MCP6271R Microchip Technology Inc., MCP6271R Datasheet - Page 29

• Manufacturer Part Number: MCP6271R
• Description: 170 ?a, 2 Mhz Rail-to-rail Op Amp
• Manufacturer: Microchip Technology Inc.

Using The Basics For ADC Range Control
By Bonnie C. Baker, Microchip Technology Inc.
You can use a high-resolution ADC to simplify or eliminate most
of the sensitive analog front-end circuitry in your system. I know
that this sounds too good to be true, but consider the pressure-
sensing example of a single-supply signal path that measures
pressure from the real world to a controller. The electronic
portion of this signal path starts at the pressure sensor. The
signal then travels through an instrumentation amplifier (the gain
stage with a voltage reference) into a 5 th order analog low-pass
filter (i.e., the two op amp noise reduction stage), digitized by
a 10- or 12-bit ADC and finally into the microcontroller. In the
microcontroller or processor, you finally implement code that
calibrates the data and reduces error. Wow, can you say all of
that in one breath? At minimum, this signal path requires seven
active devices.
This is a long-winded approach, especially when you consider a
high-resolution converter, such as a 24-bit sigma-delta converter,
as an alternative. With the sigma-delta converter, the only
external circuitry before the converter is the pressure sensor
and a few 1 st order, R|C filters. This is a great strategy because
you can reduce noise and layout errors with the sigma-delta
converter. You can also calibrate system-offset and gain errors
while getting your 12-bit range back in the controller or processor.
Normally, you would adjust the 24-bit converter’s offset error at
the code transition between 000h and 001h. Since converters
have some degree of transition noise, you need to sample this
transition area several times to verify the transition voltage. You
can then quickly remove the offset error of the converter using
the following formula:
This formula is a good start. We can translate it into a usable
form to operate at any point in the ADC transfer function (also
see Figure 1).
Analog-to-Digital Converters
Offset error = (V[0:1] – 0.5 (V
V
V[0:1] = analog voltage of first transition
V
n = number of converter bits
Offset error (with level shift) = (V[x:(x+1)] – 0.5 V
“x” is the code that is produced by your tested or selected
offset value.
ILSB
REF
= full-scale voltage
= V
REF
/ 2 n = ideal LSB voltage size
ILSB
) ) / V
ILSB
ILSB
) / V
ILSB
With this new offset-error formula, you can inject a level shift
into the output data with the controller. From this point, you can
calculate the gain error for the region of interest. The formula
that calculates the gain error across the full-scale range of an
ADC is:
If you translate this gain error formula to match the range of
interest, it becomes:
You can apply the offset and gain equations above to the output
data of any ADC, regardless of the converter’s resolution. High-
resolution ADCs can eliminate amplifier gain stages, high-order
filter stages and level-shift circuitry by taking advantage of a
portion of the million-plus possible output codes. With the offset
and gain formulas above, you can use the controller or processor
computation power to center on the portion of the conversion
that is of interest for your application.
Figure 1: You can isolate the usable output-code range of an ADC
by using modified offset and gain error equations.
Gain error = (V
V
Gain error = (V
V
Where capital “N” is equal to the number of bits that you are
going to use in your system.
ILSB
ILSB
Output
Offset error (with level shift) = V [x:(x+1)] - 0.5 V
Gain error = (V
Digital
Code
Analog and Interface Guide – Volume 2
0
Your new full-scale range
REF
REF
REF
– 2 V
– 2 V
1/4 FS
Analog Input Voltage
- 2V
ILSB
ILSB
ILSB
1/2 FS
Your new offset voltage
- V[(2
– V[(2
– V[(2
N
-2):(2
3/4 FS
n
N
N
–2):(2
–2):(2
-1)] - V[0:1]) / V
FS
ILSB
n
N
– 1)] – V[0:1] ) /
) / V
– 1)] – V[0:1] ) /
ILSB
ILSB
27