Repository navigation
Script needed: Check the HK bias error #12
Description
This script should obtain HK data from those inputs which do not have any signal connected. Thus the values obtained will provide calibration information on the bias error associated with all the HK channels.
It should be determined what, if anything, should be done with this calibration information. Tests on 6.1 hardware indicate the bias error will be vanishingly small, and so this test would become simply verification that any specific board is performing similarly.
Activity
ebokhour commentedon Dec 30, 2015
A functioning version of this exists as an iPython notebook at
https://github.com/TESScience/FPE_Test_Procedures/blob/master/John%20Doty's%20Global%20Calibration%20of%20the%20Housekeeping%20Data%20Collection.ipynb
xcthulhu commentedon Jan 25, 2016
This is done in the following script: https://github.com/TESScience/FPE/blob/master/scripts/bias_test
ebokhour commentedon Jan 28, 2016
The bias_test only outputs the power supply readings. This is a very useful subset of values, but it's not what the bias test is supposed to report. It should report the bias offsets for the unused channels, to indicate the analog bias error. It may be useful to make the power supply readings a separate script (e.g. rename this script), but that's a discussion for another thread.
xcthulhu commentedon Jan 28, 2016
How do I know if a channel is unused?
... looks like it will have the word "bias" in the name.
The expected value of these is always supposed to be 0 μA?
ebokhour commentedon Jan 28, 2016
Sorry - I was unclear because I was thinking of 6.1. In 6.2 and 7.0 there is a signal in the Heater Group called HK Bias. That's the one we need to measure. I referred to it as 'unused' because there is literally nothing wired to it. But we will use it to measure the input bias, which turns out to apply to all the housekeeping signals except the temperatures. It should read a small number of uA. When we did this in Jupyter we averaged 50 readings. We need to make sure the value is small for each board we test.