GlassWare

AUDIO DESIGN

High-Voltage Power Supply & Heater Regulator

USER GUIDE
04/02/2010

Page 1 PS-3 PCB Overview

1 Introduction to Power Supplies
2 Standard Power Supplies
2 The Function of a Power Supply
2 Wall Outlet Voltage Problems
4 The Parts of a Power Supply
6 Rectifiers
6 Capacitors
9 Types of Electrolytic Capacitors
10 Capacitor Safety
Contents 11 Rectifier Circuits
12 Voltage Doubler Circuits
13 Voltage Regulators
14 PS-3 HV Schematic & Part Values
15 Heater Power Supply & Regulator
17 Modifications & Part Upgrades
17 Operation
18 Assembly and Testing
19 Grounding
20 PS-3 PCB Dimensions
 DANGER!
This PCB holds a high-voltage power supply; thus, a real—
and possibly—lethal shock hazard exists.

Ideally, a variac should be used to slowly power up the

regulator, as it is better to have a mis-oriented electrolytic
capacitor or a mis-located resistor blow at low voltages,
rather than at high voltages. Remember that the danger
increases by the square of the voltage; for example, 200
volts is four times more dangerous than 100 volts and 400
volts is sixteen times more dangerous.

Once the power supply is powered up, be cautious at all

times. In fact, even when the power supply is disconnected
or shut down, assume that power-supply capacitors will have
retained their charge and, thus, can still shock. If you are not
an experienced electrical practitioner, before attaching the
transformer windings to the board, have someone who is
well-experienced in electronics review your work.

There are too few tube-loving solder slingers left; we cannot

afford to lose any more.

GlassWare
AUDIO DESIGN
www.glass-ware.com
www.tubecad.com

sales@glass-ware.com

Copyright © 2010
All Rights Reserved
 PS-3 PCB
The PS-3 circuit board makes li fe quite a bit easier for the tube fancier. On this small
four by four inch, extra thick (0.094), 2oz-copper traces, USA-made PCB resides both
a high-voltage power supply and a low-voltage regulator, with each finding i ts own raw
power supply, including the rectifiers and power-supply reservoir capacitors. In other
words, except for the power transformers, the PS-3 PCB holds all that is needed to
make a fine tube-circuit power supply.

The low-voltage regulator is meant to power the tube heaters; the high-voltage power
supply, the rest of the tube circuit. The low-voltage regulator’s power supply is flexible
enough to allow ei ther a conventional full-wave-bridge rectifier arrangement or a two-
recti fier voltage-doubler configurati on, allowing for more latitude in power
transformer selection; thi s regulator can be set to either 6.3V or 12V or 12.6V.

The hi gh-voltage portion of the PS-3 power supply uses a high-speed solid-state full-
wave bridge rectifier circuit and a si mple RC filter to achieve a fairly qui et B+ voltage,
which can span from 200V to 300V. (The series resi stor can be replaced by an external
choke.)

In addition, the PS-3 can DC reference the heater regulator to ground or 1/4 of the B+
voltage or left floating, with only a coupling capacitor to ground.

Introduction to Power Supplies

For most modern electronic products—printers, alarm clocks, LCD photo frames—the
power supply is the least important portion of the package. Which helps explain why
cheap, small, switching, universal, wall-wart power supplies fill every available outlet i n
most houses. All proficient tube-circuit designers must master the subtleti es of creating
a clean DC voltage. For although a heater may be powered by either an AC or DC
power supply, a DC power supply is always necessary, as the remaini ng portion of the
tube circuit requires solely DC voltage.

With audio circuits, the cleaner, the steadier, the lower and the flatter in output
impedance across the audi o band of frequencies, the better suited that power supply is
for audio use. A good metaphor is that the audio gear is the film in a movie projector
and the high-intensity li ght source is the power supply; the whiter and steadier the
light, the better the picture.

Paradoxically enough, hi gh-voltage power supplies are at once both the easi est and the
most challenging element of tube electronics. Requiring as li ttle as one transformer,
diode, and capacitor, what could be simpler? Or utilizing as much as a several power
transformers, many precision voltage references, twelve rectifier bridges, a score of
high-voltage MOSFETs, a fistful of OpAmps, dozens of capaci tors, a cup of individual
diodes, and several hundred resistors, what could be more complex?

Thi s high voltage power supply belongs to the simpler camp. Its tube rectifier allows a
slow start up voltage and provides that certain something, which only tube rectifiers
can deliver. The RC fi lter used i s a simple affair, consisting of single power resistor
and addi tional shunting capacitor. What follows i s an overview of power supplies i n
2 PS-3
Standard Power Supplies
A “standard power supply” is one that is not regulated. A standard power supply can
be simple or complex, unipolar or bipolar, low voltage or high voltage, low current or
high current, vacuum tube rectified or solid state rectified, choke-free or choke-filled—
the only disqualification is having a regulator. As most regulated supplies must also
incorporate a standard power supply as their voltage source, an understanding of the
standard power supply is primary to understanding voltage regulation.

The Function of a Power Supply

Obviously, a power supply’s purpose is to provide a power source for a circuit.
However, we should take one step back and consider why the wall outlet cannot serve
directly as a power supply for a tube circuit, since it does so for a lamp or toaster. The
simple answer is that the wall outlet provides an AC voltage and audio tube circuits
require a DC voltage. A more complete answer is that a power supply not only
converts the AC voltage into DC, additionally, it converts the value of the AC voltage
and it modifies the resulting DC voltage by smoothing, stabilizing, current limiting,
and removing much of the radio frequency interference (RFI) from it. All of these
functions are not needed by a lamp or toaster, but are essential for a tube circuit.
Actually, the only safe assumption concerning the wall voltage is that its time base is
accurate. Beyond frequency, there is much latitude—sadly, too much latitude.

Wall outlet voltage problems:

 Voltage varies
 RFI contamination
 AC voltage may contain a DC component
The wall voltage is not as fixed as you might imagine, as the wall voltage varies from
city to city, and within the same city, from hour to hour. It has slowly risen over the
decades. For example, truly old transformers bear stickers that specify a 110Vac or
112Vac input voltage; whereas, a newer transformer sticker might read 117Vac or
120Vac. In many new developments, the wall voltage runs hot, at 122Vac or hotter;
whereas in an old Victorian apartment, it can run cold, at 110Vac or less, because of
the long skinny wall wiring used. The wall voltage varies depending on the varying
load presented to the power line feeding your house. During peak use it drops and,
when unburdened, it climbs. (Indeed, if a power line on the street is ruptured, thus
unloading the load down from the break, the voltage can climb to scary heights.)
Although a power-supply designer would disagree, as far as the utility company is
concerned, +/-10Vac variations in wall voltage are perfectly acceptable. But these
voltage variations can cause trouble. A standard power supply’s output voltage will
vary with the change in wall voltage. For example, a power amplifier’s B+ voltage
could drop or rise 50 volts from being used in one city versus another. Now, +/-50
volts is a lot of voltage variation; in fact, it might be more than half the voltage drop
across a pass device in a high voltage regulator.
 GlassWare Audio Design 3
Maximum and Minimum Wall Voltages A reasonable guess on the absolute upper
limit of sustained higher line voltage would be around 127vac in the USA, as the
power company—which could be held liable for damaging your microwave oven, if
the voltage became too hot—is very quick to prevent an over-voltage. The lower limit is
usually set to around 108vac here in the USA. You can, however, choose to design
around an even wider variation, say 100-130vac.
RFI The art and science of stripping away RFI can fill a large book. Radio frequency
interference comes from more than just radio stations. Light dimmers, computers,
cable TV, hair dryers, switching power supplies, industrial electronics—all can pollute
the wall voltage with RFI. In general, assume that the wall voltage is moderately
contaminated with RFI and provide some means for scrubbing it clean. Sometimes
adding a small shunting capacitor after the fuse on the primary is all that is needed; at
other times, many complex RF filters are required on both the primary and secondary.
AC Plus Some DC Every electronic student knows that the wall outlet provides AC
voltage; few, however, know that it might carry a DC component as well. This DC
voltage can be as great as a few volts and it is usually the result of an industrial power
supply back-feeding some DC into the line. Since the power supply contains a
transformer, we imagine that this DC voltage can have no effect on our power supply.
Unfortunately, we are not that lucky.
The transformer was designed for a strictly AC use, which means that its core was not
designed to be used in the presence of a DC current. If a DC current passes through a
transformer winding, the core will tend to polarize, that is the molecules within the
iron core material will move from their flexable orientation into one unified
direction. The result of polarization of the core is that the transformer begins to
resemble an electromagnet, which can quickly saturate the core material and radically
decrease its effective inductance. In other words, the presence of DC across a winding
of a normal AC transformer will make the transformer much less effective at
transferring energy.
Paradoxically, the better the transformer, the worse the saturation problem. A toroidal
transformer is much more sensitive to a DC current, as it is a much tighter inductive
design. In contrast, the standard EI-core transformer is a looser inductive design and
this looseness allows it to sustain a greater amount of saturation. What is going on
here is that all transformers contain some sloppiness in their electromagnetic circuit,
as their physical construction is not perfect, and this sloppiness constitutes a parasitic
air gap. This air gap works like the purposeful air gap in a single ended transformer by
increasing the saturation threshold of the transformer’s core.
Beyond reduced efficiency and poorer regulation, what are the ramifications of a DC
component on the wall voltage? Here is a possible scenario: you have just built a
supremely wonderful single ended amplifier, an amplifier so sweet and bewitching that
if Shakespeare had heard it, he would have been compelled to write an additional
twenty sonnets. Yet, when you bring it to your friend’s house, it sounds unbelievably
noisy and harsh. What happened? Probably his wall connection contained a unhealthy
amount of DC and this so saturated the core of the power transformer that its
secondary voltage dropped, its ripple component increased, its ability to contain its
magnetic field reduced to the point of radiating the field across the input circuitry
and output transformer, which then coupled it to the speakers.
4 PS-3

The Parts of a Power Supply

Transf ormers Power transformers are used to isolate, to convert, and to filter the wall
voltage. A transformer is made up at least two tightly coupled coi ls, whi ch usually
enshroud a ferromagnetic core. The coils are both physically cl ose and
electromagnetically coupl ed. By means of the electromagnetic induction, AC voltage is
transformed from one voltage to another from one coi l to another, whi le the
frequency remains the same. If the voltage on the primary and the secondary equal ,
the transformer is called an i solati on transformer. If the vol tage increases, the
transformer is called a step-up transformer; if it decreases, a step-down transformer.
The rati o of turns in each winding determi nes the rati o of voltage change. Thus, if the
pri mary wi nding, the wi ndi ng that is connected to the wall outlet, has a hundred
turns and the secondary windi ng, the winding that feeds the rectifyi ng circui t, has a
thousand turns, the winding ratio will be 1 to 10, or 1:10, which means that the
transformer will step up the wal l voltage by a factor of ten. As there are no free
lunches in physics, if the voltage on the secondary i ncreases by tenfold, the current
must decrease by tenfold. In other words, the wi nding ratio speci fi es the vol tage ratio
and the inverse of the current ratio.

Standard EI-Core Transformers Most tube power supplies use an EI-core

transformer. The outer appearance of this transformer may be an industrial looking
open frame construction or classic tube amplifier end bell style or a shiny potted case.
Nonetheless, internal to each is an ferromagnetic core, which is made up of flat, thin,
pieces of soft iron called laminations that come in two shapes: an E and an I. These
two shapes are placed side by side so as to define a closed H shape. One E is followed
by one I on the layer, but on the next layer an I is followed by an E, then back to an
E followed by an I for the next layer and on and on…until the desired thickness is
achieved. This interlacing of E’s and I’s makes for an electromagnetic circuit almost
as tight as a core made of solid iron.
This begs the question: Why not use a core of solid metal instead of the many
laminations? The answer lies in the fact that a solid core would suffer from a
deleterious effect known as Eddy currents. These electrical currents rob the
transformer of efficiency. Imagine a doughnut made of iron, around a section of
which several turns of wire have been wound. Now, if a DC current is applied to the
winding, the induced lines of magnetic flux will circle the around the core, like bb’s
in a hula hoop, which is what we desire. But the same circling induces an
electromagnetic field within the core itself, which creates perpendicular currents in
the core material. These currents, as they only generate heat within the core material,
drag down the power transfer.
 GlassWare Audio Design 5
Fortunately, Eddy currents are proportional to the square of the core thickness.
Therefore, if the core is made thin enough, we can ignore the Eddy currents. This what
the laminations do, as they are very thin and electrically insulated from each other by
wax or shellac.
Large and heavy, typical EI-core transformers do not seem suited to modern sleek and
slender enclosures. Additionally, they present relatively higher winding DC resistances
and, as a result, deliver a more load-dependent secondary voltage. Their greatest failing,
however, is the large radiated magnetic fields that emanate from the bell sides of the
transformer. In other words, placement of this type of transformer within a chassis is
critical, as hum can be injected into an otherwise quiet circuit by a ill-placed
transformer. Yet, in spite of all of this, EI-core transformers often set the standard in
audio gear and should not be demised out of hand.
Flat Pack Transformers Often mounted on printed circuit boards, these fairly
modern designs pack a good amount of power in a small package. Wider and longer
than they are tall, the flat pack transformer is also made up of E and I laminations and
four bobbins of wire. This last feature is what most interests us. Most transformers
have one winding, primary and secondary, sitting on top the other, as it helps to
further couple them. The disadvantage of such an arrangement lies the increased
capacitive coupling afforded by the closeness of the windings. This capacitance will
couple undesirable RFI noise from the primary to the secondary. In contrast, the flat
pack’s greater distance between the primary and the secondary greatly reduces the
capacitive coupling between windings; thus, greatly reducing the RFI transference. (The
core is just too slow to react to the extreme high frequencies of the RFI.)

Flatpack Toroidal

Toroidal Transformers Theoretically, the toroid is one of best choices of the easily
available transformers. But, like so many good theories, practice proves difficult. First,
the attributes: low radiated magnetic fields, better regulation, smaller, lighter, and
cooler running. The obvious downside is higher cost. Where a conventional EI power
transformer can be easily and quickly assembled by simple machines, toroids require
much more labor and elaborate machines. The next disadvantage is surprising: toroids
can be dreadfully noisy, not in terms of magnetic fields, but in terms of mechanical
buzzing. Even the wire itself can buzz. Finding a low-mechanical-noise toroid can be
quite difficult. Additionally, toroids often pass much more RFI than a comparable
clunky EI power transformer would. Because the toroid’s primary and secondary lie
next to each other, there is more capacitive coupling between the windings.
6 PS-3
Furthermore, toroids easily saturate in the presence of a sustained DC current through
a winding. Because the magnetic circuit is so tight in a toroid, any DC current flow
can magnetize the toroidal core, thereby robbing the transformer of its effectiveness. In
other words, toroids are perfect, except that they aren’t. My recommendation is to only
use a toroidal power transformer when you have no other choice because of space
limitations. And when you do use one, be sure to use it conservatively (which means
that much of the smaller size benefit will be lost because of larger toroid being used).

Rectifiers
Rectifier only allow unidirectional current flow, an attribute needed in converting AC
into DC. Tube and solid-state rectifiers can be used in high-voltage power supplies, but
the vacuum tube’s limited current rating and high voltage drop prohibits its use in a
low-voltage power supply. One great advantage the tube rectifier holds over its solid-
state brother is its slow turn on (not all tube rectifiers are slow to conduct, however);
additionally, tube rectifier can more often survive high-voltage surges from the wall-
socket or inductive kickbacks. Sonically, tube rectifiers usually win in shootouts, but
the best solid-state rectifiers, such as HEXFREDs, easily beat the worst tube rectifiers.

Capacitors
Just as a resistor is the practical embodiment of pure resistance, a capacitor is the
practical embodiment of pure capacitance. Unfortunately, a perfect embodiment is
difficult to produce. The lead wires and electrodes/plates have some resistance and all
physical dielectrics (the stuff between the electrodes/plates) have some conductance.
The dielectric material acts as an amplifier of sorts, as it “amplifies” the small capacity
to hold a charge that would otherwise exist in its absence. By allowing a small
capacitor to hold a large charge, however, the dielectric’s imperfections are also
“amplified.” Since most materials change both in shape and in resistance with
temperature, we would expect many actual capacitors to degrade in performance with
changes in temperature and they do. In other words, a perfect capacitor could be made
if no material were used in its construction, just as perfect society could be formed if it
were not populated with people. Therefore, the dielectric makes for the biggest
deviation between the perfect capacitor and the actual capacitor.
In a power supply, the capacitor stores and smoothes the rectified DC voltage. In
between audio stages, a coupling capacitor freely passes the AC signal while isolating
the different DC levels. As can be expected, each function demands its own preferred
capacitor style, with the power supply capacitor relying on its quantity of capacitance
and the AC signal transferring capacitor relying heavily on the quality of the capacitor.
A capacitor has four frequently published specifications: its amount of capacitance, its
voltage limit, its temperature limit, and its tolerance. An added piece of information is
the polarity indicator on polarized capacitors, such as electrolytics.
Voltage Limits Capacitors often bear two voltage ratings on their bodies: a DC rating
and an AC rating. The AC rating will be the lower of the two. A sinusoidal wave peaks
at 1.4 times its equivalent DC voltage. Consequently, if the capacitor only bears one
voltage rating, assume that it refers to the DC rating and de-rate the capacitor when
using it in an AC application.
 GlassWare Audio Design 7
Here’s an extreme example. Fearing ultra-high-frequency noise on the power supply’s
output, an audiophile places a small-valued capacitor across the power transformer’s
secondary. Now, since the power supply yields a nominal 400-volts, a 400-volt high-
frequency-shunting capacitor is soldered in place, say 0.1-µF. But as the power supply
comprises a full-wave center-tapped-transformer arrangement, the capacitor sees not
400-volts, but 800-peak-volts! Even worse, if a large choke connects directly to the
diodes (choke loading), then the peak voltage across the leads is 1130-volts, not 400-
volts. Considering that the wall voltage may hold momentary peaks of much higher
voltages, a safe rating might be 1600-volts, quite a distance from 400-vdc.

Temperature Limits There has to be a temperature limit to every capacitor, as no

capacitor can survive on the Sun’s surface for example. With film capacitors, such as
Mylar or polystyrene, the temperature-limit is seldom printed on their bodies,
although this limit can be found in the manufacture’s spec sheet for the capacitor.
Because plastic films melt at a much lower temperature than do glass, mica, porcelain,
or ceramic, care should be taken not to momentarily overheating film capacitors
during soldering, particularly with polystyrene capacitors. For example, a 1% capacitor
can easily be thrown off a few percent by such an overheating, as the heat causes the
plastic to change in shape. In contrast, electrolytic capacitors always (well, almost
always) display a temperature limit. This limit should never be approached let alone
exceeded, as the capacitor can be easily damaged and it just might rupture. In
addition, an electrolytic capacitor’s useful life expectancy is shortened drastically with
excess heat, as the wet part (the electrolyte) evaporates with heat, leaving the capacitor
greatly reduced in capacitance.
Equivalent Series Resistance (ESR) Ideally, a capacitor should hold no ESR, as an
ideal capacitor dissipates no heat and loses no voltage across its leads, but most
capacitors do both. Real capacitors are made with metal, thin metal at that, so we
should expect some resistance due to the metal, from the leads, the electrodes, and the
slight resistance from the solder that binds the leads to the electrodes, but not the
dielectric effective residual resistance. Plastic film-foil capacitors display a lower ESR
than metalized-film capacitors, as foil has a lower resistance. Placing multiple
capacitors in parallel lowers the effective ESR.
Heat is the capacitor’s enemy. So keeping a capacitor cool is important and ESR
works against us. High ripple current causes a capacitor to self-heat, just as a resistor
would facing the same currents. A capacitor’s construction, unfortunately, limits its
heat transfer to its surroundings. The heat dissipation in a capacitor is equal to I2 x
ESR, where I is the RMS ripple current.
Equivalent Series Inductance (ESL) Like ESR, ESL represents the inductance in
series with the capacitance in the capacitor. This is an important consideration for
high-frequency power supply filtering and other high-frequency applications as well.
Placing multiple capacitors in parallel lowers the effective ESL, but the best plan
might be to use specially-designed high-frequency electrolytic capacitor instead.
Dissipation Factor (DF) DF is equal to ESR/XC. DF is much more important in
power supply applications than it is in signal-coupling applications. DF is used in
predicting the amount of heat a capacitor will dissipate.
Quality Factor (Q) The inverse of DF is the quality factor (Q). Thus, Q = ESR/XC.
8 PS-3
Electrolytic Capacitors This type of capacitor displays the poorest temperature
stability, the most leakage, the most noise, and the shortest lifespan of any capacitor.
NASA prohibits the use of electrolytics in space flight, because of high failure rate and
their possible venting of noxious fluids and gases. The reason we put up with the
electrolytic is because of its volumetric efficiency; we hunger for capacitance, lots of
capacitance. A tube power amplifier’s power supply needs at least 50-µF and solid-state
amplifier’s need closer to at least 10,000-µF. 10,000-µF of polypropylene would be
hugely expensive (more than most cars) and hugely bulky (about the size of a small
car). So instead we make do with the electrolytic. And it must be admitted: electrolytic
capacitors have increased in quality substantially over the last decade.
All electrolytic capacitors belong to the polarized group and none should ever be
reversed biased, i.e. hooked up backwards, as the capacitor conducts extremely well in
the reverse direction. In fact, as little as 0.4 volts (depends on capacitor temperature)
of reversed voltage can damage an electrolytic permanently. This bring up the issue of
the electrolytic’s short life span, which is measured in thousands of hours, as in 2,000-
hrs. Long life electrolytics might be good for 4,000-hrs. Why so short a lease on life?
Basically, the volumetric efficiency came at a price. No doubt some are thinking that a
lifespan of just 2,000 hours sounds far to short, as they own 40 year old stereo
equipment, stereo equipment filled with electrolytics, stereo equipment that still
works. The answer lies in what the 2,000-hour life span means. The life rating spells
out how long the capacitor will survive under its maximum voltage and temperature
ratings. Reducing the applied voltage by only 20% increases its lifespan by threefold
and reducing the applied voltage by 50% increases its lifespan by 8 times. The formula
for estimating electrolytic capacitor life is:
3 (T/10)
L = LR x (ER/EO) x 2
where,
L = the estimated resulting lifespan
LR = the life at rated limits
ER = rated voltage limit
EO = operating voltage
T = rated temperature minus core temperature in C
Capacitor life roughly doubles for each 15°C reduction in operating temperature; so if
the electrolytic capacitor is used in a cool environment at a reduced voltage, the
capacitor lifespan should exceed the useful life of the piece of equipment.
Unfortunately, with tube equipment that turns out to be a big “IF,” as the high-
voltage and many heat radiating parts work against us. For example, a solid-state
amplifier with a 30-volt power supply rail would most likely use a 50v electrolytic
capacitor, which means that the B+ voltage is only 60% of the capacitor’s maximum
rating. On the other hand, to make the same ratio, a tube amplifier with a 450-volt
rail would need a 750v capacitor, which is unlikely. More than likely, a 450v or 500v
capacitor would be used, giving only a 0% to 10% voltage safety margin.
Paradoxically, electrolytic capacitors must be used so as not to go bad. When left
uncharged, the electrolytic capacitor loses capacitance over time and must be
“reformed” to restore its full capacitance and reduce its leakage.
 GlassWare Audio Design 9
Reforming of the dielectric requires applying a current limited voltage to the
capacitor, until the excessive current draw falls off. Usually, a high-ohmage resistor,
say 1-10-meg, is placed in series with the capacitor and a variable voltage power supply
set to the capacitor’s maximum voltage. Once the voltage drop across the resistor falls
to a reasonable level, the capacitor will have been reformed. This can take as little as
one minute or hours. I recommend not using old electrolytic capacitors or at least
throwing away any capacitor that takes over 10 minutes to form. The resistor will get
hot, so avoid touching it and wear safety goggles when you form old electrolytic
capacitors. And never plug-in and turn on an old piece of electronic equipment that
hasn’t been used for decades or even years, as the electrolytic capacitors might vent or
even explode. Instead, plug the equipment into a variac and slowly bring up the
voltage over a few minutes. This procedure works well with solid-state equipment or
tube gear with solid-state rectifiers, but less well with tube rectifier based gear. The
work around is to tack solder high-voltage solid-state rectifier across the tube rectifier’s
socket pins and slowly bring up the voltage to about 80% of the wall voltage (100%
could damage the capacitors, as the solid-state rectifiers yield higher voltages); then
turn off the unit and discharge the capacitors before removing the solid-state rectifiers
.

Types of Electrolytic Capacitors

Electrolytic capacitors are a headache, but not all electrolytic capacitors are equal. The
capacitor manufacturers have developed several optimized electrolytic capacitors that
are worth looking into:
Computer Grade Big and blue, these capacitor stand out. The computer grade
electrolytic capacitor gives up volumetric efficiency for reliability. This capacitor is
designed to take a beating from high ripple currents. Their use is limited to the power
supply.
HF Electrolytic Panasonic and others make electrolytic capacitors that trade some
volumetric efficiency for improved high frequency behavior. ESL is reduced in these
capacitors, making them effective at higher frequencies than a generic electrolytic
capacitor. Highly recommended for cathode resistor bypassing, power supply use, and
as coupling capacitors in OTL amplifiers. Unfortunately, they are usually only
available up to 100-volts.
Low ESR Electrolytic ESR and high current do not go together. Consequently, low
ESR capacitors are useful as a power supply reservoir capacitor, as this capacitor must
see a high ripple current.
Photo Flash A special type of electrolytic capacitor, these capacitors were designed
for use with a camera’s flash unit. They store a relatively large high-voltage electric
charge and they can deliver the charge quickly, which handsomely suits the needs of a
flash bulb. What about audio; how well do these capacitors work with tube
equipment? Fairly well is the quick answer, if a few key provisions are followed. The
photo flash capacitor’s large volumetric efficiency betrays that these capacitors cannot
tolerate a steady and heavy use. Using them near their rated voltage or rated
temperature is sure to shorten their life. So too using them in a power supply, as the
main filter capacitor, is sure to kill them.
10 PS-3
So what good are photo-flash capacitors? These capacitors are quite suitable for low-
ripple power supply use; say for the first stage of an amplifier’s RC filter, after being
fed by a high-ohmage voltage-dropping resistor. Another use is as the output coupling
capacitor in an OTL amplifier. In fact, one OTL manufacturer found that they
worked perfectly in that position. Wouldn’t the high current demands of the
loudspeaker ruin this capacitor? No, although the current peaks can be quite high,
unless you are exclusively listening to organ music at high volumes, you are not
actually delivering very much steady current into the loudspeakers.
Bypass Capacitors Because the power supply workhorse known as the electrolytic
capacitor is so clunky, many advocate bypassing it with a smaller-valued, high-quality
film or oil capacitor. This can work supremely well, but it can also backfire. Many
electrolytic capacitors present a large amount of inductance (ESL), which will oscillate
in combination with a pure capacitance, which many bypass capacitor approach. The
workaround is to place a resistor in series with the bypass capacitor (usually equal to
the electrolytic capacitor’s ESR).
Two high-voltage capacitor in series High-voltage capacitor can be hard to find,
particularly if the voltage rating needs to be over 450V. This predicament can be
overcome by placing two high-voltage capacitor in series, which effectively doubles the
voltage rating, but halves the total capacitance. The danger in trick lies in the
capacitors not uniformly charging up, which will expose one to excessive voltage,
thereby destroying or damaging the capacitor. The workaround is to place a high-
voltage, high-wattage, high-ohmage (100k to 1M) resistor in parallel with each
capacitor in the series. These added resistors will work to prevent dissimilar voltage
drops across either capacitor, dividing the DC voltage equally between capacitors.

Capacitor Safety
 Always assume that capacitor holds charge and you will never be shocked to
find that in fact it did. Oil capacitors are notorious for retaining a large
charge after having been discharged.
 Always use a resistor to short out a capacitor and never short the capacitor’s
leads together, as the capacitor’s own lead wire can melt and sputter from a
heavy discharge without the current limiting that the resistor imposes.
 Always observe a polarized capacitor’s orientation, as some polarized
capacitors have gone off like little hand grenades after being inserted
backwards.
 Always shroud the exposed leads of high-voltage capacitors to reduce possible
shorts or shocks.
 Always ensure adequate clamping of any axial capacitor much larger than 5-
watt resistor, as an unsupported heavy capacitor will bounce and jiggle with
vibration or movement and break free of one its leads. This is particularly
true with power supply capacitors, as the constant vibration from the power
transformer can cause the capacitor’s leads to fatigue and snap.
 GlassWare Audio Design 11

Rectifier Circuits
Tube equipment was born to run on DC and everywhere it is chained to AC wall
sockets. In other words, all power supplies that plug into wall sockets, must convert
AC voltage to DC voltage. This conversion requires rectifiers and capacitors; and there
are several ways that the rectifiers can be configured.

Full-Wave Bridge This is the most popular

power supply configuration. The entire
primary winding is used and four rectifiers
are required. This configuration is seldom
used with tube rectifiers, as the rectifier
cathodes cannot be heated by just a single
heater winding. The two diode voltage-drops
mean little in a high-voltage power supply, Iout = Iac / 1.8
but are a big liability in low-voltage power Vdc = (Vac x 1.4) - 2Vdiode
supplies.

Full-Wave Center-Tap This is the classic tube

amplifier power supply configuration. It uses only half
of secondary winding at a time and requires a center-
tapped winding and two rectifiers, saving two
compared to the full-wave bridge configuration. On
the other hand, the two rectifiers will see twice the
peak reverse voltage that four rectifiers would see in a
full-wave bridge configuration, which can greatly limit Iout = Iac / 1.27
the use of solid-state rectifiers in high-voltage power Vdc = Vac* / 1.43 - Vdiode
*entire primary
supplies; in addition, the primary is only about 75%
as effectively utilized in this configuration compared
to the FW bridge circuit. Paradoxically enough for
what many consider a purely high-voltage circuit, this
rectifier circuit is often used in extremely low-voltage
power supplies, as the single-diode voltage drop greatly
increases this circuit’s efficiency.
L
Full-Wave Choke Input This power-supply
configuration is often used with tube
rectifiers and requires quite a bit of tuning.
When designed correctly, this power supply
topology yields the flattest output voltage
over a wide range of current draw, as the
power supply capacitor is charged by a near
square waveform. On the other hand, the
large choke will be expensive, heavy, and (60Hz) L  Vout / Iout (mA)
given to loud mechanical noise. This circuit (50Hz) L  1.2 x Vout / Iout (mA)
requires a relatively low-valued bleeder Iout = Iac / 0.707
resistor to keep the output voltage from Vdc = Vac* / 2.2 - Vdiode - Rinductor x Idc
peaking dangerously high at turn on. *entire primary voltage
12 PS-3
Voltage Doublers
Full-Wave Voltage Doubler Although B+
often shunned by tube fanciers, there is
nothing wrong with this power supply
configuration. The full-wave voltage Rload B+/2
doubler circuit doubles the voltage that a
full-wave bridge rectifier circuit would
provide, but with only half the output
current capability. And the ripple is at the
Iout = Iac / 3.6
same 100Hz or 120 Hz frequency as in the
Vdc = (Vac x 2.8) - 2Vdiode
full-wave bridge rectifier circuit.
In this circuit, the transformer core is not partially magnetized as it is in simple half-
wave power-supply circuits, as the primary charges one capacitor during one half of
the cycle and then in the second half of the cycle, it charges the second capacitor.
Because these capacitors are in series with each other, the voltages and frequency of
the noise double.
Now, what normally happens in a full-wave-bridge power supply is that the current
flows through the secondary winding in one direction on one half of the AC cycle
and then it flows in the other direction the other half of the cycle: up, down, up,
down. But in the simple half-wave voltage rectified power supply, the current will flow
through the secondary winding in one direction on one half of the AC cycle and then
it not flow on the other half of the cycle: up, off, up, off. These repeated single
direction conduction cycles are similar in effect to hooking up a DC voltage source
across the transformer's secondary, which would definitely work to magnetize the
transformer’s core. Thus, simple half-wave rectifier circuits should only be used where
little current draw is expected, such as in establishing a negative bias voltage power
supply.
Half-Wave Voltage Doubler The
B+
principle of operation for a half-wave
C2 C1
voltage doubler, however, is different from
that of the simple half-wave rectified power Rload
supply, being much closer to the full-wave
rectifier circuit. They share the "half-wave"
name because of same noise frequency and Iout = Iac / 3.6
half cycle charging of the output capacitor, Vdc = (Vac x 2.8) - 2Vdiode
but not because of the conduction cycle C2 = C1 / 2
through the secondary.
The half-wave voltage doubler's transformer conducts through both halves of the AC
cycle just like the full-wave bridge rectifier circuit. On the first half of the cycle, the
capacitor attached to the transformer's secondary is charged and the output capacitor
receives no charging. On the second half, the both capacitors are charged
simultaneously. So the transformer's secondary experiences a current conduction on
both halves of the cycle and is thus not subject to the magnetizing of its core. In
contrast, the full-wave voltage doubler circuit, while it does experience conduction
through both halves of the cycle, works differently from the half-wave version.
 GlassWare Audio Design 13

Voltage Regulators
A voltage regulator is an electronic circuit that strives to maintain a constant output
voltage—in spite of variations in either the wall voltage or the current drawn by the
load. A voltage regulator does with complex circuitry what a car battery does by brute
force: it maintains a steady and accurate output voltage over a large range of load
current. Regulators fall into two groups: switching and linear. In terms of sales,
flexibility, conversion efficiency, and weight to power output, the switching regulator
beats the linear regulator. Yet, the switching regulator is rarely used in high-end audio
products. Why? Because of noise, high-frequency noise. Switching regulators operate at
supersonic frequencies, 40kHz and much higher; and these high frequencies can easily
escape and pollute the rest of the circuit.
Linear Regulators Linear regulators, on the other hand, are relatively inefficient; but
are also lower in noise, require fewer parts, and are usually much less complex and
finicky than a switching regulator. For example, an excellent 1A, 5-volt linear regulator
in a three-pin package that requires few supporting parts can be had for less than a
quarter. Linear voltage regulators further subdivide into series and shunt regulators.
Series Regulators The series regulator is the overwhelmingly most used voltage
regulator—for good reasons. A series regulator is like a class-B amplifier: it does not
dissipate power unless it required to. For example, if the tubes are pulled from an
amplifier, the series high-voltage regulator stops conducting until the tubes are
plugged in again. When the amplifier is drawing full current from the power supply,
the series regulator is at peak dissipation.
Shunt Regulators In contrast, the shunt regulator is like a class-A amplifier: it is
always on, it dissipates the most power at idle, and it is unburdened when power is
shifted into the load. For example, if the tubes are pulled from an amplifier, the shunt
high-voltage regulator must increase its conduction (and thus its dissipation) to
compensate for the current that the tubes had drawn. When the amplifier is drawing
full current from the power supply, the shunt regulator is at its lowest dissipation.
Shunt regulators offer many sonic advantages over series regulators; they usually find a
shorter and cleaner path to the ground reference, and they absorb noise rather than
transmit it, but at the cost of decreased efficiency. Inefficiency, however, is something
with which tube audio designers have learned to accept and some believe to be virtue:
“the less efficient, the better the sound,” as they say.
The principle of operation of a shunt regulator in AC terms is simple enough: the
noise signal on the output is compared to ground, then the current through the shunt
regulator’s active current shunting device is varied in response to that signal, and then,
in turn, this current develops a varying, inverting voltage across a series resistor or
choke that cancels this same signal. In DC terms: the power supply rail voltage is
compared to a voltage reference and the idle current through the shunt regulator is
varied to bring the power supply voltage in line with the reference. Like the series
regulator, the shunt regulator needs three elements to work: a shunting device with a
variable conductance, a voltage reference that is AC bypassed to ground, and a
feedback mechanism to control the flow of current through the shunt regulator.
Amazingly enough, the triode comprises all three elements at once.
14 PS-3

PS-3 HV Schematic & Part Values

R7
B+
D7 D8

R8 C9 R3 R4 C10

to heater AC
negative C15 C14 C13

C16 R9 C12 R6 R5 C11

D10 D9
CT
GND
J1
Chassis Ground

Chassis Ground Use jumper J1 to ground the chassis to the PCB ground.

Typical Part Values

C9, C10, C11, C12 = 100pF to 0.1µF - 1kV
C13, C14 = 150µF - 400V
C15 = 0.1 to 1µF - 100V
C16 = 0.01 to 0.1µF - 100V

D7, D8, D9, D10 = HER108 - Superfast 1A/1kV

D5, D6 = 1N4007 - 1A/1kv

R3, R4, R5, R6 = 10 ohms 1W

R7 = 100 - 10k 10W
R8 = 300k 1W
R9 = 100k 1W

Full-Wave CT Vac = 60Vac to 250Vac (requires using D7, D8, D9, D10)*
Full-Wave Bridge Vac = 120Vac to 500Vac (requires center-tap an only D7 & D8)*

*Current rating depends on load. Usually,

100mA is adequate for full-wave bridge;
50mA, full-wave CT.

With center-tapped transformer, use only D7 & D8 and R3 & R4 and C9 & C10.
With non-center-tapped transformer, use D7 to D10 and R3 to R6 and C9 to C12.

The limiting factors are the 10W RC resistor R7, the 150µF/400V power supply
capacitors (C13 & C14). The power transformer’s selection is somewhat cri ti cal , as too
high a pri mary voltage wi ll create too high a B+ vol tage. Unfortunately, most high -
voltage power transformers are truly high-voltage, offering far too much secondary
voltage. Remember that power supply's output voltage wi ll be much higher when the
power supply is unloaded; for example, when the tube-based audio circuit it powers is
still col d and warming u p. In general , high vol tage power transformer exhibit very
poor regulation, thus extra care must be taken to prevent exposing the power su pply
capacitor to excessive voltage.
 GlassWare Audio Design 15
Heater Power Supply & Regulator
The PS-3 PCB also holds a low-voltage power supply and voltage regulator. Below is
the heater power supply schematic used on the PS-3 PCB. As can be seen, the power
supply can accept many different rectifier configurations: full-wave bridge and full -
wave center-tapped and full -wave voltage doubler. In other words, this regulator can
deliver +6.3Vdc from three different power transformer secondary vol tages, 5Vac,
7Vac, or 12Vac C T; or 12.6Vdc from a 6.3Vac, 12.6Vac, or 24Vac CT primary.

When used as a full-wave-bridge or a full-wave-center-tapped recti fier circuit, the two

power supply fil tering capacitors are oriented with their positive leads pointi ng to the
top of the PCB, where the heatsi nk sits; confi gured as a voltage doubler, these
capacitors are rotated 90 degrees clockwise, so the positive leads poi nt to the right edge
of the PCB

Full-Wave Bridge AC input to pads A & C

D6
Heater+
Out In
Regulator C5
D5 C3 D3 D4 C4
Adj
R1
A C
C8
B

C2 D2 D1 C1
C7 C6
R2
CT
Heater-

Capacitor C5 & C6 positive leads pointing to heatsink

Fullwave-Bridge Rectification. Raw DC voltage = 1.414Vac - 2V
AC voltage is applied to pads A and C only.

Full-Wave Center-Tap AC input to pads A & C & CT

D6
Heater+
Out In
Regulator C5
D5 Adj C3 D3 D4 C4
R1 A C
B
C8

C2 D2 D1 C1
C7 C6
R2

CT
Heater-
Capacitor C5 & C6 positive leads pointing to heatsink
Fullwave-Bridge Rectification. Raw DC voltage = 1.414Vac - 2V
AC voltage is applied to pads A and C only.
16 PS-3
The PCB can even accept a full-wave center-tapped confi gurati on. The mai n advantage
behind using a center-tapped power transformer is that the rectificati on only loses one
rectifi er voltage drop, rather than the usual two drops. A 1V versus a 2V voltage drop
can make a decisi ve difference in a low-voltage regulated power supply, such as a 5V
regul ator.

Due to rectifier losses and the need for regulator voltage headroom, the popular
6.3Vac heater windi ng is just too low to realize a 6.3Vdc regulated outpu t voltage.
Fortunately, the PS-3 PCB can be configured to create a voltage doubler, effectively
doubling the 6.3Vac to 12.6Vac, which allows a 12Vdc or 12.6Vdc output. Note the
change in capacitors C5 & C6's position; the capacitors are no longer in paral lel , but
in series. Since only two recti fiers are used in the voltage doubler circuit, rectifi er
bypass capacitors, C1 & C 4, are al so not needed.

Voltage Doubler AC input to pads A & B

D6
Heater+
Out In
Regulator C5
D5 Adj C3 D3
R1 D4 C4
A C
B
C8

C2 D2 D1 C1
C6
C7
R2
CT
Heater-
Capacitor C5 & C6 positive leads pointing to tube rectifier
Fullwave-Volatge-Doubler Rectification. Raw DC voltage = 2.828Vac - 2V
AC voltage is applied to pads A & B only; D1, D4, C1, C4 are left off PCB.

Typical Part Values

C1, C2, C3, C4 = 0.01µF - 50V
C5, C6 = 10,000µF - 16V
C13, C14 = 1kµF - 16V
D1, D2, D3, D4 = MUR410G
R1 = 124
R2 = 499 ohm for 6.3Vdc
1.07k for 12Vdc
1.13k for 12.6Vdc

Regulator = LD1085 or LM350 or LT1085

Vac Input = 7-8Vac for 6.3Vdc (FW Bridge)

12-12.6Vac CT for 6.3Vdc (FW Bridge CT)
12-12.6Vac @ 2.5A for 12Vdc or 12.6Vdc (FW Bridge)
24-28Vac CT for 12Vdc or 12.6Vdc (FW Bridge CT)
6.3-8Vac for 12.6Vdc (Voltage Doubler)
 GlassWare Audio Design 17

Modifications & Part Upgrades

Rectifiers The recti fier offers an easy upgrade, such as the HEX FRED ultra-fast and
ultra-soft recovery recti fiers.
Resistors & Capacitors The range of prices for resistor and capacitor i s huge, wi th
resistors costi ng as little as $0.02 and as much as $25. Fortunately, the PS-3 is quite
economical i n its number of criti cal parts. Resistor R7 and capacitor C15 are the most
sonically important.
Power Transformer Most wi ll desi re usi ng a si ngl e power transformer, but using two
power transformers, one for the heater power supply and one for the high-voltage B+
power supply, holds real merit. Why two? Imagine a single power transformer; the low-
voltage solid-state recti fi ers and large-valued capacitors in the heater power su pply will
allow huge current surges that occur for i n only brief spurts, while charging its power-
supply capacitors. These large current-conduction spikes can easil y couple to the B+
wi ndi ng, as the two windings share the same core. But when you use two power
transformers, the heater power supply’s current spikes are isolated, applying to only its
separate transformer. In addition, two power transformers al low a staggered turn on,
wherein the heater power suppl y can be turned on first before appl ying the B+ voltage.
Moreover, the single power transformer that was adequate to power four 6SN7s wi th
AC on the heater may longer prove adequate when i t must del iver rectified DC, due to
the losses i nvolved in rectification and regulation.

Operation
Grounding Laying out the ground wires in an audio piece of equi pment is both a
science and an art. The best procedure is to avoi d ground loops by using a single star
ground, which may hold several ground spurs, say one for each channel in a dual-
mono setup, with two sets of hookup wire leading to the PS-3 PC B. Be sure to attach
only the ground pad (and the B+ pad) at the top of the PCB to the audio circuits
being powered, using ti ghtly twisted wi res. See page 19 for much more i nformation on
groundi ng. The chassis should also be grounded; attach a wire from the PS-3 high-
voltage CT pad to the chassis where the AC power enters the chassis.

Staggered AC With this optional setup, you can turn on the heaters fi rst, so the tubes
are gi ven a chance to heat up, whi ch will create an el ectron cloud over the cathodes,
protecti ng them from the B+ voltage. Two switches (and possibl y two power
transformers) will be needed.

Fusing Usually a fuse is placed in seri es wi th the power transformer’s primary, but
both the B+ output and the heater output can receive a fuse as well. Use a slow blow
fuse for the heater and quick blow fuse wi th the B+ output.

Caution Do not touch the heater regulator’s exposed metal tab while powered, as it
wi ll be both extremely hot and possibly charged to 1/4 the B+ vol tage; it can be a
shock hazard. In other words, test with probes before you touch with your fingers. In
additi on, give the PS-3 regulator lots of room and good air flow to wi ck away heat.
18 PS-3
Assembly & Testing
Assembly Cleanliness is essential. Before soldering, be sure to clean both sides the
PCB with 90% to 99% i sopropyl alcohol. Do not use dull-looking solder; solder
should shine. If it doesn’t, first clean away the outer oxidation with some steel wool or
a copper scouring pad. If the resistor leads look in the least gray, clean away the
oxidation wi th ei ther steel wool or a wire sniper’s sharp edges. Admittedly, with new
resistors and a fresh PCB, such metal dulling is rare; but if the parts have sat i n your
closet for a year or two, then expect a good amount of oxidation to have developed.

First, solder all the small diodes i n place, and then solder the resistors, recti fiers,
capacitors, and heatsinks. Be consistent in orienti ng the resistors; keep all the
tolerance bands on the resistor’s body at the right side as you face the resistor straight
on. This will pay divi dends later, if you need to locate a soldered a resistor in the
wrong l ocation. Because the board is double sided, wi th traces and pads on each side,
it is easier to solder the resistors from their top side. As the PCB i s so overbuilt, it is
extremely difficult to remove an incorrectly placed part. Be sure to confirm all the
electrolytic capacitor orientations, as a reversed polari zed capaci tor can easily vent (or
even explode) when presented with high-voltage. Confirm twice, solder once.

It is often easier to attach the LD1085 (low-voltage regulator) to its heatsi nk first and
then to solder both the heatsink and regulator to the PCB at once, rather than
attaching the heatsink first and then attaching the regulator to the heatsink.

Testing Wear safety eye goggles, as a venting power-supply capacitor will spray hot
caustic chemicals. Make a habit of usi ng only one hand, with the other hand behind
your back, while attachi ng probes or handling high-voltage gear, as a current flow
across your chest can result in death. In additi on, wear rubber-soled shoes and work in
dry environment. Remember, safety, first, second, and last.
1. Attach only the heater power supply’s transformer winding, leavi ng the high-
voltage transformer leads unattached and electrical tape shrouded.
2. Use a variac and slowly bring up the AC voltage, while looking for smoke or
part discoloration or bulging.
3. Measure heater regulator’s output voltage without and with a load. If the
heater regulator fails to regulate or provide enough DC voltage, try either
lowering the heater voltage a tad, for example 12V instead of 12.6V, as the
0.6V difference might be enough to bring the regulator back into regulation.
4. Next, power down the heater regulator and attach the high-voltage windings
5. Attach the power transformer to a vari ac and slowly bring up the AC voltage.
6. Measure the high-voltage output vol tage without an external l oad. The
output voltage must never exceed 400V; if the voltage is greater than 350V,
use a different transformer.
7. Once the B+ voltage has stabilized, measure the AC voltage on both sides of
Resistor R7; on the output side, the AC voltage should be substantially lower
than on the resistor’s other l ead.

Only after you are sure that both the heater regulator and B+ power supply are
working well, should you attach a line stage ampli fi er (or phono stage, etc.).
 GlassWare Audio Design 19

Grounding
Ideally, every audi o circui t finds only one central signal ground point, which the rest
of the circuit uses as i ts signal reference. Ground loops, however, are extremely easy to
introduce. For example, if the i nput and output RCA jacks are not isolated from the
chassis, then the twi sted pair of wi res that connect the PCB to the jacks will each
define a ground loop. The solution is either to isol ate the jacks or use only a single hot
wi re from jack to PCB (the wire can be shielded, as long as the shield only attaches at
one end). Thus, the best plan is to plan. Before assembling the audio circuit, stop and
decide how the grounding is going to be laid out, then sol der.

Three different schools of thought hold for groundi ng a piece of audio gear. The Old-
School approach was to treat the chassis as the ground; period. Every ground
connection is made at the closest screw and nut. This method is the easiest to follow
and it produces the worst sonic results, as steel , zinc, and alumi nu m are poor
conductors.

The Semi-Star ground method uses several ground “stars” that are often cal led spurs,
which then terminate in a si ngle star grou nd point, often a screw on the chassis. This
system can work beautifully, if carefully execu ted. Unfortunately, often too much is
included in each spur connection. For example, all the input and output RCA jacks
share ground connection to a long run of bare wire, which more cl osely resembles a
snake than a spur ground. In other words, the spurs should not be defined ju st
physical proximi ty, but signal transference. Great care must be exercised not to double
ground any spur point. For example, the volume control potentiometer can create a
ground loop problem, i f both of its ground tabs are soldered together at the
potentiometer and twisted pairs, of hot and cold wires, arrive at and leave the
potentiometer, as the two cold wires attaching to the PCB will defi ne a ground loop.

The Absolute-Star grounding scheme uses a l ot of wire and i s the most time
consu ming to lay out, but it does yi eld the best sonic rewards. Here each inpu t signal
source and each ou tput lead gets its own ground wire that attaches, ultimately, at one
star ground point; each RCA jack is isolated from the chassis.

House Ground The third prong on the wall outl et attaches to the house’s ground,
usual ly the cold water pipe or a copper-plated steel pipe buried under the house. The
audio circuit's chassi s can (and should) also attach to this ground connecti on, which is
certainly the safest approach, as i t provides a discharge path should the B+short to the
chassis. Unfortunatel y, this setup often produces a hu m problem. Some simply fl oat
the ground, whi ch is not safe, while others use several solid-state rectifiers in parallel to
attach the chassis ground to the house ground (NOT NEUTRAL) via the thi rd prong,
and others still use a power 10-ohm resistor shunted by a small capacitor.

House Chassis Signal 10

0.01µF
Ground Ground Ground 250V
20 PS-3

PS-3 PCB (4" by 4")

Layout & Dimensions

4.0 in

4.0 in
3.6 in

3.6 in

Let me know what you think

If you woul d like to see some new audio PCB or kit or recommend a change to an
existing product or if you need help figuring out the PS-3's assembly, drop me a line
by e-mail to sales@glass-ware.com (begin the subject li ne with either “Ai kido” or
“tube” or the spam filters are sure to eat your message).