arga

Back in November I posted a circuit for filtering the crank signal on a 99 so Megasquirt would see it as a tach signal. 3 times since I've had a problem where my engine looses power. I assumed it was a bad solder joint but tonight my laptop was plugged in. The RPM reading was 1/2 what it should have been (2000 instead of 4000). The problem stayed on until I dropped back to idle and it synced back up. This would cause and extreme lean condition. I'm NA so no problem but under boost it could have been bad. This isn't a repeatable problem so I'm not exactly what's going on. Could be interferance. My original design had 10uf capacitors across the chips to cut down noise but when I shrunk the circuit down I left them out.

Kills me to post this but I don't want someone frying their engine because of my pride.

Frank

rotaryjunky

iTrader: (3)

Thank you, this site is about information and that is important information. Are you going to replace the caps?

arga

Yeah, I'll replace the caps this week.

I upgraded to 2.85 beta (uses a different pulse predictor algorithm) and ran some logs today. The RPM was rock solid. There was no noise at all. Of course, RPM is pretty heavily filtered.

My guess at this point is that my circuit dropped a pulse right as I started to accelerate. Then Megasquirt's tach pulse predictor program started filtering every other pulse as noise because I had accelerated past the tolerances.

My next step will be to open the tollerences all the way up and log the rpm again.

Frank

arga

Might not be my circuit. Might be the beta code I was running.

http://www.msefi.com/viewtopic.php?t...1b98ba7834503a

http://www.msefi.com/viewtopic.php?t=20669

Frank