Transcript Analog to Digital Conversions

Analogue to Digital Conversion
By Cap’n Tim Johnson, PE
Prepared 11/12/2013
Digital World
• The digital in A/D conversion consists of binary or
hex numbers that will come to stand for a real
world phenomenon.
• Say, for instance we have four-bits we use to
represent an analog value that’s been converted.
• In binary, the numbers would range from 0000 to
1111 and there are (24 = 16) sixteen different
values represented.
• These number in base 10 could range from 0-15
Typical example of converting a binary
number to a decimal number
While this is very
useful in converting
binary number to
decimal and back…
analog to digital
conversion is similar
but different.
A/D is converting a binary number
into something else
Something
Else
We want to change
the decimal number
column into real
world values that
don’t necessarily
increment by whole
numbers but do
increment by binary
numbers.
This example shows the conversion of
real world values to binary numbers
Real World
Values
Here if we let each
binary bit value or
weight represent a
1/3 of a volt, as the
binary numbers
increment, the
voltage goes up 1/3 a
volt at a time.
0.00
0.33
0.66
1.00
1.33
1.66
2.00
2.33
2.66
3.00
3.33
3.66
How do we get the real world values?
1. We need to know the maximum input
voltage or the range between Vmax – Vmin. If
Vmin is 0 or ground then all we need is Vmax.
2. We need to know the number of bits used in
the A/D converter.
3. Vmax/(2#bits) = Resolution
4. Resolution is the value of the incremental
voltage change per bit or the difference
between 0001 and 0000.
What else does the table tell us?
The binary numbers
are stored after a
conversion in the A/D
results register. By
using a simple formula
the binary number can
tell us what the input
voltage is.
Real World
Values
0.00
0.33
0.66
1.00
1.33
1.66
2.00
2.33
2.66
3.00
3.33
3.66
How?
Convert the binary
number in the A/D
results register to its
decimal equivalent.
That tells us how many
bits are needed to
multiply the resolution
to equal the input
voltage.
𝑅𝑒𝑠𝑜𝑙𝑢𝑡𝑖𝑜𝑛 ∗ (𝐴/𝐷 𝑣alue10)
101121110
⅓ * 11 = 3 ⅔ volts
Real World
Values
0.00
0.33
0.66
1.00
1.33
1.66
2.00
2.33
2.66
3.00
3.33
3.66
Quick Summary
• You need to know the maximum voltage that will
be present on the input. This voltage can not
exceed the device’s upper limit.
• You need to know the number of bits the device
uses to report the value in the A/D results
register
• Calculate the Resolution: Vmax/(2#bits)
• Convert the value in the A/D results register to
base 10 to determine the decimal number of bits
• Multiply the decimal number times the
Resolution to determine the input voltage.
General Rule
• Your results will never be higher than Vmax
• For instance, if the results register reads $FF,
all the bits are filled in a 8-bit ADC.
• The complete formula is:
•
𝑉𝑚𝑎𝑥
2#𝑏𝑖𝑡𝑠
∗ 255 
𝑉𝑚𝑎𝑥
256
∗ 255  255
256
∗ 𝑉𝑚𝑎𝑥
• The fraction 255/265 ≈ 1 thus you have 1*Vmax
• The error in the fraction = Quantization error
Other Rules of Thumb
• If you get $80 (1000 0000) in the result register,
the answer is ½ Vmax ($80/$FF is ½)
• If the last four zeros have some value then 1000
XXXX is approximately ½ Vmax
• You could read just the first 2 bits and round off
to fourths of Vmax Ex.: 0100 XXXX ≈ ¼ Vmax
• If you read just the first 3 bits you can round off
to eights of Vmax
• Ex.: 0110 XXXX ≈ ⅜Vmax & 1110 XXXX ≈ ⅞Vmax
What else could we do with A/D?
• First we have to think of what it is that we want
digitized. Rotation around an axis is a useful number
especially if we were rotating a telescope horizontally.
• We’d need to digitize the number of degrees we
rotated a telescope. The maximum number of degrees
is 360°. Then suppose our A/D converters was limited
to 5 volts input maximum and 8-bits…that would be
~1.4 degrees/bit.
• But if we limited our rotation to 180° that would be ~.7
degrees per bit. Use the ADC we can determine if the
telescope turned 1 ° with each change of a bit.
Another look at ADC
LSB value
This graph shows the fitting of a linear voltage to an ADC in theory. Visible is the
quantization error which is the overhang of the digital step higher and/or lower than
the ideal value. The ideal quantization error is ± ½ LSB. The LSB is aka Resolution.
Role of ADC module in microcontrollers
The ADC module is one of a number of modules on a microcontroller. Others
include General Purpose I/O, Comparators, Digital to Analog Converters , Op
Amps, Timers, Real-Time clock, AES encryption, Brown-Out Reset, Direct Memory
Access, Memory Protection Units, Supply Voltage Supervisors, Capacitive Touch
Sense I/O, Communications (UART,SPI, I2C, IrDA, USB, Ethernet), LCD, and other
specialty designs.
This is the block diagram for the TI MSP430 Sigma-Delta 16 bit ADC
A/D Results Register