I'm hoping someone can help me out with my cold/warm restarts. I get huge oscillations in AFR and the idle is completely erratic. The car runs beautifully once it is up to temperature. I get some slight oscillation when I'm idling with the car at speed, but nothing too bad.

The behavior also doesn't happen all the time, but it does happen more often than not. Seems to be worse when running E85.

I've uploaded a log file. If there other PIDs that are needed to diagnose the issue please let me know.

Thank you for your help everyone.

 

Your timing is far less than what it could be. Looks like there's an adder at work here, needlessly removing your timing down to sub zero numbers. Just at a quick glance, a few degrees isn't going to hurt you even on a the coldest of mornings. There is absolutely no reason it should be firing past TDC. Check your spark adders and limit them *some* and see if that improves your erratic idle. You don't want to zero out these tables. Just adjust where you are seeing the issues. Erratic AFR on cold start is fairly normal. That's why no one tunes MAF or VE on a cold engine.

 

Maybe it is the catalyst light-off routine running. I'll have to dig through the tables and see if that's something I can turn off.

I'll start there.

Thanks Chopperdoc

 

Whoever did this tune changed some **** they did not understand, or loaded a glitchy program. As Chopperdoc pointed out, there is no reason for the timing to be that low, but I see other irregularities with timing vs coolant temp that seem way off. Below 100* coolant, the timing seems to just stay far too low, but at 100*, it looks like it is trying to hunt down a reasonable number while still generally ranging too far and usually too low.. Then, after around 117*, it seems to shift back to some ridiculous low range. I would put back in bone stock values and incrementally work toward a proper tune, because what you have now is for some reason just a comedy of error.

 

No reason to be rude, this is a learning experience for me. I'm working on this tune myself and I've gotten it pretty far as is. When I get a chance I will post my tune file, but the thing is, not much has changed as it relates to timing and different coolant compensations down in the idle range. Most of my changes have been to the injector table for the ID1000s and to the VE table and the upper areas of the spark table. It's possible that it's something I did, but I would be just as willing to believe its something funky in the software. When I was running MAF only it never seemed to hunt like this.

 

There is a cranking spark table. Set to 10 degrees and try again. Might need to increase startup airflow adder also

 

I did not intend to be rude. I have loaded tunes before that just did not load correctly and had random glitches.

 

So I took a look through things. The ECT spark adder table is not changed from stock, and there's nothing else obvious to me that would be causing the issue. I did change the open loop Gas tables a touch, but I don't know if this would be related.

Anyway, I've posted both my tune and the stock file if anyone has time to take a look.

 

So as I suspected, when I revert to a MAF only tune the weird timing issues seem to go away. Timing is still pretty retarded at start up but it advances to my target idle spark pretty quickly. I don't understand why spark would act different for MAF or MAP based tunes.

Data attached.

 

IMO, what I see is in your VVE model, there are very large fueling changes for very small changes in RPM or map readings in the idle region of your VVE. Then when you plug in your maf, it overrides the VVE and stabilizes fueling, which stabilizes everything else. I also see your maf fails on second error. Should set to first error for SD tuning.

Anyway, it then appears as though your spark over/under speed is trying to compensate for AFR swings, so your spark ranges from -3 to +24. Before anything else, the fueling needs to be stabilized. MAP also looks like a sine wave, but much of this is due to the rpm changes and some is likely the TB trying to open and close to help control idle speed.

Bottom line, your idle region fueling in SD mode looks like the root cause of everything else going haywire - and since it isn't like the genIII ECU where you can just key the number into the VE table, this is why you see lots of guys go to maf only on genIV ecu's.

BTW - the ecu goes to low octane tables when MAF fails...

 

That's helpful thank you.

I haven't been using the MAF because the flow is quite turbulent (due to the supercharger intake), but a couple nights ago I installed a 1/2" thick piece of aluminum honeycomb about 1" before my MAF and it cleaned things up a fair bit. Now that the signal is usable I've started to reference the MAF again. The car is running significantly better at this point.

Two notes on your comments:
- I could be wrong, but I do not think that genIV reverts to low octane when the MAF fails, at least from what I can tell it appears to be referencing my main table
- My understanding is that the MAF will fail as soon as it violates the fail frequency, HOWEVER, it will not SET THE MIL until it sees two drive cycles where the MIL is faulted. At least this is how it works in Ford logic.

You are probably correct about the VVE table. Using coefficients to calculate the table makes it difficult to get the surface I want.

Thanks for your feedback.

 

Gotcha. You're probably right. Those were minor points anyway. However, I still think the VVE fueling swings are what's causing the issues, and getting your maf in and the airflow readings stabilized is addressing the root cause. No amount of anything was going to compensate for AFR swings from 0.9 to 1.4. Slight increase in idle speed may help stabilize too, if that's an option worth considering for you.

 

Its funny that you mention that, I ended up trying that about 2 hour ago. Increased target idle speed from 675 to 700 and it seems much more stable.

For now the MAF has cleaned up a lot of the issues. Next time I get on a dyno I'm going to have to pick through the VVE table.

Thanks for your help.