nearly impossible to do accurately with a circuit board which just shifts a narrowband
oxygen signal as you just do not know what the air/fuel ratio really is with a narrowband sensor.

However, for anyone who wants to try adding a circuit board to alter a narrowband sensor signal to produce a
leaner mix on a vehicle, the following description may be of help. It is possible to buy a ready-made circuit board,
although using a completely different operating technique, from the very reputable Eagle Research, via their
website: http://www.eagle-research.com/products/pfuels.html where the relevant item is shown like this:

10 - 20
This unit generates a small voltage, using a 555 timer chip as an oscillator, rectifying the output to give a small
adjustable voltage which is then added to whatever voltage is being generated by the oxygen sensor. This
voltage is adjusted at installation time and is then left permanently at that setting. Eagle Research also offer for
sale, a booklet which shows you how to build this unit from scratch if you would prefer to do that.

If you wish to use a circuit board with a narrowband oxygen sensor, then please be aware that there are several
versions of this type of sensor. The version is indicated by the number of connecting wires:

Those with 1 wire, where the wire carries the signal and the case is ground (zero volts)
Those with 2 wires, where one wire carries the signal and the other wire is ground.
Those with 3 wires, where 2 (typically slightly thicker) wires are for a sensor heater, and
1 for the signal while the case is ground.
Those with 4 wires (the most common on current model cars), where there are
2 (slightly heavier) for the sensor heater,
1 for the signal , and
1 for the signal ground.
(Sensors with 5 wires are normally wideband devices.)

Look in the engine compartment and locate the oxygen sensor. If you have difficulty in finding it, get a copy of the
Clymer or Haynes Maintenance Manual for your vehicle as that will show you the position. We need to identify
the sensor wire which carries the control signal to the fuel control computer. To do this, make sure that the car is
switched off, then

For 3 and 4 wire sensors:

Disconnect the oxygen sensor wiring harness,
Set a multimeter to a DC voltage measurement range of at least 15 volts,
Turn on the ignition and probe the socket looking for the two wires that provide 12 volts.
These are the heater wires, so make a note of which they are,
Shut the ignition off, and reconnect the oxygen sensor.

The two remaining wires can now be treated the same as the wires from a 2-wire sensor, one will carry the sensor
signal and one will be the signal ground (for a single wire sensor, the signal ground will be the engine block).
Jesper Ingerslev points out that the Ford Mustang built since 1996 has 2 oxygen sensors per catalytic converter,
one before the converter and one after. Some other vehicles also have this arrangement. With a vehicle of this
type, the circuit board described here should be attached to the sensor closest to the engine.

Find a convenient place along the wires. Don’t cut these wires, you will cut the sensor wire here at a later time,
but not now. Instead, strip back a small amount of the insulation on each wire. Be careful to avoid the wires
short-circuiting to each other or to the body of the vehicle. Connect the DC voltmeter to the wires (the non- heater
wires). Start the engine and watch the meter readings. When the engine is warmed up, if the oxygen sensor is
performing as it should (i.e. no engine check lights on), the voltage on the meter should begin toggling between a
low value near zero volts and a high value of about 1 volt. If the meter reading is going negative, then reverse
the meter leads. The black multimeter lead is connected to the signal 'ground' (zero volts) and the red lead will be
connected to the wire which carries the signal from the sensor. Connect a piece of insulated wire to the stripped
point of the sensor wire and take the wire to the input of your mixture controller circuit board. Connect a second
insulated wire between the signal ‘ground’ wire, or in the case of a 1-wire sensor, the engine block, and the circuit
board zero-volts line. Insulate all of the stripped cables to prevent any possibility of a short-circuit:

10 - 21
More specific detail:
However, the situation is by no means a simple one which allows a single simple adaption which will work on
every vehicle for many years. Les Pearson has been investigating this situation in depth for three years along
with a friend who is an Electronics Engineer. Having built and tested EFIEs, the oxygen sensor circuit shown
below, several versions of MAP controllers, coolant/air temperature hacks, professional systems, etc. and
discovered that many vehicle ECUs ("Electronic Control Units") learnt to adapt to the new conditions and return to
the highly inefficient excess fuel injection condition. This return to the original fuel injection is different for each
design of ECU and there are many different designs.

Les says: "To understand the solution, first you have to understand the dilemma with all the other ECU control
tricks. The EFIEs, MAP adjusters, temp hacks etc. do get good results for a short time, and then the performance
deteriorates again. Why should this be? It is because the ECU learns to deal with the new situation with them in
place. This is because the ECU knows that the feedback from most of it's controls, and sensor's are not linear,
nor should they be. All of the electronics, and adjustment methods used by the Hydrogen On Demand people are
linear, and that is not an adequate way to deal with the problem.

For example, we may add a couple of hundred millivolts to the oxygen ("lambda") sensor signal in order to return
an unduly rich signal to the ECU, and so make it respond with a lower level of fuel injection. This makes the ECU
think that the Air/Fuel Ratio is say, 15:1 or 20:1. Now the mass air maps are all wrong, we need to show less air
so that the ECU adds less fuel in order to match the fuel trims. We now need to adjust the ignition timing to take
advantage of this combustion change. The problem is that all the EFIEs, MAP/MAF adjusters, and attempts at
changing timing by manipulating air temp are a static fixed offset, producing just a single change.

An Air/Fuel Ratio of 20:1 is not appropriate under moderate or heavy acceleration when you don't, and usually
can't, add enough hydroxy gas to allow for these conditions. A set 15% to 20% leaner setting across the entire
fuel map is not appropriate at all loads, and engine speeds. Adding 50 to 75 degrees F to the air temp is not
appropriate when the outside air is already at 110 degrees F. The ECU knows this and makes appropriate
changes to it's long-term settings, and so it cancels out the changes which our electronic additions have made.

While it may be OK for an experimenter, and mechanic with all the tools at his disposal to play with these
techniques, and retune his engines every week or so in order to get great mileage, it is not realistic to expect the
average person to do this. The cost in equipment alone, would undermine any fuel savings. Also, most people
don't even change their oil at appropriate intervals.

This was my train of thought says Les. As I pondered a solution, and began searching, my search was for a
control solution which could firstly, alter the air-flow readings, lambda readings, and ignition timing and secondly,
respond to changes in engine speed and engine load. To my surprise I found several products already on the
market which are capable of doing this, and which have been available for quite some time. People in the engine-
tuning industry have been using them for years. They are custom programmable, piggy-back chips. Several
companies make them, and while most do not advertise O2 ("lambda") sensor control, many are quite capable of
altering it.

We became a distributor for one brand which seemed to be the best fit for our purposes, and we began testing.
The results are perfect. We started with an 2002 Saturn SL. The average mpg for this car started at 26 mpg
(highway and city combined). We installed the chip, tested several tuning methods, and found the one which
worked best. The car now averages 44-46 mpg. This is not special 'grandma driving' to try and coax a few extra
miles per gallon. This is a courier vehicle for a local printer, and it is driven daily like it was stolen. We have all
the same benefits of increased torque (pulling power), better throttle response, etc. The car has been driven
around for three months now with our programmed chip installed, and it achieves the mid-40s in mpg all of the
time. there are no code changes needed, no start-up problems, no driving problems of any kind whatsoever. If

10 - 22
you weren't told, you would never know that hydroxy gas was being added, except for the fact that you can go
over 500 miles on a single tank of fuel.

The only problem is that this is definitely not a do-it-yourself solution. You need a laptop computer with
proprietary software to tune the chip, and the scanning, and diagnostic equipment to know when you have it tuned
correctly. However, I have thought of a do-it-yourself solution. It still requires you to buy a few electronic kits, and
you need a lot of know-how, but we are circumventing a highly sophisticated control architecture, so anyone who
thinks it will be easy, is delusional. The main item is a Digital Fuel Adjuster or "DFA" kit from JayCar electronics
based in Australia. Their website is http://www.jaycar.com. The adjuster kit number is KC5385, and you have to
have the hand controller to program it, that kit number is KC5386. The adjuster doesn't have an enclosure
supplied, but the controller does. You need one controller and two adjusters. The controller can be reused to
program multiple adjusters. Once you have the kits, it will take several days of soldering to build them, and it is
definitely not a beginner's first-time project.

After the two adjusters and the controller have been built, the first one is wired in to the MAP/MAF sensor signal
as shown in the instructions guide. Next the second DFA is used on the oxygen sensor signal. If there are two
oxygen sensors, then the DFA is wired through the common ground for the upstream sensors ONLY. This places
the voltage offsets in tandem, which makes it unnecessary to use two DFAs (or EFIEs for that matter) for
"lambda" control. Now control has been established over fuel maps, and a "lambda" control which is responsive
to engine load has been achieved. I believe that these kits also come with an option to make them responsive to
rpm.

For ignition timing, the temperature offsets will probably still be necessary, but now you have a fuel control which
if tuned properly, the ECU will not learn its way around. I have found that the maps for "lambda" control are very
simple. Tune for the leanest Air/Fuel Ratio appropriate at very low loads, and richen up a bit in increments as the
load increases. As you get close to full throttle, but before you switch to open-loop operation, your lambda offset
should be zero (the stock setting). To tune the air-flow or fuel maps, watch the OBD II scan gauge, and decrease
the MAP signal so that your short term fuel trim ("STFT" on a scanner) is no greater than about ± 7% at each load
interval. Drive for about 20 minutes, and check that your long term fuel trim ("LTFT") never goes beyond the "7s"
either. Now the ECU cannot "see" your changes because the fuel map, and lambda readings "agree" at every
load range.

The Jaycar kits are not as sophisticated as the chip sets, but they are about 20% of the cost, if you want to put the
time and effort into them. The adjuster itself simply adds to, or subtracts from, whatever voltage runs across
them, and it can be set to change that offset value in correspondence to whatever voltage value is present at the
signal input pin. You would of course put your TPS signal to the signal input pin. The device itself is very
versatile, and could be used for many different applications. If you wanted to use one to control a Pulse-Width
Modulator attached to a cell, then that would be possible and it would provide a variable gas rate that responds to
changes in engine load. I hope you can put this to good use, and feel free to spread the word around. Perhaps
you know someone who could build a similar device or give us a schematic to build one, after looking at a JayCar
kit. The only drawback to the kits is that rpm sampling gets a little complex, and while I don't think it is absolutely
necessary, it would be beneficial. Although the kits have only 125 data points between the "closed", and "fully
open" throttle settings, and do not interpolate between data points, they seem to work very well. The professional
chip sets have 96,000 data points between CT ("Closed Throttle"), and WOT ("Wide-Open Throttle"), and they do
logically interpolate between set data points. The professional chip sets run about $650 USD programmed, and
installed.

I think that the "more is better" hydroxy gas idea that a lot of people are stuck in, is seriously flawed. There is
definitely a point of diminishing returns. I tune most systems to deliver about 1 lpm. The lower the amperage you
can do this at, of course the better. I have found that not only does it take more amperage to produce higher
volumes at a less effective rate of return, but it does not add much to the efficiency of the "boosting". With the
cells which I build I get 1 lpm of hydroxy gas at about 8 to 10 amps. I'm using direct DC with a 5-cell, 6 plate
array, similar to a "Smack's Booster", but with better plate isolation in the bath. We've spent thousands
developing fancier, and slightly more efficient cells. We've used PWMs to get better production, and be able to
attenuate gas production with duty cycle. We had a really, really advanced system. Then I applied Occam's razor
to it. We can make enough gas to support ultra-lean combustion above fairly low load conditions - so what good
is it to be able to decrease when you barely make enough already? The PWM does help, and is relatively
inexpensive so we kept that component if the customer wants, but we don't change the duty cycle.

The 6 plate "Smacks" style cell works fine. It is small, easy to build, and is efficient enough for the production we
need. Engine control was the biggest issue. I can get great mileage with just a little hydroxy gas, if I can control
what fuel table the ECU looks at every load range, and rpm. The problem with EFIEs, and MAP/MAF adjusters is
that they tell the computer to look at inappropriate fuel tables at higher load ranges. The ECU picks up on that,
adjusts it's long-term fuel trims, and goes back to an unmodified state. If I can make the ECU look at very lean
tables at cruise, and then more or less unmodified tables at higher loads it never "sees" the trick. Since we only
10 - 23
make enough hydroxy gas to affect low loads anyway, that is all I need to be concerned with. You could think of it
as an ultra-lean cruise mode: when you aren't at low cruise you aren't changing anything. When you are cruising
you are running at a very lean Air-Fuel Ratio. So far, it works great.

The super fancy system that used a PWM with a duty-cycle controlled by our chip, and made up to 3 lpm at
20amps would have cost over USD $2500 just for the parts, and equipment to cover production cost, and turn a
profit, and it only gained us about 10% efficiency. The system we are working on now should be less than USD
$1500 as a 'turn-key' system. Our little Saturn just came back at 88mpg on a carefully driven run with this system.
It typically gets high 40s to mid 50s in mpg under normal driving conditions.

I have tried adding just 0.6 lpm or so, and letting the ECU trim out to compensate. This has given me mixed
results. Sometimes I can get 25% to 30% reduction in consumption, and sometimes it makes no difference at all.
It has a lot to do with the ECU's programming, and the driver's habits. I don't really know why it doesn't work all
the time theoretically it should. The hydroxy gas makes the petrol more volatile so you should be getting more
energy per gram of fuel. That should correlate to higher exhaust gas temperatures, and the ECU should see that
and take away some fuel, but sometimes it does just the opposite. The ECU sees a lean condition due to
increased Exhaust Gas Temperature at the manifold, and lower temperature at the catalytic converter, and so it
richens up the air-fuel mix.

Another possible option that I have not explored would be an EFIE designed to change it's output to a set voltage
controlled by the vehicle's throttle position sensor. The challenge here is that it is not a linear change. The steps
between load sites would not be equal. They would need to be able to be manually set for what the application
needed. The DFAs allow you to do this, and can add voltage just like an EFIE. You can use one DFA for
MAP/MAF control, and one for oxygen sensor control. So even with a modified EFIE you would need a DFA or
something similar to provide non-linear MAP/MAF control". You can contact Les at
lespearson(at)hotmail(dot)com.

Construction:
If you wish to build an oxygen sensor controller circuit, then here is a suggestion as to how you might do it. This
description assumes very little knowledge on the part of the reader, so I offer my apologies to those of you who
are already expert in these matters. There are many different ways to design and construct any electronic circuit
and each electronics expert will have his own preferred way. In my opinion, the way shown here is the easiest for
a newcomer to understand and build with the minimum of tools and materials.

This circuit can be constructed on a printed circuit board or it can be built on a simple single-sided stripboard as
shown here:

Stripboard (often called “Veroboard”), has copper strips attached to one side of the board. The copper strips can
be broken where it is convenient for building the circuit. Component leads are cut to length, cleaned, inserted
from the side of the board which does not have the copper strips, and the leads attached to the copper strips
using a solder joint. Soldering is not a difficult skill to learn and the method is described later in this document.

When all of the components have been attached to the stripboard and the circuit tested, then the board is
mounted in a small plastic case as shown here:

10 - 24
Insulating posts can be made from a short pieces of plastic rod with a hole drilled through its length. The
mounting bolt can self-tap into a hole drilled in the case, if the hole is slightly smaller than the diameter of the bolt
threads. Alternatively, the holes can be drilled slightly larger and the bolt heads located outside the case with nuts
used to hold the board in place. This style of mounting holds the circuit board securely in place and gives some
clearance between the board and the case.

You will need building equipment, namely, a soldering iron, a 12 volt power supply such as a battery pack and an
accurate digital volt meter for this project. If the 12 volt supply is a main-powered unit, then it needs to be a well-
filtered, voltage-stabilised unit. Lastly, you will need a variable voltage source that can go from 0 to 1 volt to
imitate the output from the vehicle’s oxygen sensor when testing the completed circuit board. This is simple
enough to make, using a resistor and a variable resistor.

A series of components will be needed for the circuit itself. These can be bought from a number of different
suppliers and the ordering details are shown later in this document. Shown above is a resistor. The value of the
resistor is indicated by a set of three colour bands at one end of the body. The reason for doing this rather than
just writing the value on the resistor, is that when the resistor is soldered in place, its value can be read from any
10 - 25
angle and from any side. The component list shows the colour bands for each of the resistors used in this circuit.
If you want more information on basic electronics, then read the Electronics Tutorial which can be found at
http://www.free-energy-info.com /Chapter12.pdf

Other components which you will be using, look like this:

The MPSA14 and the BC327 devices are transistors. They each have a “Collector”, a “Base” and an “Emitter”
wire coming out of them. Please notice that the two packages are not identical, and take care that the right wire is
placed in the correct hole in the stripboard before soldering it in place.

The 1N4007 diode has a ring marked at one end of the body. The ring indicates the flat bar across the symbol as
shown on the circuit diagram, and in that way it identifies which way round the diode is placed on the stripboard.

The Light-Emitting Diode (the “LED”) will be familiar to most people as it is used so extensively in equipment of all
types.

The toggle switch has six contacts - three on each side. The centre contact is connected to one of the two outer
contacts on its side, which one, depends on the position of the switch lever.

The two capacitors (which are called “condensers” in very old literature) look quite different from each other. The
electrolytic capacitor has it’s + wire marked on the body of the capacitor, while the ceramic has such a small value
that it does not matter which way round it is connected.

The main component of the circuit, is an integrated circuit or “chip”. This is a tiny package containing a whole
electronic circuit inside it (resistors, capacitors, diodes, whatever, ....). Integrated circuit chips generally look like
this:

A very common version of this package has two rows of seven pins each and it goes by the grandiose name of
“Dual In Line” which just means that there are two rows of pins, each row having the pins in a straight line. In our
particular circuit, the chip has eighteen pins, in two rows of nine.

Now to the circuit itself. If you find it hard to follow, then take a look at the electronics tutorial on the web site as it
shows the circuit diagram symbol for each component and explains how each device works.
10 - 26
The circuit contains three capacitors, eight resistors, two diodes, one LED, one IC chip, two transistors, one toggle
switch and two types of component not yet described, namely: two preset resistors and one rotary switch.

The preset resistor is very small and is adjusted using a flat bladed screwdriver. It is used for making an
adjustable setting which is then left unchanged for a long time. The Rotary switch has a central contact which is
connected to a row of outer contacts in turn when the shaft is rotated from position to position. The switch shaft is
made of plastic and so can easily be cut to the length needed to make a neat installation, and the knob is locked
in place by tightening its grub screw against the flat face of the shaft, although some knobs are designed just to
push tightly on to the shaft. There is a wide range of knob styles which can be used with this switch, so the
choice of knob is dictated by personal taste.

This is the circuit diagram:

Electronic circuits are normally “read” from left to right, so we will look at this circuit that way. The first
components are the 100 microfarad, 35 volt electrolytic capacitor with the tiny 100 nF capacitor across it. These
are put there to help iron out any variations in the voltage supply. The BZX85C zener diode is a 24-volt type and
it protects the integrated circuit from voltage spikes coming along the +12-volt line from other equipment in the

10 - 27
vehicle, preventing the circuit from being fed more than 24 volts for even a fraction of a second as that would
damage the integrated circuit.

The next item is the On/Off dashboard switch. When switched to its Off position as shown here:

the connection from the oxygen sensor is passed straight through to the vehicle’s fuel computer, bypassing the
circuit board completely. This switch allows the whole circuit to be switched Off should you want to do this for any
reason.

In it’s On position, as shown in the circuit diagram, the varying voltage signal coming from the oxygen sensor is
passed into the circuit, and the output voltage from the circuit is passed back to the fuel computer, instead of the
original sensor voltage. This allows the circuit to manipulate the voltage sent to the fuel computer.

The next set of components (four resistors, one ceramic capacitor and one preset resistor) shown here:

are needed to feed the incoming sensor voltage to the Integrated Circuit chip, and make the chip operate in the
way that we want, (the chip manufacturer allows more than one way for the chip to work). You can just ignore
these components for now, just understand why they are there.

The Integrated Circuit chip has ten outputs, coming out through Pins 1 and 10 through 18 inclusive:

If the input voltage coming from the oxygen sensor is low, then all of these ten outputs will have low voltages on
them. When the input voltage rises a little, the voltage on Pin 10 suddenly rises to a high value, while the other
output pins still have low voltages.

If the input voltage rises a little higher, then suddenly the voltage on Pin 11 rises to a high value. At this point,
both Pin 10 and Pin 11 have high voltage on them and the other eight output pins remain at low voltage.

10 - 28
If the input voltage rises a little higher again, then suddenly the voltage on Pin 12 rises to a high value. At this
point, Pin 10, Pin 11 and Pin 12 all have high voltage on them and the other seven output pins remain at low
voltage.

The same thing happens to each of the ten output pins, with the voltage on Pin 1 being the last to get a high
voltage on it. The circuit is arranged so that Pin 10 provides the output signal for the richest air/fuel mixture for
the vehicle, and the mix gets progressively leaner as the output on Pins 11, 12, ... etc. are selected to be fed to
the fuel computer.

As there is the possibility of engine damage if the fuel mix is too lean, only six of the outputs are taken on into the
circuit. However, if the engine is being fed hydroxy gas from an electrolyzer to improve both the miles per gallon
performance and reduce emissions to zero, then it is likely that the engine will run cooler than before and engine
damage is most unlikely to occur. It is quite safe to leave the remaining output pins of the Integrated Circuit chip
unconnected. However, if this unit is to be used with the Nitrogen Hydroxide cell described in the D18.pdf
document, then it is quite safe to connect Pins 16, 17, 18 and 1 and set the rotary switch to ten positions.

The output pin to be used by the remainder of the circuit is selected by the rotary switch mounted on the
dashboard:

A standard single-pole rotary wafer switch has twelve positions but the switch operation can be restricted to any
lesser number of positions by placing the end-stop lug of the switch just after the last switch position required.
This lug comes as standard, fits around the switch shaft like a washer, and is held in place when the locking nut is
tightened on the shaft to hold the switch in place. The lug projects down into the switch mechanism and forms an
end-stop to prevent the switch shaft being turned any further. With six switch positions, the circuit provides five
levels of leaner air/fuel mix which can be selected. This should be more than adequate for all practical purposes.

The next section of the circuit is the BC327 transistor amplifier stage which provides the output current for the fuel
computer:

Here, the switch “SW1” connects to one of the output pins of the Integrated Circuit. When the voltage on that pin
goes low, it causes a current to flow through the transistor Base/Emitter junction, limited by the 2.7K (2,700 ohm)
resistor. This current causes the transistor to switch hard On, which in turn alters the voltage on its Collector from

10 - 29
near 0 volts to near +12 volts. The 2.7K resistor is only there to limit the current through the transistor and to
avoid excessive loading on the output pin of the IC.

The transistor now feeds current to the LED via the two 1N4007 diodes and the 1K (1,000 ohm) resistor. This
causes the Light Emitting Diode to light brightly. The 1K resistor is there to limit the amount of current flowing
through this section of the circuit.

Part of the voltage across the LED is fed back to the fuel computer:

By moving the slider contact on the preset resistor “VR2”, any output voltage can be fed to the fuel computer.
This voltage can be anything from the whole of the voltage across the LED, down to almost zero volts. We will
use VR2 to adjust the output voltage when we are setting the circuit up for use. In this circuit, VR2 is acting as a
“voltage divider” and it is there to allow adjustment of the output voltage going from the circuit to the fuel
computer.

The final section of the circuit is the MPSA14 transistor and its associated components:

This circuit is a timer. When the circuit is first powered up (by the vehicle’s ignition key being turned), the 470
microfarad capacitor “C1” is fully discharged (if it isn’t, then the oxygen sensor will already be hot). As it is
discharged and one side is connected to the +12 volt line, then the other side (point “A”) looks as if it is also at
+12 volts. This provides a tiny current to the Base/Emitter junction of the MPSA14 transistor, through the high
resistance 470K (470,000 ohm) resistor. The MPSA14 transistor has a very high gain and so this tiny current
causes it to switch hard on, short-circuiting the LED and preventing any voltage developing across the LED.

As time passes, the tiny current flowing through the MPSA14 transistor, along with the tiny current through the
3.9M (3,900,000 ohm) resistor “R1”, cause a voltage to build up on capacitor “C1”. This in turn, forces the voltage
at point “A” lower and lower. Eventually, the voltage at point “A” gets so low that the MPSA14 transistor gets
starved of current and it switches off, allowing the LED to light and the circuit to start supplying an output voltage
to the fuel computer. The purpose of the section of the circuit is to shut off the output to the fuel computer until the
oxygen sensor has reached it’s working temperature of 600 degrees Fahrenheit. It may be necessary to tailor this
delay to your vehicle by altering the value of either “R1” or “C1”. Increasing either or both will lengthen the delay
while reducing the value of either or both, will shorten the delay.

We want the time delay to occur if the engine is off for some time, but not to occur if the engine is switched off
only briefly. For this to happen, it is suggested that a diode is placed across the timing resistor. This will have no
effect when the circuit is powered up, but it will discharge the capacitor when the circuit is powered down. We can
slow down the rate of discharge by putting a high-value resistor in series with the discharge diode and that would
make the circuit:

Circuit Operation:
Now that we have looked at each part of the circuit separately, let us look again at the way that the circuit
operates. The main component is the LM3914 integrated circuit. This device is designed to light a row of Light
10 - 30
Emitting Diodes (“LEDs”). The number of LEDs lit is proportional to the input voltage reaching it through it’s Pin 5.
In this circuit, the integrated circuit is used to provide a reduced voltage to be fed to the fuel computer, rather than
to light a row of LEDs. When the operating switch is set in it’s ON position, the sensor voltage is fed to Pin 5
through a 1 megohm resistor.

The sensitivity of this circuit is adjusted, so that when 500 millivolts (0.5 volts) is applied to Pin 5, the output on Pin
10 is just triggered. This is done by adjusting the 10K linear preset resistor “VR1” while placing a test voltage of
500 millivolts on Pin 5. This LM3914 Integrated Circuit is normally switched so that it samples the sensor voltage.
The LM3914 chip provides ten separate output voltage levels, and the circuit is arranged so that any one of
several of these can be selected by the rotary switch “SW1”. These output voltages range from 50 millivolts on
Pin 1 to 500 millivolts on Pin 10, with each output position having a 50 millivolt greater output than it’s
neighbouring pin. This allows a wide range of control over the sensor feed passed to the fuel computer.

The input resistor/capacitor circuit provides filtering of the sensor signal. Because this circuit draws very little
current, it is easily knocked out of correct operation through it’s input line picking up stray electrical pulses
produced by the engine, particularly the vehicle’s ignition circuit. When the exhaust sensor heats up, the signal
becomes cleaner and then the circuit starts operating correctly. The circuit includes a delay so that after start up,
the output is held low for a few minutes to simulate a cold sensor. The sensor must be operating correctly before
we send signals to the computer. The most common problem, if we don’t have this delay, is that the output will be
high simply from the noise on the signal line. The computer will think the sensor is working, because it is high, and
will cut back the fuel to make the signal go low. If that were to happen, we would end up with an over-lean fuel
input to the engine, producing very poor acceleration.

The front panel LED is not just to show that the device is operating, but forms a simple voltage regulator for the
output signal to the computer. When the engine is warmed up and running normally, the LED is lit when the
output is high, and not lit when the output is low, so this LED should be flashing on and off.

The earth connection for the oxygen sensor is the exhaust system, which is firmly bolted to the engine. The
computer earth is the vehicle body. A difference of just 0.5 volts can make a large difference to the mixture. If the
engine is not securely earthed to the vehicle body, then a voltage difference can exist between the two, and in this
situation a voltage difference of just 0.5 volts would normally go unnoticed. We can’t afford to have that sort of
voltage difference when trying to control the mixture accurately, so some investigation and adjustment is needed.

To do this, start the engine, switch the headlights on to high beam, then measure the voltage between the engine
and the body. Use a digital volt meter. Any more than 50 millivolts (0.05 volts) means that there is a bad earth
connection which need cleaning and tightening. Modern cars usually have more than one connection so look
around. If you have trouble achieving a really good connection, then earth your circuit board directly on the
engine rather than connecting it to a point on the bodywork of the vehicle. The most important item is to have a
good quality signal voltage coming from the sensor, since the operating range consists of quite low voltages. The
components and tools needed for building this circuit are shown later, but for now, consider the setting up and
testing of the unit so as to understand better what is needed.

Adjusting on the Bench:

When the circuit has been constructed to the testing stage, that is, with all components in place except for the
timing capacitor “C1”, and before the power is turned on, plug the Integrated Circuit chip into its socket mounted
on the board. Be very careful doing this as the chip can be destroyed by static electricity picked up by your body.
Professionals wear an electrical earth wrist strap when handling these devices, so it would be a good idea to
touch a good earth point such as a metal-pipe cold water system just before handling the chip.

It is vital that you install the IC chip, the correct way round or it may be damaged. The circuit board layout shows
which way round it goes. The chip has a semi-circular indentation at one end to show which end is which, so be
careful that the indentation is positioned as shown on the board layout in the section which shows how the board
is built. Some manufacturers use a dot rather than a semi-circular indentation to mark the end of the chip which
has Pin 1 in it.

10 - 31
Make up the test voltage device. We need something to give us an adjustable voltage in the range 0 to 1 volt. A
very easy way to get this is to use a 10K resistor and a 1K variable resistor (called a “potentiometer” by some
people) and connect them across the 12 volt battery, as shown here:

This gives us a voltage in the correct range when the shaft of the variable resistor is turned. Power up the circuit
board by switching the 12 volt battery through to the board. Adjust the test-voltage source to 500 millivolts (0.5
volts) and apply it to the board’s input (where the sensor connection will be made when it is installed in the
vehicle). Set the switch to the “Richest” position, that is, with the switch connected to Pin 10 of the chip.

Now, using a flat-blade screwdriver, adjust the sensitivity control preset resistor “VR1” so that the output LED is
just lit. Leave the preset resistor in that position and adjust the test voltage lower and higher to test that the LED
turns on and off in response to the varying voltage at the input to the circuit. The LED should come on at 0.5
volts, and go off just below 0.5 volts. The other outputs, which can be selected by the rotary switch “SW1”, will be
about 50 millivolts lower for each position of the switch away from it’s “Richest” setting on Pin 10.

Now, with the output high and the LED lit, use a flat-bladed screwdriver to adjust the preset resistor “VR2” to set
the output voltage being sent to the computer to about 1.0 volts. When this has been set, lower the input voltage
so that the LED goes out. The output voltage should now be at zero volts. If this is what happens, then it shows
that the circuit is operating correctly.

If this board is not in place, the sensor will cause the fuel computer to make the fuel mixture richer so as to
maintain a 500 millivolt voltage from the sensor. With the circuit in place and set to its “Richest” setting, exactly
the same thing happens. However, if the rotary switch is moved to its next position, the fuel computer will
maintain the fuel feed to maintain a 450 millivolt output, which is a leaner fuel-to-air mixture. One step further
around and the fuel computer will make the mix even leaner to maintain a 400 millivolt output from the circuit
board, which the fuel computer thinks is coming from the exhaust oxygen sensor.

If your circuit board does not operate as described, then power it down and examine the circuit board again,
looking for places where the solder connections are not perfect. There may be somewhere where the solder is
bridging between two of the copper strips, or there may be a joint which looks as if it is not a good quality joint. If
you find one, don’t solder anywhere near the IC chip as the heat might damage the chip. If necessary, earth
yourself again, remove the chip and put it back into the anti-static packaging it came in, before repairing the
board. If the components are all correctly positioned, the copper tracks broken at all the right places and all
solder joints looking good and well made but the board still is not working correctly, then it is likely that the IC chip
is defective and needs to be replaced.

Next, install the delay capacitor “C1”. Set the test voltage above 500 millivolts and turn the power on again. It
should take about three minutes for the LED to come on. If you want to shorten this delay, then change the timing
resistor “R1” for a resistor of a lower value. To lengthen the delay, replace the timing capacitor “C1” with a
capacitor of larger value. If you find that the oxygen sensor heats up quickly, then you can reduce the length of
the delay. Having too long a delay is not ideal, since the computer will be adding extra fuel to make the mixture
richer.

It is suggested that the rotary switch should be set to have only six switch positions (by moving it’s end-stop lug
washer), so initially, connect the IC chip output pins 10 through 15 to the switch. You can choose to connect the
wires to the switch so that the mixture gets richer when you turn the knob clockwise, or if you prefer, you can wire
it in the reverse order so that the mixture gets richer when you turn the knob counter-clockwise.

Testing in the Car:

You can now test the device in the vehicle but don’t install it yet. Look in the engine compartment and locate the
oxygen sensor. If you have difficulty in finding it, get a copy of the Clymer or Haynes Maintenance Manual for
10 - 32
your vehicle as that will show you the position. If your vehicle has two sensors, then select the one nearest to the
engine. If your sensor has five wires running to it, then it is a “wideband” sensor which measures both the
oxygen content and the amount of unburnt fuel, and unfortunately, the type of circuit described here will not
control it.

Start the vehicle and allow the oxygen sensor to warm up for a couple of minutes. Remember that there is a
delay built in to the circuit, so after a few minutes you should see the LED start to flash. Rev the engine and the
LED will stay on. When you release the throttle, the LED will go out for a while. A flashing LED is what you want
to see. The rate of flashing will be somewhere between 1 and 10 times per second, most likely around 2 per
second. Confirm that the LED goes out when you switch off the circuit board On/Off switch mounted on the
dashboard.

Now comes the exciting bit, cutting the oxygen sensor wire and inserting the controller. Turn the engine off and
cut the wire in a convenient place. Use crimp connectors on the wire ends. Use a matching pair on the wire
which you just cut, in case you need to reconnect it, as shown here:

When set up like this, the male connector furthest on the left could be plugged into the female connector furthest
on the right and the circuit board removed. Be sure to insulate the sensor and fuel computer plug/socket
connections to make quite sure that neither of them can short-circuit to any part of the body. There is no need to
insulate the earth connection as it is already connected to the body of the vehicle. Although not shown in the
diagram, you could also put a male and female crimp connector pair on the earth cable. If your sensor has only
one wire coming from it, then you best earth connection is to a solder-tag connector placed under a bolt on the
engine. If you do that, be sure to clean all grease, dirt, rust, etc. off the underside of the bolt head and the area
around the bolt hole. Push a paper towel into the bolt hole before doing this to make sure that no unwanted
material ends up in the bolt hole and use wet-and-dry paper to really clean the surfaces. The objective here is to
make sure that there is a very good electrical connection with shiny metal faces clamped firmly together.

Installing the Controller:

Now, install the circuit board in the vehicle. For the 12 volt supply, find a connection which is switched on and off
by the vehicle’s ignition switch. Don’t drive the car yet, do this test in the driveway. With the front panel switch in
it’s “Off” position, start the car and check that it runs normally. Set the front panel rotary switch to the Richest
position (connected to the IC’s Pin 10) and switch the circuit board toggle switch to it’s “On” position. The car is
now running with a modified oxygen sensor signal although the mixture is still the same. The vehicle performance
should be completely normal. Drive the vehicle with this setting for a while to prove that the system is working
reliably before changing to any of the lower settings. When you are satisfied that everything is in order, try the
next leanest setting on the rotary switch and see how it runs.

It is important that there should be no hesitation in the engine performance and no knocking or “pinking” as that is
an indication that the mix is too lean and the engine is liable to overheat. This circuit is intended for use with an
electrolyzer, so your electrolyzer should be set up and working for these tests. The electrolyzer will tend to make
the engine run cooler and offset any tendency towards overheating.

Building the Circuit Board:

Although the above information has been presented as if the board has already been built, the actual construction
details have been left until now, so that you will already have an understanding of what the circuit is intended to
do and how it is used.

It is likely that you will know somebody (neighbour, friend, relative,...) who has the necessary equipment and
skills. If so, borrow the equipment, or better still, recruit the person to help with the construction. It is very likely
that anybody owning the equipment would be very interested in your project and more than willing to help out.

10 - 33
However, the rest of this document will be written on the assumption that you cannot find anybody to help and
have had to buy all of the necessary equipment. This project is not difficult to build, so you will almost certainly be
successful straight off.

The tools which you will need, are:

1. A soldering iron with a fine conical tapering tip (probably 15 watts power rating)
2. Some “Multicore” resin solder. This is special solder for electronics construction work and is quite different from
plumber’s solder which is not suitable for this job.
3. A pair of long-nosed pliers (for holding component wires when soldering them in place)
4. Something for cutting and cleaning wires and stripping off insulation coverings. I personally prefer a pair of
“nail” scissors for this job. Others prefer a pair of wire cutters and some sandpaper. You get whatever you
feel would be the best tool for doing these tasks.
5. A 1/8 inch (3 mm) drill bit (for making bolt holes in the stripboard and for breaking the copper strips where
needed) and a 3/8 inch (9 mm) drill and bit for mounting the switches on the plastic box.
6. A coping-saw or similar small saw for cutting the rotary switch shaft to the optimum length.
7. A small screwdriver (for tightening knob grubscrews).
8. A crimping tool and some crimp connectors.
9. A multimeter (preferably a digital one) with a DC voltage measuring range of 0 to 15 volts or so.
10. (Optional) a magnifying glass of x4 or higher magnification (for very close examination of the soldering)

Soldering:
Many electronic components can be damaged by the high temperatures they are subjected to when being
soldered in place. I personally prefer to use a pair of long-nosed pliers to grip the component leads on the upper
side of the board while making the solder joint on the underside of the board. The heat running up the component
lead then gets diverted into the large volume of metal in the pair of pliers and the component is protected from
excessive heat. On the same principle, I always use an Integrated Circuit socket when soldering a circuit board,
that way, the heat has dissipated fully before the IC is plugged into the socket. It also has the advantage that the
IC can be replaced without any difficulty should it become damaged.

If you are using CMOS integrated circuits in any construction, you need to avoid static electricity. Very high levels
of voltage build up on your clothes through brushing against objects. This voltage is in the thousands of volts
range. It can supply so little current that it does not bother you and you probably do not notice it. CMOS devices
operate on such low amounts of current that they can very easily be damaged by your static electricity. Computer
hardware professionals wear an earthing lead strapped to their wrists when handling CMOS circuitry. There is no
need for you to go that far. CMOS devices are supplied with their leads embedded in a conducting material.
Leave them in the material until you are ready to plug them into the circuit and then only hold the plastic body of
the case and do not touch any of the pins. Once in place in the circuit, the circuit components will prevent the
build up of static charges on the chip.

Soldering is an easily-acquired skill. Multi-cored solder is used for electronic circuit soldering. This solder wire
has flux resin contained within it and when melted on a metal surface, the flux removes the oxide layer on the
metal, allowing a proper electrical and mechanical joint to be made. Consequently, it is important that the solder
is placed on the joint area and the soldering iron placed on it when it is already in position. If this is done, the flux
can clean the joint area and the joint will be good. If the solder is placed on the soldering iron and then the iron
moved to the joint, the flux will have burnt away before the joint area is reached and the resulting joint will not be
good.

A good solder joint will have a smooth shiny surface and pulling any wire going into the joint will have no effect as
the wire is now solidly incorporated into the joint. Making a good solder joint takes about half a second and
certainly not more than one second. You want to remove the soldering iron from the joint before an excessive
amount of heat is run into the joint. It is recommended that a good mechanical joint be made before soldering
when connecting a wire to some form of terminal (this is often not possible).

The technique which I use, is to stand the solder up on the workbench and bend the end so that it is sloping
downwards towards me. The lead of the component to be soldered is placed in the hole in the stripboard and
gripped just above the board with long-nosed pliers. The board is turned upside down and the left thumb used to
clamp the board against the pliers. The board and pliers are then moved underneath the solder and positioned so
that the solder lies on the copper strip, touching the component lead. The right hand is now used to place the
soldering iron briefly on the solder. This melts the solder on the joint, allowing the flux to clean the area and
producing a good joint. After the joint is made, the board is still held with the pliers until the joint has cooled down.

10 - 34
Nowadays, the holes in the stripboard are only 1/10 inch (2.5 mm) apart and so the gaps between adjacent
copper strips is very small indeed. If you solder carefully, there should be no problem. However, I would
recommend that when the circuit board is completed, that you use a magnifying glass to examine the strip side of
the board to make quite sure that everything is perfectly ok and that solder does not bridge between the copper
strips anywhere. Before powering up the circuit, double-check that all of the breaks in the copper strips have
been made correctly. Here is a possible layout for the components on the stripboard:

If this board is turned over horizontally, the underside will look like this:

10 - 35
This shows where the breaks in the copper strips need to be made using a 1/8 inch (3 mm) drill bit.

To construct this circuit, cut a piece of stripboard which has 18 strips, each with 32 holes. That is a board size of
about two inches (50 mm) by just over three inches (85 mm). Mount the components on the board, working from
one end as the installation is easier if you have a clear board to work across. If you are right-handed, then start at
the left hand side of the board and work towards the right, installing all components as you go. If you are left-
handed, then mount the components starting with the right hand side of the board and working towards the left
hand side.

Having said that, it is probably easier if you put all of the wire jumpers in place as the first step. The best wire for
this is solid core wire of the type used in telephone wiring, as it is easy to cut, easy to remove the insulation and it
lies flat on the board, clear of all of the other holes. So, start with the wire jumpers and then install the electronic
components working across the board.

The jumper wires lie flat on the board, and like the other components, have about 2 mm of clean wire projecting
through the copper strip before the solder joint is made.

The wires coming off the board should be of the type which have several thin wires inside the insulation, as these
are more flexible and withstand the vibration of a vehicle in motion, better than solid core wire. If you have just
one reel of wire, then be sure to label the far end of each piece mounted on the board, the moment you have
soldered it in place. These labels will help avoid errors when mounting in the case, if you do not have different
coloured wires.

The completed circuit board can be mounted in a small plastic box of the type which has a lid held in place by
screws. It may be convenient to screw or bolt the case to the underside of the dashboard and then screw the lid
in place, covering the mounting screws:

10 - 36
The components in this circuit are not critical and any near-match alternatives can be used. In the event that the
MPSA14 Darlington-pair transistor is not available, then two general-purpose high-gain silicon transistors like the
BC109 or 2N2222A can be substituted. Just connect them like this:

The emitter of the first transistor is connected to base of the second and the two collectors are connected
together. If the transistors have metal cases, then make sure the emitter/base connection cannot touch either
case as the cases are often connected internally to the collectors. If each transistor has a gain of only 200, then
the pair will have a combined gain of 40,000 times. That means that the base current need only be 40,000 times
less than the collector current of the second transistor.

The BC327 transistor can be replaced by almost any other silicon PNP transistor in this circuit as the gain does
not need to be great and the power rating is very small. The following is a list of the main electronic components
needed for the construction of this circuit as described here. There are several suppliers who are able to supply
all of these components and the most suitable depends on where you are located. If there is any difficulty, try an
internet search, and if that fails, ask for help in one or more of the Yahoo enthusiast groups such as ‘watercar’,
‘hydroxy’ or any of the electronics Groups.

Component Qty. US Supplier Code

Black plastic box with lid, size about 4” x 3” x 2” 1 Radio Shack 270-1803
Stripboard: 18 strips, 32 holes 1 Electronix Express 0302PB16
Double Pole Double Throw toggle switch 1 Radio Shack 275-636
Fuseholder, panel mounting, 1.25” 1 Radio Shack 270-364
Fuse, 2 amp slow-blow 1.25” 1 Radio Shack 270-1262 ?? (3 A)
Rotary wafer switch, 12-way single pole 1 Electronix Express 17ROT1-12
Knob for the rotary switch 1 Radio Shack 274-424
LED, any colour, 5 mm diameter 1 Radio Shack 276-041
IC socket, 18 pin DIL 1 Radio Shack 276-1992
Miniature preset resistor, 10K linear 2 Radio Shack 271-282
LM3914 LED bar driver Integrated Circuit 1 Electronix Express LM3914
BC327 PNP transistor 1 Electronix Express 2N2905
MPSA14 Darlington pair transistor 1 Electronix Express MPSA14
1N4007 Diode or equivalent 3 Radio Shack 276-1103 (2 pack)
BZX85C zener diode, 24 volt version 1 Electronix Express 1N5359
470 microfarad, 35 volt (or higher) axial lead 1 Radio Shack 272-1018
aluminium foil electrolytic capacitor
100 microfarad, 35 volt (or higher) axial lead 1 Radio Shack 272-1016
aluminium foil electrolytic capacitor
100 nF (0.1 microfarad) ceramic disc capacitor 2 Radio Shack 272-135 (2 pack)
10 megohm 1/4 watt carbon resistor 1 Radio Shack 271-1365 (5 pack)
(Bands: Brown,Black,Blue)
1 megohm 1/4 watt carbon resistor 3 Radio Shack 271-1356 (5 pack)
(Bands: Brown,Black,Green)
470K 1/4 watt carbon resistor 1 (Radio Shack) use two 1M in parallel or
(Bands: Yellow,Purple,Yellow) or 1 Radio Shack 271-1133 (5 pack 1/2 watt)
10K 1/4 watt carbon resistor 1 Radio Shack 271-1335 (5 pack)
(Bands: Brown,Black,Orange)
2.7K 1/4 watt carbon resistor 1 Radio Shack 271-1328 (5 pack)
(Bands: Red,Purple,Red) [use 3.3K]
1K 1/4 watt carbon resistor 2 Radio Shack 271-1321 (5 pack)
(Bands: Brown,Black,Red)
100 ohm 1/4 watt carbon resistor 1 Radio Shack 271-1311 (5 pack)
(Bands: Brown,Black,Brown)
Connecting wire: stranded and solid core Local supplier

10 - 37
Electronix Express http://www.elexp.com/index.htm
Radio Shack http://www.radioshack.com/home/index.jsp

And for a UK supplier:

Component Qty. European Supplier Code

Black plastic box with lid, size about 4” x 3” x 2” 1 ESR 400-555
Stripboard: 18 strips, 32 holes 1 ESR 335-010
Double Pole Double Throw toggle switch 1 ESR 218-028
Fuseholder, panel mounting 31 mm 1 ESR 187-115
Fuse, 2 amp 31 mm 1 ESR 190-220
Rotary wafer switch, 12-way single pole 1 ESR 210-012
Knob for the rotary switch 1 ESR 060-22X
LED, any colour, 5 mm diameter 1 ESR 711-540
IC socket, 18 pin DIL 1 ESR 110-180
Miniature preset resistor, 10K linear 2 ESR 998-310
LM3914 LED bar driver Integrated Circuit 1 ESR LM3914
BC327 PNP transistor 1 ESR BC327
MPSA14 Darlington pair transistor 1 ESR MPSA13
1N4007 Diode or equivalent 3 ESR 1N4007
BZX85C zener diode, 24 volt version 1 ESR 726-240
470 microfarad, 35 volt (or higher) axial lead 1 ESR 810-104
aluminium foil electrolytic capacitor
100 microfarad, 35 volt (or higher) axial lead 1 ESR 810-096
aluminium foil electrolytic capacitor
100 nF (0.1 microfarad) ceramic disc capacitor 2 ESR 871-061
10 megohm 1/4 watt carbon resistor 1 ESR 906-610
(Bands: Brown,Black,Blue)
1 megohm 1/4 watt carbon resistor 3 ESR 906-510
(Bands: Brown,Black,Green)
470K 1/4 watt carbon resistor 1 ESR 906-447
(Bands: Yellow,Purple,Yellow)
10K 1/4 watt carbon resistor 1 ESR 906-310
(Bands: Brown,Black,Orange)
2.7K 1/4 watt carbon resistor 1 ESR 906-227
(Bands: Red,Purple,Red)
1K 1/4 watt carbon resistor 2 ESR 906-210
(Bands: Brown,Black,Red)
100 ohm 1/4 watt carbon resistor 1 ESR 906-110
(Bands: Brown,Black,Brown)
Reel of multi-strand connecting wire 6 amp Red 1 ESR 054-112
Reel of multi-strand connecting wire 6 amp Blue 1 ESR 054-116
Reel of solid core (or local phone wire) 1 ESR 055-111

ESR http://www.esr.co.uk Tel: 01912 514 363

While the components listed above are the parts needed to construct the electronics board, the following items
may be needed in addition when testing and installing the board in a vehicle:

Component Use
Rubber or plastic grommets To protect wires from rubbing against the edges of the holes in the box
Crimp “bullet” connectors Male and female, one pair for each sensor wire cut
Mounting bolts, nuts and spacers To hold the circuit board securely, clear of the box.
Double-sided adhesive tape For mounting the box on the dash. Alternatively, hardware items for this.
Fuse-box connector For connecting to the fuse box to give an ignition-switched 12V supply
10K resistor and 1K Linear For bench testing with voltages of up to 1 volt, if these components are not
variable resistor already to hand
Multimeter For general checking of voltages, continuity, etc.

I should like to express my sincere thanks to the various members of the ‘watercar’ Group who provided the
technical information and patient support which made this document possible.

An alternative: As the signal coming from the oxygen sensor to the vehicle’s ECU fuel computer needs to be
raised slightly to allow for the much cleaner exhaust produced when a booster is being used, an alternative
solution has been suggested and tested. The idea is to add a small, adjustable voltage to the signal already
10 - 38
coming from the oxygen sensor. This voltage can be from a single ‘dry-cell’ battery and adjusted with a variable
resistor:

The circuit shown here allows a voltage anywhere from zero to 0.5 volts to be added to the oxygen sensor signal.
This must not be done unless a booster is running. Using it without a booster is liable to lead to engine
overheating and possible valve damage. This, of course, applies to the previous oxygen sensor signal adjusting
circuit as well.

Please Note: This document has been prepared for information purposes only and must not be construed
as an encouragement to build any new device nor to adapt any existing device. If you undertake any kind
of construction work, then you do so entirely at your own risk. You, and only you, are responsible for
your own actions. This document must not be seen as an endorsement of this kind of adaptation nor as
providing any kind of guarantee that an adaptation of this kind would work for you personally. This
document merely describes what has been achieved by other people and you must not consider it as
being a foolproof blueprint for replication by anyone else.

The “Zach West” Electrolyser.

Zach West of the USA has produced a motorcycle electrolyser. Zach’s 250 cc motorcycle can run on the output
of his electrolyser and Zach estimates the output as being 17 litres per minute of HHO gas, which seems to me to
be far too high for the current flow. This is not a COP>1 system as the output from the electrical system of the
motorcycle is very limited, and so the battery will slowly run down as time goes by. However, Zach’s design of
electrolyser is interesting, both for it’s simplicity and it’s high rate of gas output. The increased gas volume which
would be produced if this design were adapted for, and driven by, a 12-volt input could be very useful, especially if
combined with David Quirey’s system which allows the resulting modified gas to operate in unmodified engines,
as shown later on in this chapter.

The method which Zach uses is somewhat unusual as he manages to bleed off and discard most of the oxygen
produced. This means that the remaining gas is mainly hydrogen which is far less reactive than HHO which is
already in the perfect proportions for combination back into water and so is highly reactive. Instead, the resulting
gas can be compressed reasonably well, and Zach compresses it to 30 psi (pounds per square inch) in a storage
container. This helps with acceleration from stationary at traffic lights.

Zach uses a simple, modular style of construction where a series of coiled electrode pairs are each placed inside
an individual length of plastic pipe. This is a design which is neither difficult nor particularly expensive to build. In
overall broad outline, Zach’s electrolyser is fed water from a water tank to keep it topped up. The electrolyser box
contains several pairs of electrodes which split the water into hydrogen and oxygen when fed with pulsed
electrical current generated by the electronics, which is powered by the electrical system of the motorcycle. The
gas produced by the electrolyser is fed to a dual-purpose bubbler, which prevents any accidental igniting of the
gases from travelling back to the electrolyser and in addition, removes most of the oxygen from the gas by acting
as a gas “separator”. The arrangement is like this:

10 - 39
The hydrogen gas output from the electrolyser is not fed directly to the engine but instead it goes to a pressure
tank which is allowed to build up to thirty pounds per square inch before the engine is started. The majority of the
oxygen produced by the electrolysis is vented away through a 30 psi one-way valve which is included to keep the
pressure inside the bubbler (and the electrolyser) at the 30 psi level. That pressure is excessive for a high-
performance electrolyser which produces HHO which is highly charged electrically and so will ignite
spontaneously when compressed, due to it’s own electrical charge. In this simple DC electrolyser, the HHO gas
is mixed with quite an amount of water vapour which dilutes it and allows some compression.

The water supply system operates by having an air-tight supply tank positioned at a higher level than the
electrolyser. A small diameter (1/4” or 6 mm) plastic tube coming from the supply tank feeds through the top of
the electrolyser and straight down, terminating at exactly the electrolyte surface level wanted in each of the
electrolyser tubes. When the electrolysis lowers the electrolyte level below the bottom of the pipe, bubbles of gas
pass up the tube allowing some water to flow from the tank to raise the electrolyte surface level back to it’s design
position. This is a very neat passive system needing no moving parts, electrical supply or electronics but yet one
which accurately controls the electrolyte level. One essential point to understand is that the water tank needs to
be rigid so that it will not flex and the filler cap needs to be air-tight to prevent the entire water supply discharging
into the electrolyser. Another point to remember when topping up the water tank is that the tank contains a mix of
air and HHO gas above the water surface and not just plain air, and that gas mix is at 30 psi pressure.

Now, to cover the design in more detail. This 6-volt electrolyser contains eight pairs of electrodes. These
electrode pairs are coiled around in “Swiss-roll” style and inserted into a length of 2 inch (50 mm) diameter plastic
pipe, ten inches (250 mm) tall. The electrodes are each made from a 10 inch (250 mm) by 5 inch (125 mm) of
316L-grade stainless steel shimstock which is easy to cut and work. Shimstock is available from a local steel
supplier or metal fabrication company and is just a sheet of very thin metal.

Each electrode is cleaned carefully, and wearing rubber gloves, cross-scored using coarse sandpaper in order to
produce a very large number of microscopic mountain peaks on the surface of the metal. This increases the
surface area and provides a surface which makes it easier for gas bubbles to break away and rise to the surface.
The electrodes are rinsed off with clean water and then coiled round, using spacers to maintain the necessary
inter-plate gap, to form the required shape which is then inserted into a length of plastic pipe as shown here:

10 - 40
As the springy metal pushes outwards in an attempt to straighten up again, spacers are used to keep the
electrodes evenly separated along their whole length by inserting 1/8” thick vertical spacer strips. The
connections to the plates are made by drilling a hole in the corner of the plate and inserting the wire several times
through the hole, twisting it back around itself and making a wire-to-wire solder joint on both sides of the steel.
The joint is then insulated with silicone or any other suitable material. It is, of course, essential that the joint does
not short-circuit to the other electrode even though that electrode is very close by.

It is always difficult to make a good electrical connection to stainless steel plates if space is restricted as it is here.
In this instance, the electrical wire is wrapped tightly through a drilled hole and then soldered and insulated. The
soldering is only on the wire as solder will not attach to stainless steel.

10 - 41
An unusual feature of this design is that each of the electrode pairs is effectively a separate electrolyser in its own
right as it is capped top and bottom, and effectively physically isolated from the other electrodes. The water feed
comes through the top cap which has a hole drilled in it to allow the gas to escape. The electrical wires (#12
AWG or swg 14) are fed through the base and sealed against leakage of electrolyte. Each of these units has
some electrolyte stored above it, so there is no chance of any part of the electrode surface not being able to
generate gas. There is also a large amount of freeboard to contain splashes and sloshing without any being able
to escape from the container. The end caps are standard PVC caps available from the supplier of the PVC
piping, as is the PVC glue used to seal them to the pipe.

Eight of these electrodes are placed in a simple electrolyser case and connected together in pairs as shown here:

Pairs of pipe-enclosed electrode spirals are then connected in a chain inside the electrolyser as shown here:

Many years of experimentation and testing have shown that 316L-grade stainless steel is the most suitable
material for electrodes, but surprisingly, stainless steel is not highly electrically conductive as you would expect.
Each electrode causes a voltage drop of nearly half a volt, and so careful surface preparation, cleansing and
conditioning are needed to get top performance from the electrodes. This process is described in detail by the
very experienced Bob Boyce who says:

The preparation of the plates is one of the most important steps in producing an electrolyser which works well.
This is a long task, but it is vital that it is not skimped or hurried in any way. Surprisingly, brand new shiny
stainless steel is not particularly suitable for use in an electrolyser and it needs to receive careful treatment and
preparation before it will produce the expected level of gas output.

10 - 42
The first step is to treat both surfaces of every plate to encourage gas bubbles to break away from the surface of
the plate. This could be done by grit blasting, but if that method is chosen, great care must be taken that the grit
used does not contaminate the plates. Stainless steel is not cheap and if you get grit blasting wrong, then the
plates will be useless as far as electrolysis is concerned. A safe method is to score the plate surface with coarse
sandpaper. This is done in two different directions to produce a cross-hatch pattern. This produces microscopic
sharp peaks and valleys on the surface of the plate and those sharp points and ridges are ideal for helping
bubbles to form and break free of the plate.

When doing hand sanding the sandpaper is drawn across the plates in one direction only and not backwards and
forwards, as the backwards stroke always destroys the perfectly good ridges created on the forward stroke. Also,
you only need two strokes in one direction before turning the plate through ninety degrees and completing the
sanding of that face of the plate with just two more strokes (again, with no backstroke).

Always wear rubber gloves when handling the plates to avoid getting finger marks on the plates. Wearing these
gloves is very important as the plates must be kept as clean and as grease-free as possible, ready for the next
stages of their preparation. Any particles created by the sanding process should now be washed off the plates.
This can be done with clean tap water (not city water though, due to all the chlorine and other chemicals added),
but only use distilled water for the final rinse.

While Potassium hydroxide (KOH) and Sodium hydroxide (NaOH) are the very best electrolytes, they need to be
treated with care. The handling for each is the same:

Always store it in a sturdy air-tight container which is clearly labelled "DANGER! - Potassium Hydroxide". Keep
the container in a safe place, where it can’t be reached by children, pets or people who won't take any notice of
the label. If your supply of KOH is delivered in a strong plastic bag, then once you open the bag, you should
transfer all its contents to sturdy, air-tight, plastic storage containers, which you can open and close without
risking spilling the contents. Hardware stores sell large plastic buckets with air tight lids that can be used for this
purpose.

When working with dry KOH flakes or granules, wear safety goggles, rubber gloves, a long sleeved shirt, socks
and long trousers. Also, don’t wear your favourite clothes when handling KOH solution as it is not the best thing to
get on clothes. It is also no harm to wear a face mask which covers your mouth and nose. If you are mixing solid
KOH with water, always add the KOH to the water, and not the other way round, and use a plastic container for
the mixing, preferably one which has double the capacity of the finished mixture. The mixing should be done in a
well-ventilated area which is not draughty as air currents can blow the dry KOH around.

When mixing the electrolyte, never use warm water. The water should be cool because the chemical reaction
between the water and the KOH generates a good deal of heat. If possible, place the mixing container in a larger
container filled with cold water, as that will help to keep the temperature down, and if your mixture should “boil
over” it will contain the spillage. Add only a small amount of KOH at a time, stirring continuously, and if you stop
stirring for any reason, put the lids back on all containers.

If, in spite of all precautions, you get some KOH solution on your skin, wash it off with plenty of running cold water
and apply some vinegar to the skin. Vinegar is acidic, and will help balance out the alkalinity of the KOH. You can
10 - 43
use lemon juice if you don't have vinegar to hand - but it is always recommended to keep a bottle of vinegar
handy.

Plate cleansing is always done with NaOH. Prepare a 5% to 10% (by weight) NaOH solution and let it cool down.
A 5% solution ‘by weight’ is 50 grams of NaOH in 950 cc of water. A 10% solution ‘by weight’ is 100 grams of
NaOH in 900 cc of water. As mentioned before, never handle the plates with your bare hands, but always use
clean rubber gloves.

A voltage is now applied across the whole set of plates by attaching the leads to the outermost two plates. This
voltage should be at least 2 volts per cell, but it should not exceed 2.5 volts per cell. Maintain this voltage across
the set of plates for several hours at a time. The current is likely to be 4 amps or more. As this process
continues, the boiling action will loosen particles from the pores and surfaces of the metal. This process produces
HHO gas, so it is very important that the gas is not allowed to collect anywhere indoors (such as on ceilings).

After several hours, disconnect the electrical supply and pour the electrolyte solution into a container. Rinse out
the cells thoroughly with distilled water. Filter the dilute NaOH solution through paper towels or coffee filters to
remove the particles. Pour the dilute solution back into the cells and repeat this cleaning process. You may have
to repeat the electrolysis and rinsing process many times before the plates stop putting out particles into the
solution. If you wish, you can use a new NaOH solution each time you cleanse, but please understand that you
can go through a lot of solution just in this cleaning stage if you choose to do it that way. When cleansing is
finished (typically 3 days of cleansing), do a final rinse with clean distilled water. It is very important that during
cleansing, during conditioning and during use, that the polarity of the electrical power is always the same. In other
words, don’t swap the battery connections over as that destroys all the preparation work and requires the
cleansing and conditioning processes to be carried out all over again.

Using the same concentration of solution as in cleansing, fill the cells with dilute solution. Apply about 2 volts per
cell and allow the unit to run. Remember that very good ventilation is essential during this process. As water is
consumed, the levels will drop. Once the cells stabilise, monitor the current draw. If the current draw is fairly
stable, continue with this conditioning phase continuously for two to three days, adding just enough distilled water
to replace what is consumed. If the solution changes colour or develops a layer of crud on the surface of the
electrolyte, then the electrodes need more cleansing stages. After two to three days of run time, pour out the
dilute KOH solution and rinse out the cells thoroughly with distilled water.

The construction which Zach has used is very sensible, utilising readily available, low-cost PVC piping. The spiral
electrodes are inside 2” diameter pipe and Zach says that the bubbler is also 2” diameter PVC pipe. I seriously
doubt that a two-inch diameter bubbler could handle a flow as high as 17 lpm which is a substantial amount. Also.
You want the bubbles in the bubbler to be small in order that the gas comes into good contact with the water.
Consequently, using more than one bubbler where the diagram shows just one, would be sensible.

At this time, Zach only uses one bubbler, but a second one is highly desirable, located between the storage tank
and the engine and positioned as close to the engine as possible. This extra bubbler does two things, most
importantly, it prevents the gas in the storage tank being ignited by a backfire caused by a valve sticking slightly
open and secondly, it removes every last trace of potassium hydroxide fumes from the gas, protecting the life of
the engine. This is a big gain for such a simple addition.

The gas storage tank is also made from PVC pipe, this time, 4 inch (100 mm) diameter, 14 inches (350 mm) long
with standard end caps fixed in place with PVC glue as shown below. This is a compact and effective
arrangement well suited for use on a motorcycle. The majority of this extra equipment can be mounted in bike
panniers, which is a neat arrangement.

10 - 44
The electric drive to the electrolyser is from a Pulse Width Modulator (“DC Motor speed controller”) which was
bought from the Hydrogen Garage as Zach is in America. That particular PWM board is no longer available, so
especially for those people in Europe the choice might be rmcybernetics.com, although there are many suppliers
and the module should not be expensive.

As this unit was rated at just 15 Amps maximum, Zach added another 15 Amp rated FET transistor in parallel to
the output stage to raise the current capacity to 30 Amps. A fuse protects against accidental short circuits and a
relay is used to control when the electrolyser is to be producing gas. The connecting wire is #12 AWG (swg 14)
which has a maximum continuous current capacity of just under ten amps, so although the current peaks may be
twenty amps, the average current is much lower than that.

Two electromagnets outside the bubbler, positioned 2.5 inches (65 mm) above the base, are connected as part of
the electrical supply to the electrolyser, and these cause most of the oxygen and hydrogen bubbles to separate
and exit the bubbler through different pipes. There is a divider across the bubbler to assist in keeping the gases
from mixing again above the water surface. The bubbler also washes most of the potassium hydroxide fumes out
of the gas as the bubbles rise to the surface, protecting the engine as these fumes have a very destructive effect
on engines.

The objective with any HHO system is to have the minimum amount of gas between the bubbler and the engine in
order to block the ignition of the gas in the unlikely event of a backfire. In this system, the gas storage tank
contains a very large amount of gas, though admittedly it is not full HHO gas thanks to the electromagnet
separation system, but nevertheless, it would be most advisable to have a second bubbler between the gas
storage tank and the engine, positioned as close to the engine as possible. HHO gas produces a very high-speed

10 - 45
shock-wave when it is ignited so the bubbler needs to be of strong construction to withstand this. No pop-off
bubbler cap or blow-out device acts fast enough to contain a HHO shock-wave, so make the bubbler housing
strong enough to withstand the pressure wave.

Zach’s electrolyser arrangement is like this:

It must be realised that the water tank, electrolyser, bubbler/separator and hydrogen holding tank, all operate at
thirty pounds per square inch. This means that each of these containers must be robust enough to withstand that
pressure quite easily. It also means that the 30 psi one-way check valve on the oxygen venting pipe is an
essential part of the design as well as being a safety feature. As a bubble of gas from the electrolyser escapes
into the water tank every time a drop of water feeds to the electrolyser, the contents of the water tank above the
water surface becomes a stronger and stronger mix of air and HHO. Consequently, it soon becomes an energetic
mixture. It is common for static electricity to build up on a tank of this nature, so it will be very important to earth
both the tank and it’s cap before removing the cap to top up the tank with more water.

The electrolyser has a potassium hydroxide (KOH) solution in it. The electrolysis process produces a mixture of
hydrogen, oxygen, dissolved gases (air) and potassium hydroxide fumes. When the system is being used, the
water in the bubbler washes out most of the potassium hydroxide fumes, and in doing so, it gradually becomes a
dilute electrolyte itself. Potassium hydroxide is a true catalyst and while it promotes the electrolysis process, it
does not get used up during electrolysis. The only loss is to the bubbler. Standard practice is to pour the
contents of the bubbler into the electrolyser from time to time, filling the bubbler again with fresh water.
Potassium hydroxide has been found to be the most effective catalyst for electrolysis but it has a very bad effect
on the engine if it is allowed to enter it. The first bubbler is very effective in removing the potassium hydroxide
fumes, but many people prefer to take the scrubbing process a step further by placing a second bubbler in the
line, in this instance, between the hydrogen pressure tank and the engine. With two bubblers, absolutely no
potassium hydroxide fumes reach the engine.

When running with HHO gas as the only fuel, it is essential to adjust the timing of the spark so that it occurs after
Top Dead Centre. The timing on this bike is now set at 8 degrees after TDC. However, if David Quirey’s style of
bubbling the HHO through a liquid such as acetone, then no timing alterations would be needed.

This electrolyser is designed to run off the nominal six volts of a motorcycle electrics (about 7.3 volts with the
engine running), but increasing the number of tubes, each containing electrode coils, would convert the design to
a 12V system and then the electrolyser housing would probably be like this:

10 - 46
It is possible that seven sets of three or four spirals wired in parallel would be used for larger engines with their
13.8 volt electric systems. Zach uses the very simple method of allowing excess gas to be vented via the oxygen
valve if gas production exceeds the requirements of the engine. When operating on a twelve volt system it might
be more convenient to use a standard pressure switch which opens an electrical connection when the gas
pressure rises above the value for that switch:

The pressure switch just mounts on one of the end caps of the pressure tank and the switch electrical connection
is placed between the relay and the electrolyser. If the gas pressure reaches it’s maximum value of 30 psi. then
the switch opens, stopping electrolysis until the pressure drops again:

10 - 47
Caution: This electrolyser is not a toy. If you make and use one of these, you do so entirely at your own
risk. Neither the designer of the electrolyser, the author of this document or the provider of the internet
display are in any way liable should you suffer any loss or damage through your own actions. While it is
believed to be entirely safe to make and use an electrolyser of this design, provided that the safety
instructions shown below are followed, it is stressed that the responsibility is yours and yours alone.

An electrolyser should not be considered as an isolated device. You need to remember that both electrical and
gas safety devices are an essential part of any such installation. The electrical safety devices are a circuit-
breaker (as used by any electrician when wiring a house) to protect against accidental short-circuits, and a relay
to make sure that the booster does not operate when the engine is not running:

However, the system designed by Zach West is almost certainly not self-sustaining and if that is correct, then the
battery powering the electrolyser will need to be charged between trips. That does not have to be the situation as
high-efficiency electrolysers are available. First, the Shigeta Hasebe spiral plate electrolyser (shown on page 719
of the Appendix) has produced 7 lpm of HHO gas mix for an input of just 84 watts and while that 84 watts is an
inconvenient 2.8V at 30 amps, it should be possible to raise the voltage and lower the current without losing too
much of the performance. In my opinion, the electrics of a motorcycle should be able to output 84 watts and so
the motorcycle could become self-powered.

Motorcycles can definitely become self-powered as can be seen from the electric motorcycle system of Teruo
Kawai’s COP>3 design shown in chapter 2. Teruo went to America and was in a meeting aimed at getting his
design manufactured and sold in America when the meeting was interrupted and Teruo intimidated into
abandoning his venture.

You must also remember that Steve Ryan of New Zealand demonstrated running his motorcycle on treated water.
I suspect that the treated water was water which had been infused by charged water clusters as described by
Suratt and Gourley later in this chapter. Their electrolyser has an efficiency of 0.00028 kilowatt-hour or less to
generate one litre of gas. Those inconvenient units mean that to produce 1 lpm needs 16.8 watts or 7 lpm needs
118 watts. If cold water mist is added to the air entering the motorcycle engine, then it seems likely that a good
deal less than 7 lpm would be needed. If you have a good enough tank which is made of a material capable of
containing the very small molecules of this gas, then the gas can be compressed to 1000 psi and that should
allow a motorcycle to run for some time on the gas cylinder.

The “DuPlex” Booster designed by Bill Williams.

A fully-submerged design from Bill Williams in the USA is another different style of booster:

10 - 48
The construction details for this booster design, can be downloaded free from the web using the link:
http://www.free-energy-devices.com/DuPlex.pdf

There are many other designs, including those with concentric pipes, each having its own advantages and
disadvantages, some being commercially available as ready-made devices, and there are links to these boosters
on the web sites mentioned above and a general booster forum at http://tech.groups.yahoo.com/group/watercar/
and another at http://tech.groups.yahoo.com/group/HHO/ where people will answer any queries.

One problem with the use of boosters is that if the HHO gas volume is higher than it needs to be, the vehicle's
Electronic Control Unit ("ECU") is liable to detect the improved fuel burn and start pumping in excess fuel to offset
the improved conditions. How to deal with this situation is covered in the free document which can be
downloaded from http://www.free-energy-devices.com/D17.pdf

The “Hogg” Electrolyser from Selwyn Harris.

An interesting design is the Hogg electrolyser as described by Selwyn Harris of Australia. The Hogg cell has two
stainless steel mesh electrodes coiled around each other. This gives a large electrode surface area in a very
compact container. In this version of the design there are six identical cells which feed into a large bubbler. For
clarity, only two of the six electrolyser cells are shown here:

10 - 49
The items marked with a blue dot form just one of three identical sets. That is, three filters feed out from the
bottom of the bubbler, the flow through them being caused by three separate pumps and the two electrolyser cells
marked “1” and “2” along with their associated pipes and four one-way valves, are replicated to give electrolysis
cells “3” to “6” which are not shown in this diagram. Those three identical sets are connected to the central
bubbler, spaced around it evenly at 120-degree positions horizontally as shown in rough outline here:

The water is circulated through the set of cells using three small pumps and there are two water collectors built on
to the bottom of the bubbler. Also, as the ‘electrolyte’ used is rainwater, and each electrolysis cell is completely
full of electrolyte as this is an ‘electrolyte circulation’ style of electrolyser.

Each of the three pumps has it’s own filter to trap any particles coming from the cells as experience has shown
that the water can contain a considerable amount of material. The filters are standard irrigation in-line filters made
from transparent plastic filled for three quarters of their length with fine plastic sponge material.

A key feature of the cell design is the use of two powerful neodymium magnets per cell. These act directly on the
water and that causes a major increase in the gas production rate. The magnets have their North poles facing
towards each other.

The two mesh electrodes are made from stainless steel wire of 0.32 mm diameter and woven to give 2 mm holes
between the wires and a 0.65 mm overall sheet thickness. These dimensions are important as other mesh sizes
and styles do not give so good a performance. The electrodes are wider at one end to form a connection tab
which allows easy electrical connection to each electrode and they are then wired in parallel so that each cell gets
12 volts across it as shown here:

The two mesh electrodes are cut like this:

10 - 50
and:

The mesh material looks like this:

10 - 51
The six electrolyser cells and the single large bubbler are constructed using standard plastic plumbing materials:

10 - 52
The connecting pipes are a clear plastic, heat-tolerant type of 12 mm (0.5”) diameter. The bubbler is also
constructed from plastic pipe fittings:

As some people have difficulty in visualising the way in which the electrodes are combined, this simplified concept
sketch may be helpful:

The two electrodes are kept separated by the use of small diameter fibre washers held in place between them at
strategic places using super glue. The mesh itself is then treated by being immersed in citric acid in order to
make it work well with rainwater.

There are three of these pairs of electrolyser cells, each pair being connected to the upper collector reservoir.
The water being pumped out of each cell is passed through one of the three filters before entering the collector
reservoir which feeds the small pump which keeps the water circulating which in turn keeps removing any
particles which have got into the rainwater.

10 - 53
When used with rainwater from a barrel, this electrolyser is said to draw just 1.4 amps per cell, giving a total input
of about 115 watts when run on a 12-volt electrical supply. While rainwater is supposedly pure, the reality is that it
seldom is and it’s ability to carry a current varies dramatically from place to place and even more widely from
country to country. If you decide to build this electrolyser and find that you do not get anything like 1.4 amps
flowing through any one cell, then you may well have to add a small amount of electrolyte to the water in order to
get the current flowing. That is, assuming that you want a current flow of 1.4 amps per cell on 12 volts. The
output of this electrolyser is said to be capable of running a small electrical generator but that has not been
confirmed at this time.

AVA Magnetic Levitation’s Electrolyser

Adam of AVA Magnetics shows a very nicely constructed electrolyser based on an entirely different style of
electrode. His video is at https://www.youtube.com/watch?v=Iz8wuUXWuGU and his cell looks like this:

Initially built using stainless steel pipe, the electrical resistance of the long coil proved to be a problem, so Adam
overcame the problem by inserting copper pipe inside the stainless steel pipe. Copper carries electrical current
very well, while stainless steel has quite a high resistance to current flow, which many people find very strange,
having been taught in school that all metals conduct electricity. In a coil like this, electric current has to pass
through a considerable length of stainless steel in order to reach the lower turns, and that resistance to current
flow is greater than the situation with an electrolyser which uses stainless steel sheets. The copper tube down the
inside of the steel pipe makes a major difference because the current flows easily through the copper and then at
every point along the tube, the current only needs to flow sideways through a thin layer of stainless steel in order
to perform the electrolysis.

The photograph above looks like a very neatly wound coil with twelve turns in it. That is not the case. It is two
separate coils with turns 1, 3, 5, etc. being in one coil and turns 2, 4, 6, etc. being in the other coil. That
arrangement means that each turn in the coil has a turn with the opposite voltage beside it and the electrolysis
takes place between adjacent turns in the coil. The ends of each coil are sealed off carefully to prevent corrosion
of the copper pipe by the electrolyte being used. Unfortunately, most of the surface area of the circular pipe
electrode is not close to its neighbouring coil surface and that increased distance reduces the rate of gas
production for those areas. I suspect that Adam is using a fairly dilute electrolyte in order to keep the heating
effect under control and so, better performance may result when using full-strength electrolyte. However, the

10 - 54
HHO production at around 16 amps looks impressive as shown here just before the whole container is obscured
by the HHO bubbles coming off the coils:

At full power, the cell draws 18 amps when powered through a Pulse-Width Modulator manufactured by the
”HHO Powercell” company using a battery of around 12.5 volts. That 225 watts produces a gas output rate which
needs to be measured. In my opinion, the gas output pipe needs to be much larger with an inner diameter of
perhaps 15 millimetres as a small pipe hinders the escaping gas. If you don’t believe that, then try blowing
through a length of 6 millimetre diameter plastic pipe and see how hard that is to do.

Nevertheless, this is a simple, basic electrolyser with just two electrodes, although, possibly because the Pulse-
Width Modulator is feeding it a good quality square wave signal, the small bubble size suggests that there is an
efficiency gain which makes it perform better than would be expected. A single cell like this will find it hard to
escape from the basic fact that 1.24 volts is all that is used in water electrolysis and even with the voltage drop
across stainless steel you really do not want more than 2-volts between the plates. In other words, only 2-volts of
the battery’s 12-volts can actually make gas and so, the vast majority of that 225 watts goes to make heat and not
gas. The situation for 1, 2 or 3 single cells is like this:

It is probable that HHO production at 225 watts would be doubled if two of these cells are connected in series,
that is, connected in a chain. The performance would probably be tripled at that same 18 amps of current, if three
of these cells are connected in series across the voltage source. If that is done, then ideally, each cell needs to
be inside its own separate container. Adam suggests that the coil produces a magnetic field which boosts the
production of HHO, and considering the effect of the magnets in the Shigeta Hasebe cell, he may well be right.

10 - 55
The arrangement with three cells connected in series and powered through a frequency-adjustable Pulse-Width
Modulator would be like this, although the gas outlet pipes need to be connected together with short pipe
connections as those pipes will be filled with energetic HHO when in use:

As always, the battery connects to a circuit breaker “CB” or fuse before anything else is connected. The Pulse-
Width Modulator “PWM” is effectively like Dave Lawton’s most simple circuit with adjustable frequency, adjustable
Mark-Space Ratio (or “Duty Cycle” as it is sometimes called), and a high output current capability. The electrical
connections to the cells are done as shown in the diagram and the cells are all positioned the same way round as
that allows testing with the cells leads swapped over to see if that makes a difference when the best frequency
and Duty Cycle have been found. Please remember that a bubbler is essential on the gas output pipe and the
depth of water in any bubbler should be at least five inches which is 125 millimetres.

No measurement of the rate of HHO flow has been made. This can be measured with a cheap gas flow meter
from a medical equipment supplier or a result which is easily good enough for our purposes can be determined by
using a two litre or other suitable container of known capacity, filling it with water and bubbling the HHO output
into the container as shown here:

The result using this method is got by timing how long it takes for the bottle to be filled with the HHO gas mixture,
and from that, working out how many litres are being produced per minute. For example, if it takes one minute to
fill a 2-litre bottle, then the flow rate is 2-litres per minute. If it takes two minutes to fill, then the gas output rate is
1-litre per minute. If it takes 2-minutes and 30-seconds to fill, then the rate is 0.8 litres per minute. The result will
be altered slightly by the atmospheric pressure which changes from hour to hour, but we only need a rough
estimate, especially if we are just comparing performance after some small change to the method of electrolysis.

The https://www.youtube.com/watch?v=YfG6kyifq30 video shows methods of winding pipes into helical coils as
used here and the plastic spacers needed, but the man posting, ‘ANNMANN’, has a fully equipped workshop and
constructional skills not available to most experimenters. Unfortunately, he does not seem to aware of the fact
that electrodes need to be very close together for the best HHO production. He has tried using marine cable
(used for mast stays) instead of pipe and that is, of course, very much easier to use when making spiral
10 - 56
electrodes, but it is probable that four spacers would be needed rather than just the two used with pipe spirals.
There is also, no scope for using copper inside the hawser. Selwyn Harris of Australia, who is a very experienced
HHO user, states that it is very difficult to wind stainless steel pipe into an accurate spiral. He wonders if winding
the spirals using copper pipe which is very much easier to bend than stainless steel, and then electroplating the
spiral with chromium, might not be a better and more practical way of producing good working spiral electrodes.

For me, with my serious lack of accurate constructional skills, another much easier method of construction occurs
to me. First, the arrangement of two spirals is not really effective for electrolysis as shown here:

The rate of electrolysis is directly proportional to the area of the electrodes. It is also highly affected by the
separation of the electrodes. With the wires or pipes shown on the left, only a small percentage of the surface
area is spaced at the best distance which your construction allows (marked by the red arrows). The two plate
electrodes shown on the right have their entire surface area at the optimum spacing. So, if we are going to use
pipes or wires, then we want to increase the most active surface area, and in fact, increase the electrode area
overall. Working with steel is much more difficult than working with plastic, especially since most of us do not
have the tools, workspace and skills to help us succeed, using marine-grade stainless steel cable is a very
appealing option. This material is shown at https://www.youtube.com/watch?v=11Qn4CGlZp4 and looks like this:

10 - 57
It is available from ship’s chandleries world wide and is constructed from stainless steel wires. If we use this, then
we have to accept the fact that it’s electrical resistance will be high, but the big advantage is that we only have to
make the plastic spacers.

This brings us to the problem of increasing the surface area and decreasing the gaps between the wires. While
ANNMANN has excellent video presentations, please do not be misled by the excessively large gaps which he
had between his wires. The largest gap that we want is 3 millimetres, and ideally, we would like 2 millimetre gaps.
So, for this I suggest that a plastic plate is drilled like this:

For this, the hole size matches your cable diameter, say, 6 mm, and the gaps between the holes are 2 millimetres
if you are confident, and 2.5 or 3 millimetres if you are not confident of your skill in drilling plastic. This plate will
be used to wind four nested spirals. As the cable is flexible, we need to use two additional spacer strips like this:

10 - 58
These are to make sure that the desired gap is maintained along all of the cable length, by threading the cable
through these as each turn is made. That way, the cable is supported every quarter turn, maintaining the gap size
all the way around every turn.

Although only one type of cable is used, colours are used in the following diagrams merely to make it easier to
see where the turns in each winding start and finish. All windings start at the bottom and end at the top. The first
winding uses the innermost holes:

This winding is made with two separate lengths of cable, one shown in green and one shown in red. The lower
end of the red cable terminates just this side of the plastic sheet. The lower end of the green cable terminates just
beyond the other side of the plastic sheet. Unlike the diagram, the cable is NOT cut off short as shown but is left
long so that it can be connected outside the plastic cell container.

We now double the electrode surface area by winding a second pair of coils just outside the first one. The
arrangement looks like this:
10 - 59
This second layer is wound in exactly the same way as the previous one. The outer cable turns are not exactly
above the spiral turns underneath, but they are very close to being directly above. When we connect the cables
electrically, we arrange for the cables surrounding any cable to be the opposite polarity, so that the electrolysis is
maximised.

The next step is to wind the third layer:

And then the fourth layer is wound using the remaining holes:

10 - 60
Connected electrically as shown, every cable in the first and fourth layers is surrounded on three side with cables
of opposing polarity. The cables in layers two and three have every turn surrounded on all four sides by cables of
the opposing polarity. So, this arrangement has overcome the need for difficult pipe bending, has increased the
electrode area considerably and has achieved a reasonable gap between the electrodes and has improved the
area of best electrolysis by 87%. We now improve matters considerably by using three of these cells in series in
order to use the applied power more efficiently.

There is, of course, no reason why there should not be more than three of these cells in series and I would expect
five to be the likely maximum due to excessive voltage drop across the stainless steel cable. There is also no
reason why there should not be five or more spiral layers in each cell and each additional layer raises the overall
efficiency of that set of spirals.

However, while it seems very probable that this construction would yield a good performance, please remember
that this is only a suggestion and at time of writing, it has not been built and tested. When first testing the unit, the
Pulse-Width Modulator is set to give a low to medium current flow through the cells. Then the frequency control is
used to get the maximum gas output without altering the current setting. If the current drops at the point of peak
gas production during this test, that is not a problem. With multi-layer spirals of this type, a gas outlet pipe of 20
millimetres would be recommended. Also, the gas outlet should have a baffle plate across it to prevent tiny
droplets of electrolyte escaping with the gas, as shown elsewhere in this chapter. That should be done even if the
electrolyser is to be used in a stationary location such as feeding a generator. Electrical connections to the ends
of the spiral wires can be made with large screw connectors not unlike the ordinary hardware store type but
intended for much larger wires. These are also available through ship’s chandlers as boats often use high-current
twelve volt wiring of large diameter.

Advanced DC Boosters.
All of the practical construction details on electrical safety, gas safety, engine connections, type of water, safe
mixing of electrolyte, etc. already discussed, apply to all kinds of electrolysers and boosters of every design. So,
please understand that these are universal features which need to be understood when using any design of
booster.

It is possible to produce large volumes of HHO gas from a DC booster, enough gas to run a small motor directly
on it. For this, we need to pay attention to the efficiency factors already covered in this document. The person

10 - 61
who is outstanding in this field is Bob Boyce of the USA who has kindly shared his experience and expertise freely
with people who want to use serious electrolysers.

Bob's attention to detail when constructing high-performance electrolysers has resulted in efficiencies which are
more than double those of the very famous Michael Faraday whom most scientists consider to be the final word
on electrolysis.

Bob Boyce’s High-Efficiency Electrolyser.

We are now moving from the "casual" style of booster to the "serious" style of electrolyser. In this category, you
will find that the units built are not cheap, weight a considerable amount, require considerable skill to make and
usually are quite large physically. I will mention two designs here. First, the very well-known design from Bob
Boyce. For this electrolyser, Bob makes solid stainless steel electrode plates act as cell partitions as well as
being electrodes. This is a clever technique but it takes a very high level of construction accuracy to make a box
with slots in the side and base, so that the stainless steel plates can be slid into the box and when there, form a
watertight seal between the cells, preventing electrical current bypassing the places by flowing around them.

The number of cells in the electrolyser depends on the electrical DC voltage supply which is produced from the
electrics of the vehicle. This higher voltage is created by using a standard off-the-shelf "inverter" which produces
high-voltage alternating current ("AC") meant to be the equivalent of the local electricity mains supply. In the
USA, the voltage produced is in the 110 to 120 volt region, elsewhere, it is in the 220 to 230 volt region.

If you are not familiar with electrical jargon, then check out chapter 12 which explains it step by step. The AC
output from whatever inverter you buy, is changed back into DC by using a component called a "diode bridge" and
a reservoir device called a capacitor. When this is done, the resulting DC voltage is 41% greater than the quoted
AC voltage, so a 110-volt inverter will produce about 155 volts and a 220-volt inverter about 310 volts. As you
want about 2 volts per cell, the number of cells would be about 80 or 150 depending on which inverter is used.
This large number of stainless steel plates each sized at six-inches (150 mm) square, creates a substantial weight
which then is increased by the weight of the case, and the electrolyte. The overall arrangement (without the
capacitor) is like this:

A very high-precision box for this style of electrolyser can be made using the design of the late Ed Holdgate of
Florida if you are a skilled fabricator:

10 - 62
The gas production rate is so high that the gas outlet pipe has to have holes drilled along the top in order to try to
exclude spray and moisture from the massive rate of bubbles bursting at the surface of the electrolyte. The high
efficiency of Bob's electrolysers is due to his meticulous preparation and construction methods. You will notice
that Bob recommends the use of a particle filter with a 1-micron mesh, between the engine and the HHO system.
Apart from ensuring that everything entering the engine is very clean, the particle filter with a mesh of that small
size, also acts as a flashback-preventer as flame can't pass through it.

Firstly, the stainless steel plates are cross-scored with sandpaper to create a specially shaped plate surface
which helps high-speed bubble release. Secondly, the plates are put through a rigorous "cleansing" process
where they are subjected to repeated periods of electrolysis followed by rinsing particles off the plates and filtering
the electrolyte solution. When no further particles break free from the plates, they are then put through a
"conditioning" process which develops a catalytic layer on the plate surfaces.

This processing and the various construction details are provided in the following free download document,
thanks to Bob's generosity in sharing his experience with us: http://www.free-energy-info.com/D9.pdf and there is
a forum for Bob's design: http://tech.groups.yahoo.com/group/WorkingWatercar/ where questions are answered.
Here are the main details:

Bob Boyce is a most experienced and knowledgeable series-cell designer, and sincere thanks are due to him for
sharing his design freely with everybody and for his continuous help, advice and support of the builders of
electrolysers. Bob achieves a massively increased gas production rate by using an electrolyser with a large
number of cells in it. Bob’s electrolyser is easily the most effective available at this time. It uses one hundred
cells (101 plates) and applies a sophisticated pulsing waveform which raises the operational efficiency far above
that envisioned by the science textbooks available today. Units with just 60 cells are inclined more to brute-force
DC electrolysis, tending to mask the gains produced by pulsing. As there is a voltage drop across each stainless
steel electrode plate, it is usual to allow about 2 volts across each cell for DC operation. However, Bob finds that
for high-efficiency pulsing, the optimum voltage for each cell with 316L-grade stainless-steel electrode plates is
about 1.5 volts. This means that a voltage of about 1.5 x 100 = 150 volts is needed to power it to its maximum
pulsed output.

To get this higher voltage, Bob uses a 110 Volt inverter. An inverter is a common, commercially available
electronic circuit which usually has a 12 Volt DC input and generates a 110 Volt AC output. These are readily
available for purchase as they are used to run (US) mains equipment from car batteries. The output from the
inverter is converted from Alternating Current to pulsing Direct Current by passing the output through four diodes
in what is called a ‘Diode Bridge’. These are readily available at very low cost from electronic component
suppliers.

Obviously, it would not be practical to use a hundred self-contained cells daisy-chained together to act as the
series-connected electrolyser cell. There would not be enough physical space in the engine compartment for that,
so a different style of cell construction is needed. The view looking down on several separate electrolyser cells
could be represented something like this:

Here the plus side of each cell is connected to the minus side of the next cell to provide a set of six interconnected
cells acting in series. The current flowing through the electrolyser goes through each cell in turn and so each cell
receives exactly the same current as the other cells. This is the same sort of arrangement as using six self-

10 - 63
contained cells in a daisy-chain. To reduce the physical size of the unit, it is possible to construct the electrolyser
as shown here:

In this arrangement, the individual cells have just one positive plate and one negative plate. The plates slot into
the sides and bottom of the housing so that the electrolyte is trapped between the plates and an air gap is formed
between the plus plate of one cell and the minus plate of the next cell.

These air gaps are wasted space. They contribute nothing to the operation of the electrolyser. Each consists of a
metal plate, a gap and a wire connection to the next metal plate. From an electrical point of view, the two metal
plates at the opposite ends of these gaps, being connected by a wire link, are effectively the same plate (it is just
a very thick, hollow plate). These air gaps might as well be eliminated which would save one metal plate and one
wire link per cell. This can be difficult to visualise, but it produces an arrangement as shown here:

The only air gaps remaining are at the ends of the electrolyser. The plates in the middle are notionally touching
each other. The positive plates are marked in red and the negative plates are shown in blue. In reality, there is
only one metal plate between each cell and the next cell - the red and blue marking is only a notional device to try
to make it easier to see that the diagram actually shows six separate cells in a single housing. They are separate
cells because the metal electrode plates extend into the base and sides of the housing, thus isolating the six
bodies of electrolyte from each other. It is very important that the different bodies of electrolyte are fully isolated
from each other, otherwise the electrolyser will not act as a series-connected unit and most of the current will skip
past the middle plates and just run from the first plate to the last plate around the sides of the other plates. So,
the plates need to be a fairly tight push-fit in grooves cut in the sides and base of the housing. The electrolyte
level must always be below the top of the plates as shown here:

An electrolyser with a hundred cells, built in this style will have 101 metal plates and 100 separate bodies of
electrolyte. In spite of these large numbers, the size of the overall unit does not have to be excessive. The
spacing between the plates is set to, say, 3 mm (1/8 inch) and the plate thickness might be 16 gauge (1/16 inch),
so the width of a 100-cell electrolyser is about 20 inches. In actual practice, the gaps at the end of the
electrolyser may also contain electrolyte although that electrolyte takes no part in the electrolysis process.

The size of the plates may be determined by the space available in the engine compartment. If there is a large
amount of spare space, then the plate size may be selected by allowing from two to four square inches of area on
both sides of each plate, per amp of current. Each side of every plate is in a different electrolysis cell so a 6-inch

10 - 64
by 6-inch plate will have 36 square inches on each face and so would carry between 36 / 4 = 9 to 18 amps of
current. The choice of current is made by the builder of the electrolyser and it will be influenced by the size and
cost of the inverter chosen to drive the electrolyser and the allowable current draw from the battery. This is for
straight DC electrolysis where the battery is connected directly across the electrolyser. Using Bob’s triple-
oscillator electronics pulser card, the electrolyte level has to be kept down to about three inches from the top of
the six inch plate because the gas production rate is so high that there has to be substantial freeboard to stop the
electrolyte being splashed all over the place.

Bob usually uses a 6” x 6” plate size. It is essential that every item which contains HHO gas is located outside the
passenger compartment of any vehicle. Under no circumstances should the electrolyser or bubbler be located in
the passenger area of the vehicle as the ignition sound generated is so great that permanent hearing damage
would be a serious possibility.

For straight DC operation of an electrolyser of this type, the circuitry is very straightforward. The inverter should
be mounted securely, preferably in the stream of air drawn in to cool the radiator. Using a diode “bridge” of four
diodes converts the stepped up AC output of the inverter back into pulsing DC and produces the electrical
arrangement shown here:

As mains voltage is quoted as an average figure (“root-mean-square”) it has a peak voltage of 41% more than
that. This means that the pulsing DC has a voltage peak of just over 150 volts for the nominal 110 volt AC output
from the inverter. The bubblers and the particle filter remove all traces of electrolyte fumes from the gas as well
as protecting against any accidental igniting of the gas caused by the engine misfiring.

The very famous Michael Faraday who was an exceptionally gifted experimenter, placed two electrodes in water
and determined how much gas was produced per amp of current. Using an electrolyte and recent tenchology
when running on DC, Bob Boyce would not consider an electrolyser properly constructed, cleansed and
conditioned until it was producing more than double Faraday’s gas production rate. A typical working electrolyser
made by Bob would have about 216% of Faraday’s result. People taught in universities and unaware of current
technology, use Faraday’s result in calculations and those calculations indicate that it would take more energy to
produce HHO gas than could be produced by then burning the hydrogen produced. Their calculations are wrong.
The energy in freshly made HHO gas is typically four times more energetic than hydrogen is and so those
calculations are too low by a factor of more than eight times. Also, the majority of energy from burning HHO
comes from “charged water clusters” (see chapter 10) and not from the hydrogen, and most of these good people
doing the calculations have never even heard of charged water clusters, and so, they accept the “can’t be done”
verdict without thinking about it.

Pulsed Operation
If you have already read chapter 10, you will know that the next step forward in raising HHO production is to apply
a suitable pulsed waveform to the electrolyser terminals rather than just a straight DC voltage. Doing this with the
design of Bob Boyce raises the cell efficiency to around ten times the result produced by Michael Faraday. Bob
Boyce’s highly efficient pulsed electrolysis system has been very generously shared freely by Bob so that anyone
who wishes may construct one for their own use without the payment of a licence fee or royalties. Just before
presenting the details, it should be stressed that in order to get Bob’s performance of up to 1,000% of the Faraday
(supposed) maximum gas output, each step needs to be carried out carefully exactly as described. Much of the
following text is quoted from Bob’s forum posts and so should be considered as his copyright, not to be
reproduced without his permission.

10 - 65
Your Responsibility:
If you decide to construct an electrolyser of this, or any other design, you do so wholly on your own responsibility,
and nobody is in any way liable for any loss or damage, whether direct or indirect, resulting from your actions. In
other words, you are wholly responsible for what you choose to do. I say again, this document must not be
construed as an encouragement for you to construct this or any other electrolyser.

Bob’s electrolyser splits water into a mixture of gases, mainly hydrogen and oxygen. That gas mixture, which will
be referred to as “HHO” is highly reactive and must be treated with respect and caution. A fairly small volume of
HHO gas ignited in air is quite liable to cause permanent hearing loss or impairment due to the shock waves
caused by the ignition. If HHO gas is ignited inside a sealed container, then the resulting implosion is liable to
shatter the container. Bob uses two bubblers and a one-way valve to protect against this occurrence, and details
of these are given in this document.

To make the water inside the electrolyser carry the necessary current, potassium hydroxide (KOH) is added to
distilled water. This is the best electrolyte for an electrolyser of this type. Potassium hydroxide is also known as
“caustic potash” and it is highly caustic. Consequently, it needs to be handled carefully and kept away from
contact with skin, and even more importantly, eyes. If any splashes come in contact with you, it is very important
indeed that the affected area be immediately rinsed off with large amounts of running water and if necessary, the
use of vinegar which is acidic.

This electrolyser design uses a toroidal transformer to interface the electronics to the electrolyser cells. It is vital
that this transformer be used very carefully. Under no circumstances may this transformer be powered up by the
electronics when connected to anything other than the filled electrolyser cells as they act as a safety buffer.
When driven by Bob’s electronics, this transformer draws additional energy from the environment. While this is
very useful for electrolysis, there are sometimes unpredictable energy surges which can generate as much as
10,000 amps of current. If one of these should occur when the transformer is not connected to the electrolyser
which is able to soak up this excess, the resulting electrical conditions can be very serious. If you are lucky, it will
just burn out expensive components. If you are not lucky, it can cause a lightning strike which is liable to hit you.
For that reason, it is absolutely essential that the toroid transformer is never powered up with the secondary
winding connected to anything other than the filled electrolyser.

Patenting:
It should be clearly understood that Bob Boyce, has released this information into the public domain and it has
been displayed publicly since early in 2006. It is not possible for any part of this information to be made part of
any patent application anywhere in the world. This prior public disclosure of the information prevents it being
patented. It is Bob’s intention that this information be freely available to people world-wide. It should also be
emphasised that all of the quotations of Bob’s words which is a very extensive part of this document, remain the
copyright of Bob and may not be reproduced for display or sale without his prior written permission.

The Objective:
This is a “HHO-On-Demand” (“HOD”) system. It is very difficult indeed to generate HHO gas fast enough to
power an internal combustion engined vehicle under all road conditions. Moving from standstill to rapid
acceleration causes such a massive sudden requirement for additional volumes of HHO gas, that it is difficult to
provide that volume instantly.

A better solution is to use an electric engine for the vehicle. This can be an electric vehicle which was designed
from scratch as such, or it can be a standard vehicle which has been adapted for electric engine use. These
electric vehicles are usually limited in how far they can travel, but a good solution to this is to use an electrical
generator to charge the batteries, both when the vehicle is in use and when it is parked. This electrolyser can be
used to run such a generator on water as shown in chapter 10. With this arrangement, there are no CO2
emissions and the vehicle is very environmentally friendly. The batteries provide the necessary sudden
acceleration demands and the generator recharges the batteries during normal driving.

Overview:
Bob’s pulsed system has the following components:

1. An electrical connection to the vehicle’s electrical system (with safety features built in).
2. An “inverter” which raises the electrolyser voltage to about 160 volts.
3. Bob’s specially designed circuit board which generates a complicated water-splitting waveform.
4. Bob’s specially designed toroidal transformer which links Bob’s circuit board to the electrolyser.
5. Bob’s specially prepared 101-plate series-connected electrolyser.
10 - 66
6. A dual-protection system for linking the electrolyser safely to the internal combustion engine.

None of these items is particularly difficult to achieve, but each needs to be done carefully and exactly as
described, paying particular attention to the detailed instructions.

Building the Case:

The case needs to have very accurate slots cut in it. If you do not have a milling machine, then you might
consider getting a fabrication shop to mill the slots for you. The case has two ends, two sides, one base and one
lid. Of these, the two sides and the base need 101 accurate grooves cut in them. The grooves are there to hold
the electrode plates securely in position, and have to be cut extremely accurately. The groove width is set at
0.0003" less than the actual, measured plate thickness. This prevents any electrical flow around the plates.

Many people ask about moulding the slotted sides but this is physically impossible to do to the accuracy needed
and the cell performance depends on plate spacing to very high accuracy and slot width to even higher accuracy.
This is not a backyard construction quality job and there are very, very few people with both the equipment and
skill to complete the construction to this degree of accuracy.

The base and two sides of the cell could have grooves cut in them to take the plates. This is not a good idea for
various reasons, including the fact that the steel plates expand when they warm up and are liable to crack the
acrylic case unless the slots are cut deeper than normal. Also, it is difficult to cut very accurate slots in acrylic due
to the heat of the cutting blade causing the acrylic to deform in the immediate area. Grooved acrylic is very much
weaker and breaks easily due to the planes of weakness introduced into the material.

Using Ultra High Molecular Weight Poly Ethylene or High Density Poly Ethylene (food chopping-board material)
strips is a much better technique as that material does not have the same cutting heat problem and it can also
take the plate expansion much better, so it is the construction method of choice. It is also a cheaper material.

The grooves which are cut for the plates should be three thousandths of an inch wider than the thickness of the
plates. A good plate thickness is 16 gauge sheet which is one sixteenth of an inch thick or 0.0625 inch (1.5875
mm), so the recommended groove width for that is 0.0655 inches which is not a convenient fraction being about
four and one fifth sixty-fourths of an inch. The grooves are 1/8” (3 mm) deep.

10 - 67
The supplier of the acrylic sheet needed for making the case, will be able to supply “glue” specifically designed for
joining acrylic sheets together. This glue actually welds the plates together so that the sheets become one
continuous piece of acrylic along the joint. Start by mating the sides and the base. Insert two or three plates into
the slots to be quite sure that the alignment is spot-on during the joining process. Line the ends up during jointing
to be sure that the sides are completely square when being joined to the base.

Concerns have been expressed about the strength of the acrylic casing under severe road conditions. So it has
been suggested that the acrylic components be constructed from sheet which is 3/4” to 1” thick (18 mm to 25 mm)
and the corners reinforced with angle iron secured with bolts tapped into the acrylic as shown below.

10 - 68
Here is a photograph of a 101-plate housing built by the late Ed Holdgate who worked to a very high standard of
accuracy:

This housing looks very simple and straightforward, but this is highly misleading and the materials are very
expensive, so any error is costly. The construction accuracy needed is very high indeed with many opportunities
for a total and expensive disaster. Ed Holdgate has built several custom fixtures to ease the construction, but
construction is still very difficult even with these specialist fittings and his years of experience. "Sikaflex 291" or
"Marine Goop" marine bedding compound can be used to seal between the two slotted sides and the slotted
base, and between the slotted sides and the two end inserts, in order to prevent any leakage between the acrylic
and any of these inserts.

The accuracy required for the slots to hold the stainless steel plates is 0.0003” and the plates are tapered with a
belt sander on both sides along all four edges so that when they are forced into the slots they will not cut into the
sides of the slots. This produces excellent leakage characteristics, but don’t lose sight of the very high accuracy
of the slot cutting needed for this. The edges of the slotted inserts receive a bead of Sikaflex marine bedding
compound attaching them to the acrylic box and the compound is allowed to cure before construction is
continued.

The end plates with the stainless steel straps welded to them are used to connect the electrical supply to the
plates, keeping any connection which could possible work loose and cause a spark, completely outside the
housing. Even though the straps are welded and there is no likelihood of them coming loose, the welds are still
kept below the surface of the electrolyte.

10 - 69
Getting and Preparing the Plates:
A set of 101 plates is needed for the electrolyser. The material used when making the plates is very important. It
should be 16-gauge 316L-grade stainless steel as it contains a blend of nickel and molybdenum in the correct
proportions to make it a very good catalyst for the pulsing technique. You can try your local steel stockists to see
if they can supply it and what their charges would be. One satisfactory stainless steel supplier which Bob has
used is Intertrade Steel Corp., 5115 Mt. Vernon Rd SE, Cedar Rapids, IA 52406. Do not buy from eBay as you
have no real comeback if the plates supplied are dished due to having been flame cut.

It is very important indeed that when asking for a quote that you make sure that the supplier is aware of the
accuracy you require. The plates need to be flat to a tolerance of +/- 0.001 inch after cutting and this is the most
important factor. That level of accuracy excludes any kind of flame cutting as it produces inevitable heat distortion.
With shearing, expect +/- 0.015 inch on the cuts and +/- 0.001 inch on flatness. Laser cutting produces much
higher accuracy and you can expect as good as +/- 0.005 inch on cuts and there is no spec needed for flatness
since laser cutting does not distort the edges like shearing does.

The plates are square: 6 inches by 6 inches, but that does not represent 36 square inches of active surface area
as some plate area is inside the grooves and some of each plate is above the surface of the electrolyte. Another
point to remember is that 101 steel plates this size weigh a considerable amount and the completed electrolyser
with electrolyte in it will weigh even more. It is essential therefore to have a case which is strongly built from
strong materials, and if a mounting bracket is to be used, then that bracket needs to be very robust and well
secured in place.

The preparation of the plates is one of the most important steps in producing an electrolyser which works well.
This is a long task, but it is vital that it is not skimped or hurried in any way. Surprisingly, brand new shiny
stainless steel is not particularly suitable for use in an electrolyser and it needs to receive careful treatment and
preparation before it will produce the expected level of gas output.

The first step is to treat both surfaces of every plate to encourage gas bubbles to break away from the surface of
the plate. This could be done by grit blasting, but if that method is chosen, great care must be taken that the grit
used does not contaminate the plates. Stainless steel plates are not cheap and if you get grit blasting wrong, then
the plates will be useless as far as electrolysis is concerned. A safe method which Bob much prefers is to score
the plate surface with coarse sandpaper. This is done in two different directions to produce a cross-hatch pattern.
This produces microscopic sharp peaks and valleys on the surface of the plate and those sharp points and ridges
are ideal for helping bubbles to form and break free of the plate.

Bob Boyce uses a specially widened 48-inch belt sander which is good for preparing the plates using 60 or 80 grit.
However, most people don’t have this equipment and do the sanding by hand. Bob stresses that when doing
hand sanding the sandpaper is drawn across the plates in one direction only and not backwards and forwards, as
the backwards stroke always destroys the perfectly good ridges created on the forward stroke. Also, you only
need two strokes in one direction before turning the plate through ninety degrees and completing the sanding of
that face of the plate with just two more strokes (again, with no backstroke). Most people want to sand the
plates far too much and if overdone to a major degree, that can reduce the plate thickness and cause electrolyte
leakage through the slots around the plates. So, to say it again, to sand one face of a plate, use just two strokes

10 - 70
in one direction, turn the plate through ninety degrees and finish that face with just two more strokes, both in the
same direction.

Always wear rubber gloves when handling the plates to avoid getting finger marks on the plates. Wearing these
gloves is very important as the plates must be kept as clean and as grease-free as possible, ready for the next
stages of their preparation.

Any particles created by the sanding process should now be washed off the plates. This can be done with clean
tap water (not city water though, due to all the chlorine and other chemicals added), but only use distilled water for
the final rinse.

A point which is often missed by people constructing electrolysers is the fact that electrolysis is not just an
electrical process, but it is also a magnetic process. It is important for maximum operating efficiency that the
plates are aligned magnetically. In theory, stainless steel is not magnetic, but much of the stainless steel actually
supplied to builders is slightly magnetic. When the plates arrive from the supplier each plate may have random
magnetic characteristics. The easiest way to deal with this situation is to try to give the plates a mild magnetic
orientation. This can be done quite simply by wrapping a few turns of wire around the stack of plates and passing
some brief pulses of DC current through the wire.

Obviously, the plates need to be kept in the same direction when being slotted into the case. The next step in the
preparation process is to make up a weak solution of potassium hydroxide. This is done by adding small amounts
of the potassium hydroxide to water held in a container. The container must not be glass as that is not a suitable
material in which to mix the electrolyte.

Potassium hydroxide, also called KOH or “Caustic Potash”, which can be bought from various suppliers such as:
http://www.essentialdepot.com/servlet/the-13/2-lbs-Potassium-Hydroxide/Detail
http://www.organic-creations.com/servlet/the-653/caustic-potassium-hydroxide-KOH/Detail or
http://www.nuscentscandle.com/ While Potassium hydroxide (KOH) and Sodium hydroxide (NaOH) are the very
best electrolytes, they need to be treated with care. The handling for each is the same:

Always store it in a sturdy air-tight container which is clearly labelled "DANGER! - Potassium Hydroxide". Keep
the container in a safe place, where it can’t be reached by children, pets or people who won't take any notice of
the label. If your supply of KOH is delivered in a strong plastic bag, then once you open the bag, you should
transfer all its contents to sturdy, air-tight, plastic storage containers, which you can open and close without
risking spilling the contents. Hardware stores sell large plastic buckets with air tight lids that can be used for this
purpose.

When working with dry KOH flakes or granules, wear safety goggles, rubber gloves, a long sleeved shirt, socks
and long trousers. Also, don’t wear your favourite clothes when handling KOH solution as it is not the best thing to
get on clothes. It is also no harm to wear a face mask which covers your mouth and nose. If you are mixing solid
KOH with water, always add the KOH to the water, and not the other way round, and use a plastic container for
the mixing, preferably one which has double the capacity of the finished mixture. The mixing should be done in a
well-ventilated area which is not draughty as air currents can blow the dry KOH around.

When mixing the electrolyte, never use warm water. The water should be cool because the chemical reaction
between the water and the KOH generates a good deal of heat. If possible, place the mixing container in a larger
container filled with cold water, as that will help to keep the temperature down, and if your mixture should “boil

10 - 71