Digital to Analogue Conversion
Analogue Digital Conversion
Analog and digital data were briefly mentioned at
the start
A digital signal is an approximation of an analog
one
Levels of signal are sampled and converted to a
discrete bit pattern.
Digital signal processing is used, for example, to
enhance and compress images, to process
sounds to generate speech, etc, etc.
Step (discrete) Approximation
“stair-step”
approximation of
sample original signal
level
more samples give greater accuracy
time
hold time for sample
Objectives
To understand how a digital value can be converted to an
analogue value
To draw circuits and explain the operation of two digital to
analogue converters: the binary weighted resistor network and the
R-2R ladder network
To draw the block diagram and explain the operation of three
analogue to digital converters: flash, counter ramp and
successive approximation
To be able to calculate the conversion time for an analogue to
digital converter
To be able to explain the sampling rule
To be able to describe the basic design of a sample and hold
circuit and explain how it works
The binary weighted resistor network
Comprises of a register and resistor network
Output of each bit of the register will depend on whether a 1
or a 0 is stored in that position
e.g. for a 0 then 0V output
for a 1 then 5V output
Resistance R is inversely proportional to binary weight of each
digit
R
MSB
2R RL VL
4-bit
4R
register
8R
LSB
Buffering the resistor network
Best solution is to follow the resistor network with a buffer
amplifier
Has high impedance, practically no current flows
All input currents sum at S and go through Rf
Vo = -IfRf
R I1 Rf
MSB
4-bit register
2R I2 If
4R -
I3 S I
LSB 8R I4 +
Vo
Vo If R f (I1 I2 I3 I 4 ) Rf
Digital-to-Analogue Example
Calculate the output voltage for an input code
word 0110 if a logic 1 is 10V and a logic 0 is 0V,
and R = RF=1k
I1 = I4 = 0
I2 = 10v / 2R = 10 / 2k = 5 mA
I3 = 10v / 4R = 10 / 4k = 0.25 mA
Vo = -If x Rf = -(0.0075) x 1000 = -7.5 volts
Vo If Rf (I1 I2 I3 I4) Rf
The binary weighted resistor network
Seldom used when more than 6 bits in the code
word
to illustrate the problem consider the design of an
8-bit DAC if the smallest resistor has resistance R
what would be the value of the largest resistor?
what would be the tolerance of the smallest resistor?
Very difficult to manufacture very accurate
resistors over this range
The R-2R Ladder Resistor Network
Has a resistor network which requires resistance values that
differ 2:1 for any sized code word
The principle of the network is based on Kirchhoff's current
rule
The current entering N must leave by way of the two resistors
R1 and R2
I N R2
R1
•
The R-2R Ladder Resistor Network
Works on a current dividing network
A B
I R I1
2R 2R 2R
I1 I2
Resistance to right of B = 1/(1/2R + 1/2R)
Resistance to right of A = R +2R/2 = 2R
Current divides I1 = I/2 I2 = I/4 divides again
The R-2R Ladder Resistor Network
The network of resistors to the right of A have an
equivalent resistance of 2R, and so the right hand
resistance can be replaced by a copy of the network
I I/2 R I/4 R I/8
Bit Current
3 I/2
2 I/4
2R 2R 2R 2R
1 I/8
0 I/16
I1=I/2 I2=I/4 I3=I/8
bit 3 bit 2 bit 1 bit 0
The R-2R Ladder Resistor Network
I R R R The state of the bits is
used to switch a
voltage source
Vs 2R 2R 2R 2R 2R
I/2 I/4 I/8 I/16
Rf
+
Vo
4-bit register
MSB
LSB
Vo -R f (b 3 I 2 b2 I 4 b I 8 b 0 I 16)
1
Example
I R R R
Vs 2R 2R 2R 2R 2R
I/2 I/4 I/8 I/16
Rf
+
Vo
4-bit register
MSB
LSB
Vo -R f (b 3 I 2 b2 I 4 b I 8 b 0 I 16)
1
For the circuit shown above with I = 10 mA and Rf = 2k, calculate the
output voltage V0 for an input code word 1110.
Example
I = 10mA
Rf = 2k
input code word 1110
Vo = -2000( 0.01/2 + 0.01/4 + 0.01/8 + (0 x 0.1)/8 )
= - 2000 * (0.04 + 0.02 + 0.01) / 8
= 17.5 volts
Quantisation
Suppose we want to use a D-A converter to generate the
sawtooth waveform (graph shown on the left)
End up with stair-case waveform (graph shown on the right)
The 16 possible values of the D-A converter output are called
the quantisation levels
The difference between two adjacent quantisation levels is
termed a quantisation interval
voltage
voltage
time
0000
0000
0001
0010
time
1111
...
Quantisation Error
Difference between the two waveforms is the quantisation
error
Maximum quantisation error is equal to half the
quantisation interval
One way to reduce the quantisation error (noise) is to
increase the number of bits used by the D-A converter
quantisation interval
111
110
bands or quanta
1001
101 1000
0111
100
0110
011 0101
0100
010 0011
0010
001 0001
0000
000
samples
Quantisation Noise
The voltage produced by the DA convertor can be
regarded as the original signal plus noise:
This is the quantisation noise.
Summary
We have looked at techniques for converting a digital
codeword into an analogue voltage using a weighted resistor
network. In particular:
the binary weighted network (not suitable for large resolution D-A
converters)
the R-2R ladder
The addition of an amplifier minimises the loading effects on
the weighted network
The conversion from digital to analogue involves a quantisation
process that limits the resolution and introduces the
quantisation noise.
This quantisation error can be reduced by increasing the
number of bits in the converter.
