Datasheet

SYSTEM CALIBRATION USING THE PGA
TransferFunction
withOffsetError+GainError
TransferFunction
withGainErrorOnly
IdealTransferFunction
GainError
V
FS_ACTUAL
V
FS_IDEAL
AnalogInput
DigitalOutput
V 1LSB-
REF_ADC
0V
0000h
0FFFh
OffsetError
V
Z_IDEAL
V
Z_ACTUAL
PGA112 , , PGA113
PGA116 , PGA117
SBOS424B – MARCH 2008 – REVISED SEPTEMBER 2008 ............................................................................................................................................ www.ti.com
In practice, the zero input (0V) or full-scale input
(V
REF_ADC
– 1LSB) of ADCs cannot always be
Analog-to-digital converters (ADCs) contain two major
measured because of internal offset error and gain
errors that can be easily removed by calibration at a
error. However, if measurements are made very close
system level. These errors are gain error and offset
to the full-scale input and the zero input, both zero
error, as shown in Figure 75 . Figure 75 shows a
and full-scale can be calibrated very accurately with
typical transfer function for a 12-bit ADC. The analog
the assumption of linearity from the calibration points
input is on the x-axis with a range from 0V to
to the desired end points of the ADC ideal transfer
(V
REF_ADC
– 1LSB), where V
REF_ADC
is the ADC
function. For the zero calibration, choose
reference voltage. The y-axis is the hexadecimal
10%V
REF_ADC
; this value should be above the internal
equivalent of the digital codes that result from ADC
offset error and sufficiently out of the noise floor
conversions. The dotted red line represents an ideal
range of the ADC. For the gain calibration, choose
transfer function with 0000h representing 0V analog
90%V
REF_ADC
; this value should be less than the
input and 0FFFh representing an analog input of
internal gain error and sufficiently below the tolerance
(V
REF_ADC
– 1LSB). The solid blue line illustrates the
of V
REF
. These key points can be summarized in this
offset error. Although the solid blue line includes both
way:
offset error and gain error, at an analog input of 0V
the offset error voltage, V
Z_ACTUAL
, can be measured.
For zero calibration:
The dashed black line represents the transfer function
• The ADC cannot read the ideal zero because of
with gain error. The dashed black line is equivalent to
offset error
the solid blue line without the offset error, and can be
• Must be far enough above ground to be above
measured and computed using V
Z_ACTUAL
and
noise floor and ADC offset error
V
Z_IDEAL
. The difference between the dashed black
• Therefore, choose 10%V
REF_ADC
for zero
line and the dotted red line is the gain error. Gain and
calibration
offset error can be computed by taking zero input and
full-scale input readings. Using these error
For gain calibration:
calculations, compute a calibrated ADC reading to
• The ADC cannot read the ideal full-scale because
remove the ADC gain and offset error.
of gain error
• Must be far enough below full-scale to be below
the V
REF
tolerance and ADC gain error
• Therefore, choose 90%V
REF_ADC
for gain
calibration
Figure 75. ADC Offset and Gain Error
34 Submit Documentation Feedback Copyright © 2008, Texas Instruments Incorporated
Product Folder Link(s): PGA112 PGA113 PGA116 PGA117