Unstable AFR

[ka24deroadster]

I have a 1996 Chevy LT-1 with a Ford EDIS ignition system and a 5 speed manual tranny. My Megasquirt is a MSii with V2.90 code. I have been running this in my truck for over 2 year with very little problems. Every so often when I would start it, it would seam to run a little rough but if I shut it off and re-started it things would be fine. It had a little issue if you jabbed the throttle really fast because I never did the fine tuning with accel enrichment. With that being said, the truck was a reliable daily driver for over 2 years. Last week it suddenly became unreliable. When I am driving it is fine. When it is Idling it is ok. When I step on the gas pedal to get moving from a stop, the truck stumble and backfires and dies. If i slowly step on the gas it is usually fine. I hooked my computer up to the megasquirt to try and find out what is wrong. I noticed that the AFR was going extremely lean when I quickly open the throttlebody. I also noticed that the Pulse Width became huge so i think the cylinders are miss-firing from too much fuel. My AFR at idle was very rich and stayed rich. Shut the truck off and restart it and the AFR will go back to the normal 14-15 range. Sometimes when I open the throttle the PW will remain reasonable and the motor runs nice. Other times when I open the throttle the PW will jump from the low 3's to over 16 and the motor will stumble and knock. If the motor does not just die, when it recovers the AFR at idle will usually drop to around 11 and remain there at idle. I can rev the motor but when it returns to idle it will still have the rich AFR until I turn the motor off and restart it. I don't think the AFR reading shift is the o2 sensor because the idle PW shifts from the around 3.2 to around 4. Sometimes when I crack the throttle and it stumbles, I will close the throttle and the engine RPM will continues to climb and remain elevated for about 5 seconds (see datalog 2014-12-26-12.37.32 @ 1228) before returning to normal. Does anyone have any ideas what might be causing these issues?

[trakkies]

Got a few comments on your MSQ but I'm more a hardware type than tuning expert.

Normally on a VE table if you're using a sensible target AFR table (which you seem to be doing) each cell starting from the bottom left hand corner will have a higher value than its neighbour. VE analyse live should sort that out if you have a working wideband O2. It looks to me like only small parts of the VE table have actually been tuned. Also, your AFR table has one odd cell on the top line - 16.5 - and you really don't want it going weak on full throttle.

However, if it was once OK, I'd be checking for faults outside of MS before altering the tuning side. Check the fuel pressure is correct and the regulator working.
And the ignition timing with a timing light.

There also seem to me to be lots of odd things in your MSQ - but I'll leave that to those better qualified than me.

[ka24deroadster]

The tune does look like there are some weird cell values. I don't think they were like that when I tuned it 2 years ago but to be honest I don't remember all the cell values. I have not thought about any of this tuning stuff since I started driving the truck. I have not found any mechanical problems with the motor. The problem appears to be with the MS and the PW it is commanding. The AFR at idle is changing because the MS is commanding a longer PW then it should. When the engine is idling and the throttle is opened quickly the engine stumbles and dies because the MS sends a PW in the 15-20ms range and floods the engine. The normal PW at idle is only 3.2ms. When I open the throttle and the MS sends PW in the 5-6ms range the engine runs fine. I need to figure out why it sometimes does the 5-6ms PW and other times it does the 15-20ms PW. I also need to know why the MS is changing the PW at idle. When the idle goes rich the MS will continue with the wider PW and rich AFR until I shut the car off. When i restart the car,and reboot the MS, the idle PW and AFR will be correct again.