you seem to be saying that it doesn't.
Then you haven't read the thread.
I read the pdf. I know where all the methods fall on the table.
With that in mind, re-read what I just wrote, and note that TIME is
what I refer to, not voltages or diode thresholds, which you seem to be
stuck on, even though the relationship is not even linear.
If the TIME numbers are smaller, then NONE of the methods affect
operation enough to matter, since the numbers are ALL faster than the 12V
class numbers.
If the speed does matter that fucking much, shitcan the hard switches,
then he should design a direct driven circuit (custom SSRs) that switches
his signals with FETs or the like. You can even opto-isolate it. Hell,
he can opto-isolate this so he won't get any current back through his
precious chip. He could incorporate optos as a rule.
I have done all this on HV supplies that feed a huge industrial laser
printer. We even made our own HV SSRs that had strings of FETs to handle
the 6kV it was switching. The supply had like 8 different HV supplies on
it, and all were Mpu managed. An 8150C IIRC. I have photos around here
somewhere. There were like two patents filed on it. I think one may have
been the HV SSR. It was hand potted in hard stycast, and then
conformally coated. After all the boards were treated like this, they
were attached to the main motherboard, and the edge leads soldered in.
Then those nodes were also doped and coated.
We had to overcome multiple corona issues since there were several HV
leads under use at any given time that tracked near each other to the
egress side other supply. It was quite a menagerie, and we learned a lot
about corona and HV wire types, and how corona can even migrate a pinhole
through hot Teflon insulation... etc.
WOW. memories. I thought all those were gone already...