[Open_electroporator] Using the feedback signals of kvboard v0.5

John Griessen john at cibolo.com
Tue May 29 14:56:39 UTC 2018


https://www.cibolo.com/docs/kvboard_pyflex_f401_v0.5_tests-8.jpg

shows I-sense feedback of current through the cuvette return conductor.

The current is great for advance setup to decide what levels to use to zap with.
Levels of zap strength for different cuvette resistances as loaded can be in a lookup table
perhaps located in a serial flash memory on board, (not detachable so it is always there as a baseline).



https://www.cibolo.com/docs/kvboard_pyflex_f401_v0.5_tests-7.jpg

shows the HV sense feedback of voltage at the cuvette HV conductor.

The compensation caps added to v0.5 have stopped the fake overshoots seen
with v0.4 HV sense feedback.

When we start making code to do feedback control, it needs to not be very aggressive
and high gain, (not try to shoot up to the top values as fast as imaginable), or it
will be prone to oscillate, over volt, and thus kill the prototype.

I recommend all testing with  the first article of v0.5 be done with recipes
without feedback correcting to avoid that kind of burn-up.  Another way to
say that in terms of proportional integral differential control is to make
the system be over-damped to start with...


I think PID control should be simplified down to PI + some feed-forward,
(using a current measuring test before going up in volts much to decide which levels
to use from a look-up table.  Then any scripted volt profile
can run based on the levels chosen.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: kvboard_pyflex_f401_v0.5_tests-7.jpg
Type: image/jpeg
Size: 409180 bytes
Desc: not available
URL: <http://cibolo.us/pipermail/open_electroporator/attachments/20180529/32e72060/attachment-0001.jpg>


More information about the open_electroporator mailing list