Physics 120B: Lecture 8

Odds and Ends

Binary/Hex/ASCII
Memory & Pointers in C
Decibels & dB Scales
Coherent Detection
 Binary, Hexadecimal Numbers
• Computers store information in binary
– 1 or 0, corresponding to VCC and 0 volts, typically
– the CC subscript originates from “collector” of transistor
• Become familiar with binary counting sequence
binary decimal hexadecimal
0000 0000 0 0x00
0000 0001 1 0x01
0000 0010 2 0x02
0000 0011 2+1 = 3 0x03
0000 0100 4 0x04
0000 0101 4+1 = 5 0x05
etc.
1111 1100 128+64+32+16+8+4 = 252 0xfc
1111 1101 128+64+32+16+8+4+1 = 253 0xfd
1111 1110 128+64+32+16+8+4+2 = 254 0xfe
1111 1111 128+64+32+16+8+4+2+1 = 255 0xff
Lecture 7 2
 Binary to Hex: easy!
• Note separation of previous 8-bit (one-byte)
numbers into two 4-bit pieces (nibbles)
– makes expression in hex (base-16; 4-bits) natural
binary hexadecimal decimal

0000 0 0

0001 1 1

0010 2 2

0011 3 3

0100 4 4

0101 5 5

0110 6 6

0111 7 7

1000 8 8

1001 9 9

1010 A (lower case fine) 10

1011 B 11
Lecture 7 3
 ASCII Table in Hex first hex digit
0 1 2 3 4 5 6 7
0 NUL ^@ null (\0) DLE ^P SP space 0 @ P ` p
1 SOH ^A start of hdr DC1 ^Q ! 1 A Q a q
2 STX ^B start text DC2 ^R “ 2 B R b r
3 ETX ^C end text DC3 ^S # 3 C S c s
4 EOT ^D end trans DC4 ^T $ 4 D T d t
5 ENQ ^E NAK ^U % 5 E U e u
second hex digit

6 ACK ^F acknowledge SYN ^V & 6 F V f v

7 BEL ^G bell ETB ^W ‘ 7 G W g w
8 BS ^H backspace CAN ^X ( 8 H X h x
9 HT ^I horiz. tab (\t) EM ^Y ) 9 I Y i y
A LF ^J linefeed (\r) SUB ^Z * : J Z j z
B VT ^K vertical tab ESC escape + ; K [ k {
C FF ^L form feed FS , < L \ l |
D CR ^M carriage ret (\n) GS - = M ] m }
E SO ^N RS . > N ^ n ~
F SI ^O US / ? O _ o DE
Lecture 7 L 4
 ASCII in Hex
• Note the patterns and conveniences in the ASCII
table
– 0 thru 9 is hex 0x30 to 0x39 (just add 0x30)
– A-Z parallels a-z; just add 0x20
• starts at 0x41 and 0x61, so H is 8th letter, is 0x48, etc.
– the first 32 characters are control characters, often
represented as Ctrl-C, denoted ^C, for instance
• associated control characters mirror 0x40 to 0x5F
• put common control characters in red; useful to know in some
primitive environments

Lecture 7 5
 Two’s Complement
• Unsigned are direct binary representation
• Signed integers usually follow “two’s complement”
binary hex unsigned 2’s complement
0000 0000 0x00 0 0
0000 0001 0x01 1 1
0000 0010 0x02 2 2
0111 1111 0x7F 127 127
1000 0000 0x80 128 -128
1000 0001 0x81 129 -127
1111 1110 0xFE 254 -2
1111 1111 0xFF 255 -1

– rule: to get neg. number, flip all bits and add one
• example: -2: 0000 0010  1111 1101 + 1 = 1111 1110
– adding pos. & neg.  0000 0000 (ignore overflow bit)
Lecture 7 6
 Floating Point Numbers
• Most standard is IEEE format
– http://en.wikipedia.org/wiki/IEEE_754-1985

1.01×2−3, 1 is implied, exp. offset by 127

• Three parts: sign, exponent, mantissa

– single-precision (float) has 32 bits (1, 8, 23, resp.)
• 7 digits; 10±38; log(10)/log(2) = 3.32, so 223 ≈ 107; ±127/3.32 ≈ 38
– double precision (double) has 64 bits (1, 11, 52, resp.)
• 16 digits; 10±308
• The actual convention is not critical for us to understand, as
much as:
– limitations to finite representation
– space allocation in memory: just 32 or 64 bits of 1’s & 0’s
Lecture 7 7
 Arrays & Storage in C

• We can hold more than just one value in a variable

– but the program needs to know how many places to save in
memory
• Examples:
int i[8], j[8]={0}, k[]={9,8,6,5,4,3,2,1,0};
double x[10], y[10000]={0.0}, z[2]={1.0,3.0};
char name[20], state[]=“California”;
– we can either say how many elements to allow and leave them
unset; say how many elements and initialize all elements to zero;
leave out the number of elements and specify explicitly; specify
number of elements and contents
– character arrays are strings
– strings must end in ‘\0’ to signal the end
– must allow room: char name[4]=“Bob”
• fourth element is ‘\0’ by default

Lecture 7 8
 Indexing Arrays
int i,j[8]={0},k[]={2,4,6,8,1,3,5,7};
double x[8]={0.0},y[2]={1.0,3.0},z[8];
char name[20],state[]="California";

for (i=0; i<8; i++)

{
z[i] = 0.0;
printf(”j[%d] = %d, k[%d] = %d\n",i,j[i],i,k[i]);
}
name[0]='T';
name[1]='o';
name[2]='m';
name[3] = '\0';
printf("%s starts with %c and lives in %s\n",name,name[0],state);

• Index array integers, starting with zero

• Sometimes initialize in loop (z[] above)
• String assignment awkward outside of declaration line
– #include <string.h> provides “useful” string routines
• done automatically in Arduino, but also String type makes many things easier

Lecture 7 9
 Memory Allocation in Arrays
• state[]=“California”; 
each block is 8-bit char
C a l i f o r n i a \0

• name[11]=“Bob”; 

B o b \0

– empty spaces at the end could contain any random garbage

• int i[] = {9,8,7,6,5,4,3,2}; 
9 8 7 6 5 4 3 2
each block is 16 or 32-bit int
– indexing i[8] is out of bounds, and will either cause a
segmentation fault (if writing), or return garbage (if reading)
Lecture 7 10
 Multi-Dimensional Arrays
j
0 1 2 3
int i,j,arr[2][4];
0 4 5 6 7
for (i=0; i<2; i++){ i
for (j=0; j<4; j++){ 1 2 3 4 5
arr[i][j] = 4+j-2*i;
}
} in memory space: 4 5 6 7 2 3 4 5

• C is a row-major language: the first index describes

which row (not column), and arranged in memory
row-by-row
– memory is, after all, strictly one-dimensional
• Have the option of treating a 2-D array as 1-D
– arr[5] = arr[1][1] = 3
• Can have arrays of 2, 3, 4, … dimensions
Lecture 7 11
 Arrays and functions
• How to pass arrays into and out of functions?
• An array in C is actually handled as a “pointer”
– a pointer is a direction to a place in memory
• A pointer to a variable’s address is given by the & symbol
– you may remember this from scanf functions
• For an array, the name is already an address
– because it’s a block of memory, the name by itself doesn’t contain a
unique value
– instead, the name returns the address of the first element
– if we have int arr[i][j];
• arr and &arr[0] and &arr[0][0] mean the same thing: the address of the first
element
• By passing an address to a function, it can manipulate the
contents of memory directly, without having to pass bulky
objects back and forth explicitly

Lecture 7 12
Example: 3x3 matrix multiplication
void mm3x3(double a[], double b[], double c[])

// Takes two 3x3 matrix pointers, a, b, stored in 1-d arrays nine

// elements long (row major, such that elements 0,1,2 go across a
// row, and 0,3,6 go down a column), and multiplies a*b = c.
{

double *cptr; // pointer type variable: * gets at contents

int i,j;

cptr = c; // without *, it’s address; point to addr. for c

for (i=0; i<3; i++){

for (j=0; j<3; j++){
*cptr++ = a[3*i]*b[j] + a[3*i+1]*b[j+3] + a[3*i+2]*b[j+6];
// calc value to stick in current cptr location, then
// increment the value for cptr to point to next element
}
}
}

Lecture 7 13
 mm3x3, expanded
• The function is basically doing the following:

*cptr++ = a[0]*b[0] + a[1]*b[3] + a[2]*b[6];

*cptr++ = a[0]*b[1] + a[1]*b[4] + a[2]*b[7];
*cptr++ = a[0]*b[2] + a[1]*b[5] + a[2]*b[8];

*cptr++ = a[3]*b[0] + a[4]*b[3] + a[5]*b[6];

*cptr++ = a[3]*b[1] + a[4]*b[4] + a[5]*b[7];
*cptr++ = a[3]*b[2] + a[4]*b[5] + a[5]*b[8];

*cptr++ = a[6]*b[0] + a[7]*b[3] + a[8]*b[6];

*cptr++ = a[6]*b[1] + a[7]*b[4] + a[8]*b[7];
*cptr++ = a[6]*b[2] + a[7]*b[5] + a[8]*b[8];

– which you could confirm is the proper set of operations for

multiplying out 3×3 matrices

Lecture 7 14
 Notes on mm3x3
• The function is constructed to deal with 1-d instead
of 2-d arrays
– 9 elements instead of 3×3
– it could have been done either way
• There is a pointer, *cptr being used
– by specifying cptr as a double pointer, and assigning its
address (just cptr) to c, we can stock the memory by using
“pointer math”
– cptr is the address; *cptr is the value at that address
– just like &x_val is an address, while x_val contains the value
– cptr++ bumps the address by the amount appropriate to
that particular data type, called “pointer math”
– *cptr++ = value; assigns value to *cptr, then advances the cptr
count
Lecture 7 15
 Using mm3x3
#include <stdio.h>

void mm3x3(double a[], double b[], double c[]);

int main()
{
double a[]={1.0, 2.0, 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, 9.0};
double b[]={1.0, 2.0, 3.0, 4.0, 5.0, 4.0, 3.0, 2.0, 1.0};
double c[9];

mm3x3(a,b,c);

printf("c = %f %f %f\n",c[0],c[1],c[2]);
printf(" %f %f %f\n",c[3],c[4],c[5]);
printf(" %f %f %f\n",c[6],c[7],c[8]);

return 0;
}
• passing just the names (addresses) of the arrays
– filling out a and b, but just making space for c
– note function declaration before main

Lecture 7 16
 Another way to skin the cat
double a[3][3]={{1.0, 2.0, 3.0},
{4.0, 5.0, 6.0},
{7.0, 8.0, 9.0}};
double b[3][3]={{1.0, 2.0, 3.0},
{4.0, 5.0, 4.0},
{3.0, 2.0, 1.0}};
double c[3][3];

mm3x3(a,b,c);

• Here, we define the arrays as 2-d, knowing that in memory

they will still be 1-d
– we will get compiler warnings, but the thing will still work
– not a recommended approach, just presented here for
educational purposes
– Note that we could replace a with &a[0][0] in the function call,
and the same for the others, and get no compiler errors

Lecture 7 17
 Decibels
• Sound is measured in decibels, or dB
– as are many radio-frequency (RF) applications
• Logarithmic scale
– common feature is that every 10 dB is a factor of 10 in
power/intensity
– other handy metrics
• 3 dB is 2×
• 7 dB is 5×
• obviously piling 2× and 5× is 10×, which is 10 dB = 3 dB + 7 dB
– decibels thus combine like logarithms: addition represents
multiplicative factors

Lecture 7 18
 Sound Intensity
• Sound requires energy (pushing atoms/molecules through a
distance), and therefore a power
• Sound is characterized in decibels (dB), according to:
– sound level = 10log(I/I0) = 20log(P/P0) dB
– I0 = 1012 W/m2 is the threshold power intensity (0 dB)
– P0 = 2105 N/m2 is the threshold pressure (0 dB)
• atmospheric pressure is about 105 N/m2
• 20 out front accounts for intensity going like P2
• Examples:
– 60 dB (conversation) means log(I/I0) = 6, so I = 106 W/m2
• and log(P/P0) = 3, so P = 2102 N/m2 = 0.0000002 atmosphere!!
– 120 dB (pain threshold) means log (I/I0) = 12, so I = 1 W/m2
• and log(P/P0) = 6, so P = 20 N/m2 = 0.0002 atmosphere
– 10 dB (barely detectable) means log(I/I0) = 1, so I = 1011 W/m2
• and log(P/P0) = 0.5, so P  6105 N/m2

Lecture 7 19
 Sound hitting your eardrum
• Pressure variations displace membrane (eardrum,
microphone) which can be used to measure sound
– my speaking voice is moving your eardrum by a mere
1.510-4 mm = 150 nm = 1/4 wavelength of visible light!
– threshold of hearing detects 510-8 mm motion, one-half
the diameter of a single atom!!!
– pain threshold corresponds to 0.05 mm displacement
• Ear ignores changes slower than 20 Hz
– so though pressure changes even as you climb stairs, it is
too slow to perceive as sound
• Eardrum can’t be wiggled faster than about 20 kHz
– just like trying to wiggle resonant system too fast produces
no significant motion
Lecture 7 20
 dB Scales
• In the radio-frequency (RF) world, dB is used several
ways
– dB is a relative scale: a ratio: often characterizing a gain or
loss
• +3 dB means a factor of two more
• −17 dB means a factor of 50 loss, or 2% throughput
– dBm is an absolute scale, in milliwatts: 10×log(P/1 mW)
• a 23 dBm signal is 200 mW
• 36 dBm is 4 W (note 6 dB is two 3 dB, each a factor of 2  4×)
• −27 dBm is 2 mW
– dBc is signal strength relative to the carrier
• often characterizes distortion from sinusoid
• −85 dBc means any distortions are almost nine orders-of-
magnitude weaker than the main sinusoidal “carrier”
Lecture 7 21
 Coherent Detection
• Sometimes fighting to discern signal against
background noise
– photogate in bright setting, for instance
• One approach is coherent detection
– modulate signal at known phase, in ON/OFF pattern at
50% duty cycle
– accumulate (add) in-phase parts, while subtracting out-of-
phase parts
– have integrator perform accumulation, or try in software
• but if background is noisy in addition to high, integration better
– basically background subtraction
– gain more the greater the number of cycles integrated

Lecture 6 22
Raw Signal, Background, and Noise

Lecture 6 23
Modulated Signal; still hard to discern

Lecture 6 24
Integration, subtracting “OFF” portions

Lecture 6 25
 Expressed in Electronics

first op-amp just inverting; second sums two inputs, only one on at a time
has effect of adding parts when Ref = +1, subtracting where Ref = -1
clears “memory” on timescale of tint = RintC
could also conceive of performing math in software
Lecture 6 26
 Announcements
• Project Proposals due next Friday, Feb 7
• Lab 4 due following Tu/Wed (2/11, 2/12)

Lecture 7 27