I have a most idiotic problem.

On my MS3x, I have very VERY brief spikes in my TPS that are significant enough to trigger accel enrichments without touching the throttle.

Essentially the TPS value will go from 0.0 to 0.3 and back to 0 over a very short duration. The actual magnitude could be higher but not caught by the relatively slow datalog. My SDCard logs aren't reliably working right now. I did get some scaled-up screen shots of some logs that show the problem.





If it were constant noise that looked like a sawtooth I could probably fix it more easily. But since it's very infrequent and subtle, it's harder to nail down.

I have already tried decreasing the lowpass filter just before the CPU input for TPS from 700-something Hz to 70-something Hz and it had no effect. I guess a single transient spike doesn't really get filtered out there.

There are no aberrations in the other sensor inputs on the same 5V and GND (CLT, MAT) so that's also odd... and makes me suspect that the spikes are external to the Megasquirt enclosure. I plan to unplug the TPS connector and see if they persist just for funsies.

I have purloined some 3-conductor shielded wire from work in case I need it but I hate to re-wire the factory setup.

Any suggestions on where there might be noise generated or received would be AWESOME. Remember, it's not constant like the patsmx5 thread so I doubt his solution will help me. But maybe it will.

Experts?

 

Do you have a datalog where the car is idling?