VR discrete issues

[mck1117]

rusefi/rusefi#4874
rusefi/rusefi#4922
and maybe rusefi/alphax-4chan#84

Seems that TLV3202 instead of MAX9142 has slightly different behavior, and is susceptible to oscillation on the falling edge that can cause an early falling edge.

In bench testing, adding a 10nF capacitor across pins 4-5 (pin 5 to ground) solves the issue, but further testing required. @ElDominio is going to try it on his real engines that have this issue.

[mck1117]

The bodge I did on the bench:

[ElDominio]

Would I not have to put on at pin3 too? Since I see we're putting a cap across 2IN+ and GND, wouldn't one be needed between 1IN+ and GND too? just for my info

[rusefillc]

@Qwerty-OFF can you help with ^^^?

[ElDominio]

Yaris with 10nf cap across 4 and 5 still has issues

[ElDominio]

changingtocamactivewithinvertyaaris.zip

I changed from not using cam to using cam

[ElDominio]

I'm a fucking idiot, forgot to add -360

New issue: cam reads slow af and sometimes crazy
camcrazy.zip

[rusefillc]

@ElDominio please clarify

"VR discrete issue" not resolved
or
"VR discrete issue" impossible to test because of CAM issue
or
"VR discrete issue" resolved, now separate issue?

[mck1117]

he forgot to add a second cap to the cam channel, and is going to test on a rev E so he can get an SD log

[ElDominio]

Yaris
RevE with 10nF capacitors on TLV3202

OsciLogandNormalLogwithcapsRevE.zip
re_15.zip

[ElDominio]

[mck1117]

Cap seems to have made an improvement to crank signal:

before on left, after on right

Left has very not-50% duty cycle, falling edge comes very soon after rising edge. Right has near 50/50 duty, which is expected for VR signal with many teeth (like a 36-2 crank)

[rusefillc]

#247 for actual change

let's start by improving newer 2 sided version? we might end up making improved one sided version as well I guess

[ElDominio]

We might need another cap according to @mck1117 , @rusefillc . Pending testing

[mck1117]

We might need another cap according to @mck1117 , @rusefillc . Pending testing

the updated schematic has both caps, like the original little test board I did