OK, my idle valve suddenly went crazy.

Normally, its like this (nice and calm)


Right now it's:


Even when I disconnected it, it would still look like this:


Even with zeroed out PID, it would still seesaw like that. If I control using an idle valve test, it would remain stable. Same on gslender FW and official 3.3.0a.

Batt V is above the correction curve. The valve is fluctuating to a value not on any table/duty adder. Valve cleaned and inspected, still doing it.



WTF am I looking for? Wiring issues? I checked at the shop and one of the painters apparently tried cleaning the engine.

Is it strange that I get the same readings even with a disconnected valve? Hope someone has come across this before...

 

OK. Despite all the love in here, I managed to figure it out. I should have rolled back further.

It turned out to be something in the battery voltage compensation curve. We were trying to implement a "floating baseline" where the voltage correction would adjust to varying voltages (my car has lower voltage when heat soaked).

Anyway something screwed up, and even with the feature disabled, it was screwing up the valve duty. I zeroed the battery voltage duty correction and got back to the old smooth idle...



If you notice, there are some small dips because of the lack of voltage duty correction. AIA has more to do, without the idle valve compensation.

Kinda strange that it didn't appear on the first day I flashed the FW. I suppose if you wanna be on the cutting edge, you gotta expect to bleed a little.

 

Well done Greg, I wouldn't have had a clue.