There are a few options to consider
when implementing calibration in production depending on the needed accuracy and
production test equipment capability and test time. Let's take cell voltage
calibration as an example. Below are a few different options listed ranging from
least accurate (1) to most accurate (4).
- No production calibration, use default values.
- Perform calibration on a fixed number of devices and use the
average values of the calibration parameters for all packs. This might be a good
option if the variation observed from board to board is small and
acceptable.
- Calibrate each device individually in production by applying a
single accurate voltage applied to the cell input, such as 3.5 V. Averaging the
measurement over several readings can further improve accuracy.
- Calibrate each device individually in production by applying
two precise DC voltages to the cell input, such as 2.5 V and 4.5 V. Averaging
each of the measurements over several readings can further improve accuracy.
This will not only allow a more accurate calculation of cell gain, but also
enables calibration of cell voltage offset.