pcmhacking.net

I had a bunch of trouble with idle when I got the car back from a heads/cam/intake/exhaust/tune upgrade at a local shop. With some tweaking and some help from a couple people on the interwebs, I got it almost perfect. But not quite. The one thing that still doesn't work reliably is the transition from driving to idling. That does work MOST of the time, but not always.

After adding a little more throttle-follower airflow and driving around for about 30 minutes, of stoplight to stoplight, plus randomly putting in the clutch at various RPM, idle was great, and I thought that follower change must have fixed the return-to-idle problem... Every single time I put the clutch in, RPM would drop to about 1200 (because throttle cracker?) and then settle to 850 when the car comes to a stop. Yay.

But on the way back home, I put the clutch in to coast down a hill, and the car almost died. I kept driving around for another half-hour up and down hills, but couldn't figure out how to reproduce that problem reliably. It did come back a couple times though. I have a (very weakly held) theory that putting the clutch in while in DFCO might be contributing to the problem, but when I tried to deliberately induce the problem that way, the problem usually wouldn't come back.

It almost always returns to idle just fine. But once in a while, RPM drops right past 1000, and every time that happens it's a sure sign that it's going to drop all the way to about 300-400 at which point the anti-stall stuff kicks in and tries to save it (and actually does save it about half the time). The only other repeatable aspect of this problem is that when it occurs, the engine continues to refuses to idle for a while, even after repeatedly stabbing the throttle to stop it from dying, or after holding it at 1500ish RPM for a couple seconds. When I release the gas pedal, RPM plummets again. I have to cruise around in gear for at least ten seconds before the car will idle properly again.

It's like there's some condition that's required for idle-control to kick in, and that condition is usually there, but not always. I have no idea what that condition could be, though.

Anyone car to suggest changes? I've attached my latest .bin file and two XDFs that I've been alternating between.

I think data logging would probably help a lot, so I'm working on a logger now, but that's not likely to be ready for quite a while.

I'm amazed you were able to improve it without any data logging at all.

Only because I had help...

One guy pointed out that my idle spark tables had too much advance, so the adaptive spark advance wasn't able to provide a torque increase. That helped. And I did some experimenting with a bit more and a bit less advance.

Another guy, who has a similar cam, has the idle set to use about 10 g/s more base airflow than my original tune. So I tried that change and it helped. Plus some more experimenting with a bit more and a bit less airflow.

And there's the stall saver feature, which shouldn't ever come into play, but until the root cause is fixed, it does sometimes help prevent the engine from dying.

And a couple other people pointed out that you can use the 0 and 400 RPM columns of the idle spark tables to provide a bit of extra kick if RPM drops too low.

All of that and a bunch of experimentation got things working pretty well.

Except for this one last thing, where every once in a while, it just refuses to idle. WTF?!?!

heat soak? I have had idle issue before when heat soaked.

Also more recently I had a random stalling issue on my VR found it to be the throttle blade was too closed and I opened it slightly and all is good. I was chasing my tail on that one as I couldn't pick it from the logs but looking at it again it was relying on the iac too much for the base air flow and becasue they are inherently slow to react it would not catch the engine even with the stall saver and died. So I cracked it open on the stop screw a turn or so and it's been flawless ever since.

I don't think it's heat related... Today it wanted to stall right after a cold start, but usually cold starts are fine. Yesterday the problem came and went while I was just sedately driving around town, and coolant temps were around 180-200 the whole time. But I'll try to pay closer attention to ECT and see if there's a correlation.

This car has a drive-by-wire throttle, so there is no physical IAC, but I suspect that the tune still simulates IAC behavior by adjusting the throttle blade angle, and maybe that's the part of the tune that needs fixing. I'm about ready to try to reverse engineer this stuff to see if there are some previously-undocumented tables in there that might be causing the problem, but that's a pretty daunting task so I'm hoping there's a simpler explanation and solution.

I've read of similar idling issues if VSS isn't working, but don't have any first-hand experience of this.

Joe.

I think you need to collect a log while driving. It sounds likely there is a physically intermittent component in the system, and without logs you wont know whats what. The fact that its so random and you've given it a damn good shot at trying to reproduce it with no result makes me lean towards a hardware problem. If you have logs, you should be able to see whats not right if its tune or an input sensor. And if the tune and inputs look ok, then it might be time to look at the physical electronic throttle.

I agree. So instead of sleeping...

https://github.com/LegacyNsfw/PcmHacks/ ... fad3be3a1b

That looks like a good start

I know i'm resurrecting this, but you've helped me a lot and I just saw this.

I do a lot of Gen 3 LS tuning, did you get your issues sorted out? I have experienced similar issues, where 95% of the time it's great, but that 5% it just nose dives. There are several things that could contribute to it:
1) AFR. If you're cammed, running a bit rich tends to make the car idle hunt/surge when returning to idle. If you've been sitting in traffic and heat soaked, this could explain it.
2) DFCO could do it, but normally the RPM cut off is set high enough that it shouldn't cause a problem.
3) Be careful with just adding a lot of base running airflow to the idle tables! Look at the STIT PID and see what the car wants while it's warming up/warmed. Set it a bit on the high side, but not too far, because the car will trim negative, and if conditions change while driving, it will "remember" the negative trim and won't have enough airflow to catch idle.
4) There are 2 tables for idle spark: the main high/low octane tables. And there's an idle spark table. In stock tune, these two tables are not the same. main high/low octane table is generally referenced when you're moving, but can refer to idle spark table when you're completely off the gas (but the conditions can be edited). I sometimes set the conditions to use the high/low octane tables when speed > 2 mph, and then add 1-2* to the idle region on the main table. The extra spark will give you a bit of a rev hang, but can provide a better chance at avoiding stalls with a cam.