(I need to update my PJ thread. I've been super busy at work and spending my weekends on my MSM build.)

This doesn't seem like the typical stall scenario. I've had stall issues before where setting unreasonable fuel and/or air settings will catch it - potentially preventing the car from entering idle at all. In this case, it seems like the car is preparing to enter idle then just dies off. It seems to require a load in order to occur, simple revving will not trigger this problem.

What happens: Coming to a stop, RPM slowly fall, somewhat stabilize, then go into a steady freefall until the car stalls. Stabilization of RPMs will typically be within a 100-200 of target regardless of what that target may be. Most of the time, the car starts right up; but sometimes, more frequently as the car gets hot, the car will fail to restart. It will crank, stumble to a low RPM very briefly, then die. You'll see failure to restart in the attached logs, followed by a successful restart.

To the best I can tell, this problem gets worse with heat soak. I've been able to get some distance from the house before it'll die. Once the problem starts, it's rather frequent, nearly every stop. If hitting the throttle will help, I've not managed to do that fast enough.

I can replicate this in my driveway. Start the car, back out of the garage, stop on the driveway - it'll likely stall. Pull back into my garage, it'll likely stall. (Note, this is the scenario you'll see in the log.)

What doesn't happen: Stall does not occur at a standstill. I can rev when stationary and the car enters idle perfectly every time.

What I've Tried: (There may be more, this is what I remember.)

• Added and removed fuel from VE: no change.
• Open loop idle: exact same problem.
• Disable Overrun: exact same problem.
• Added air via IACV on closed loop initial table: no change.
• Increased target RPM from 850 to 1000: Problem starts at a higher RPM, but otherwise no change.
• Added air via Dashpot: seemed to delay onset of the problem, but otherwise no change.
• Swapped Cam position sensor (for potential relationship with failed restart): no change.

What I have:

• MSM engine w/+1mm
• MS3 using internal MAP
• Kraken w/Pulsar
• AEM WBO
• GM IAT
• FM brushless fans
• FM 36-2 trigger.
• GM FlexFuel
• Factory CLT

Other important details:

• Running E85.
• Engine still in break-in, about 75 miles in.
• Idle screw set to 750rpm w/IACV disconnected.
• Smoke tested intake, no leaks.
• Closed Loop Initial Table is a little wild - yes. It was ~22-36, I tried expanding it to see if it would catch.
• Idle VE might be weird for similar reasons. I believe Main VE is normal.

I appreciate any feedback. Thanks.

 

Try increasing your engine state settings, specifically "slow decel threshold". Any time you're beginning to stall, you're kicked out of idle mode. Typically I see 800/-800 for those settings, I'd start there.

It's also running extremely lean as RPM falls, you may have to run richer, then give EGO enough authority to lean it out to target. Not quite enough idle time in the log to see ASE turn off and EGO take over.

 

And, with that, I hope everyone had a loose definition of "atypical stall."

Thanks, Curly, this is good stuff.

I've not changed the engine state thresholds, thank you for the suggestion and for the background. I didn't understand why the idle state wasn't kicking in to drive fuel and IAC, that's likely why.

I had previously added 20% to the 25kpa row of the main VE, but I don't think I did that below the Idle VE threshold (1250rpm). If the engine never entered idle, then it would have needed more fuel in the lower RPM range of the main VE.

The tune used to have EGO authority 15 in the cells currently 0 in positive and negative. I removed that as I was concerned that EGO was pulling fuel from last throttle blip through the AFR dip down to 11.7, and as the table leaned out EGO wasn't keeping up. I'll worry about getting this back in place later.

Summary of my plan based on that feedback:

• Update engine states to give idle (and cruise) more room.
• Update Main VE with more fuel in the 25kpa row, 750-1500rpm bins.
• Monitor AFRs as engine RPMs fall and adjust fuel to compensate; aim to overshoot fuel requirements, use EGO to draw it back.

I'll update when I get a chance to try this, but may not be tonight. My first-world problem is not enough parking, I have to park PJ in front of my neighbors to get my MSM out. I appreciate any other feedback along the way. Thanks!

Edit: Would these changes also address the failure to start post-stall? That problem was occurring even after a key cycle.

 

With the Main and Idle VE table set where I'm idling at 10AFR (or less), it still stalls. In one of stall events it leans out into the 15s, but in the last stall it's in the 12s.

Ideas?

 

Are you 100% sure the MAP line isn't pinched at the pedals or firewall? Looks like atmo pressure in your area is about 80-85kpa, which it shows at the beginning of the log, but every time the engine goes to 0rpm, it's still at ~23kpa. I would expect it to return to atmo pressure.

 

Damn, good catch. I completely missed that, but it stands out now that you've pointed to it. Thanks, Curly.

I'm using the same vacuum line I've had in the car since I installed the MS3, but I re-routed it when I removed a mechanical boost gauge in preference for an AEM electronic gauge. The AEM sensor is in the engine bay, with a tee connecting to the old hose running through the firewall. I'll check to see if the AEM is returning to ambient after each stall, then chase down potential crimp / compression locations.

If the AEM is returning to ambient, I may connect the CAN connection and switch the MS3 to external MAP. That's what PJ is using and seems to be working great.

Update to follow when I get a chance to dig into this.

 

F..K.

Well, s..t.

It seems that what little clue I have about tuning, I have even less about routing boost/vacuum lines. I even wondered why it seemed the problem was correlated with the clutch pedal, especially since the switch is locked closed

The best news is that the car drives perfectly. I restored my Idle VE and Main VE tables and took a nice, long drive. Car enters idle perfectly every time. Amazing what happens when the MS3's MAP sensor gets real data.

I'm aware of the other wires hanging, I'll be cleaning those (and that boost line) up now that the car appears to be running correctly. I might still switch to the CANBUS AEM MAP sensor, too.

Curly, thank you for pointing out the issue with the MAP in my logs.

 

I've had this same issue with my car since I got it, and it persists even with the new engine. If I let the engine brake the car to about 1500 rpm's before pushing the clutch it won't die and it idles ok-ish. If I push the clutch in and try to coast it dies almost every time. I thought I had it figured out a few times only to have it continue. I think I even mentioned checking the lines to my MS under the dash but never did. I will definitely have to look at it asap.