Table of Contents
Advertisement
9 8640A Analog Input Interface
Detailed Calibration
A detailed calibration compensates for the offset introduced by the programmable gain amplifier
(PGA) and an offset effect caused by the ADC IC. The sign (polarity) of the PGA offset can be detec-
ted, but the ADC's cannot. Only the magnitude of the ADC's offset can be detected (see the theory of
operation for an expalanation of the polarity bit's origin).
To make the calibration you will need to:
1. Measure the offset caused by the ADC.
2. Measure the offset due to the PGA.
3. Apply the offsets to a reading from an input channel.
Conditions
The procedure is described below, but the sequence shown is not the only one possible. The procedure
assumes you have converted the signed magnitude to a real number. While you may perform bit
manipulations on the 13 bits of the raw reading, remember it is in signed binary (not two's compli-
ment) form and that zero may have positive or negative sign (polarity).
The polarity of zero is caused by offsets which are too small to measure but not too small to detect.
For the mathematical operations described in the calibration procedures, normalize zero by removing
the sign.
NOTE
Most high level languages do not permit signed zero
(±O)
in a numeric variable. You will have decide how to store
it.
1. ADC offset
A. Read the reference channel with the gain set to 1. With the gain set
to
1
you will minimize the affects of most offset sources except the
ADC chip.
B. Ignore the sign bit; this offset is always positive. This value is
ADC_Offset. The polarity bit you detected when you took the reading
was caused by whatever offset existed at the PGA.
2. PGA offset
A. Read the reference channel with the gain set to S12. This reading
will detect the combination of the PGA_ Offset_X_ S12 (PGA offset
multiplied by S12) and the ADC _Offset.
B. Subtract ADC_Offset from the absolute value of the reading. The
difference is the IPGA_Offset_X_SI21.
4-4
Advertisement
Table of Contents
