I have a coil and some plug wires laying around that I might swap in to see if the stumble clears or changes. I also might try to clean up two of the plugs that were pretty black when I pulled them before. If none of that works. I'll look into those tables!
The spark retard is very interesting. I've never touched or even looked at the sensor. I'll try to identify it when I get the chance.

 

I don't have much of an update today, but I did put some other plug wires on and the problem did not change. Since those are the only things I changed since the clogged injectors took the car out of commission, I doubt anything else relating to the ignition system is the problem. I'm really looking at fuel / the tune because those are what have been changed since this whole thing began.

I do have a couple of notes
Right after starting the car (engine cold), for like 30 seconds or maybe a minute, there is no misfire. It runs very smooth at all throttle inputs and loads. But the misfire starts up very quickly. I only got the chance to drive the car completely cooled down twice today but it did the same both times.

The misfire gets worse with more throttle, until about 75% and it clears up almost instantly. So very light throttle relates to a very light misfire, and it gets worse and worse then all of a sudden clears up pretty much all at once. If the engine is under a significant load, however, it won't misfire regardless of throttle position.

The misfire does not care about RPM, if the throttle is under about 75% and there is no significant load, the engine will misfire all the way up to the limiter (and will start as soon as I start letting out the clutch).

I tried to get the A/C to come on so I could possibly get it to misfire in neutral, but it wouldn't come on. It may be because I have my datalog plugged into the ALDL (even though the computer isn't plugged in). I didn't pursue it any farther tonight but I may look into that further, it seems like it may be valuable to get the engine to misfire while I can get out and do some physical checks on the engine.

I prepared a bin with the tuning changes you recommended, but I haven't put them on the chip yet. I've been busy today, but decided to drive it since I wasn't going far and I wanted to know more about the misfire. So here it is. I don't know if any of this is helpful but at least I have a better understanding of it than I did before.

 

Would it be a bad idea to unplug the knock sensor and see if it gets better?

 

What you're describing seems to be a problem when in closed loop. Both the starting up cold (open loop) and the heavy throttle (power enrichment) are both times where the ECM is ignoring the O2 sensor.

What you might do as a diagnostic is raise the minimum coolant temperature for closed loop fueling to the maximum value to ensure you stay in open loop. Then you can see for a longer period of time how the car runs vs closed loop.

To disable the knock sensor, don't just unplug it. You'll get an EST error which could muddy the waters on the diagnostic value of this test. Rather, go to the spark tables and find the ones for Max Knock Retard vs RPM and set all values to 0. When you do this, I'd keep it to about 50% throttle or less just for safety. But given what you're describing, it should still be informative.

But, do each thing separately. Only to one thing at a time to see what information we can glean from each step.

It's really good you're able to do tuning and take data. This kind of a troubleshooting effort over a message board would be far more difficult (and costly...probably having to blindly throw parts and money at the car).

 

Forcing open loop and disabling knock retard (separately) had no effect at all on the misfire.
i wasn't able to get xdl files for them because my computer refused to connect to the ALDL. Its very off-and-on about doing that. I can go out later and try again to get xld's, but there was absolutely no change in the car's behavior.

I think the startup thing is related more to engine temperature than open/closed loop. The "grace period" doesn't apply when the engine is already up to temperature. I did verify again today that on a cold engine, the engine runs and accelerates smooth for a short time before the misfire starts up.

 

I tend to agree based on the available evidence...

So to recap...

1.) I think we've cleared up your high BLM issue with the VE table updates, as well as catching your bin calibration up to your AIR disable and MAT sensor location changes. BLM's look really good now.

2.) Sensor readings all look good in your datalogs. So evidently at least the driver side of the engine appears to be running well.

3.) You are getting some decent amount of knock retard, but disabling knock retard in the bin has no effect

4.) Open loop when cold appears to allow the engine to run well. 30 seconds isn't enough time for the engine to develop any significant amount of coolant temperature to lean out the mixture. But it may be that start up enrichment decaying out over a period of time on top of the typical "rich" open loop mixture may be enough to mask over the problem.

What it seems to me is that you have a problem on the passenger side of the engine. Bad injector perhaps?

Have you ohmed out all your injectors? Any on the passenger side of the engine out of family with the rest?

One other thing you can try is to force a richer open loop mixture as another diagnostic to try to isolate the problem to fueling. Do this in your open loop bin.

Your open loop tables are still stock, so what I would do is this...

Open Loop Change to Fuel/Air vs Coolant Temp Table => Change the values at 8°C and above to something like 8. This will command AFR's down around 13.5:1 down to 12.5:1 depending on MAP.

It'll definitely smell rich, but if the car smooths out, it suggests that likely one of your injectors on the passenger side is potentially malfunctioning and it's taking this amount of a rich mixture to make that cylinder(s) run close to the stoich 14.7:1

 

All your points look good to me. And sounds like a good idea. I'll try this out.

All the injectors are brand new, but I replaced them because my sending unit had rusted and clogged the old ones. I cleaned out the fuel lines as best I could and replaced the sending unit. I ran fuel through the lines and it came out clear before I installed the injectors. But there is still some doubt in my mind that it might not have been good enough. So if there is an injector problem, its probably dirty or partially clogged from something that I missed, or got past the filter somehow. Which is what I'm suspecting since these past few tests haven't had any effect. I'll ohm the injectors to be sure, but I trust that FIC did their due diligence making sure they were in good shape before I got them.

I'll try out this open loop table you suggest, and see how it goes from there. I'll let you know.

 

I was just about to make the bin with this open loop AFR change you suggested, but it looks like the %change has already been set around 9. I never did this so I'm not sure how it got this way. I'll attach a picture. I set the values above 8 degrees to 14 to test it anyway. (14 is about what it was at 8 degrees). The condition has not changed.
I did manage to get a log this time, which may be useful. The first few minutes of it the engine was cold and there was no misfire. When the engine running time reads 232 I come to a stop, and when I accelerate again the misfire begins. Then throughout the rest of the log everytime I accelerate there is a misfire. It is interesting to note, there was absolutely no misfire until I came to that stop sign, then as I left it came on with full force. It's like off / on. This log may not be useful but I'll email it to you in case there is something valuable to be found in it.
It's also worth mentioning that I forgot to stop the log before shutting the engine off, so there's some bogus data at the very end where I shut the engine off. So if you want to fill out history tables, exclude the last few seconds.

 

Here's the values I found in the bin before I adjusted them. I set everything above 8 degrees to 14 for the log

 

Can you send me your latest bin?

Last time you said it went 30 seconds before the misfire started. Was that with the richened Open loop table too?

Now it's gone for 228 seconds?

BTW, there may be something going on with your EST system after all..... I'm not sure if its related, but when you go to closed throttle, your spark timing drops out. Closed throttle spark table has 20° deg as the minimum, but I'm seeing spark timing as low as 7° in some instances. Nothing else in your data seems to indicate why the ECM is doing that.

Thinking about ignition modules...I know aftermarket ignition modules can sometimes cause the final timing to be different than what the ECM commands due to different spark latency, but I don't know that I've ever seen an ignition module actually influence what the ECM commands, nor why it only seems to affect closed throttle timing. But I don't know what else could be doing this.

I posted over in the DIY PROM section in thirdgen.org as there are a number of experts over there...

https://www.thirdgen.org/forums/diy-...ml#post6548664

 

Yeah I'll send the bin.
I checked the stock bin and it has the open loop table I posted above. So everything I have done so far has been with that open loop table. Only the log I just sent you was done with the open loop table set to 14.

I have to say 30 seconds was an estimate. The first time I noticed the misfire had cleared, I wasn't expecting it so I didn't have a really accurate read. It was only about 30 seconds after I noticed it wasn't misfiring until it came back. Then after letting the car sit for several hours it was a similar amount of time (but it may not have been completely cooled down)
Probably more importantly, it has gotten significantly colder here in the past few days. I think it was in like the low 60's when I first noticed the misfire cleared on startup, and this morning it was in the mid 30's. The temperature changes pretty rapidly here and its kind of inconsistent.

The ignition advance you noticed is interesting. There's nothing I know of that would be causing that. The distributor (and therefore ICM) are relatively new but they've been in the car several months now. Hopefully the thirdgen forum can give us some insight there.
I'll follow your post in the thirdgen forum. It's useful to have some more eyes on the problem.

Oh and I'll send you the bin.

 

In open loop, should the O2 sensor still be oscillating? I know the computer isn't referencing it, but I thought it would be more stable than it is, since the fuel trims are based on a table and not being adjusted. Also, it seems like it tends to hit lean spots on acceleration (where the misfire is happening). It seems like every time I open the throttle, the O2 reads very lean (below 100), and when I close the throttle it reads rich (800-900). The O2 actually hit 0 at least once (at the beginning of the log). If this means anything, at the very least it would indicate to me that the misfire is probably a fueling issue causing a lean spot.

I don't know hardly anything about tuning so let me know if I am wrong about this.

 

A heated O2 sensor is going to report what it sees whether you're open or closed loop.

I'm still on the side of fueling too...

Regarding the lean spots.... did you ever try out what I suggested in post #39?

 

Yes I did, but to no effect. The adjusted table is still in the bin.

 

I'm looking at your bin now...

Yeah, looks like you maxed out the AE Delta MAP up to the LT1 values.

Open loop AFR vs coolant values are only 5.4 though. Did you have it up to 8-ish already? Or is this a typo in the bin? If so, maybe put it up to 8, 9, or even 10 maybe?

After this, if the misfire doesn't resolve, I'm not sure where else to go with this (I may need to chew on it for a bit to come up with other ideas).

Hopefully the discussion on thirdgen turns up something.

At this point, the two possibilities appear to be:

1.) A problem with one or more injectors on the passenger side causing a lean condition that we can't directly see

2.) A problem with the ignition module (if in fact it can actually influence the commanded spark timing... which I'm not sure about yet).

Has the severity of the misfire abated at all with these tuning changes?

 

I'm not sure why the table is different for you and me. Either way, I already set the values to 14 across the board and it had no effect.

At the moment I'm really leaning towards injectors. I had been driving the car several months perfectly with no issue until the sending unit rusted and clogged everything. I really doubt something else has failed from sitting in the garage while I worked on it. I'd bet I didn't clean thoroughly enough and something got stuck in the injectors. I won't say I'm sure that's what it is - I know better by now to never be sure with these things. But it makes the most sense to me. I'm trying to think of some ways to test them though and I'm not really sure. Maybe I can soak them in injector cleaner and hope whatever is stuck dislodges? seems like a long shot. I'll get another scholarship payout in a month or two and until then I can't spend any money. I don't mind sending them to get cleaned when that comes in, but I want to be as sure as possible that will fix the problem if I do.

The severity of the misfire has not changed at all. Nothing about it has changed at all really.

I really appreciate all the help. Even if it didn't resolve the misfire, my tune was pretty egregious and I'm very happy to have it all in line. Also reporting back here has kept me motivated to work on it at least a little bit every day. Left to my own devices, it's liable to sit for a few weeks before I get to it.
I'll stay tuned here and on thirdgen.org. The week after next is finals. So next week and the week after are gonna be fun. I'm quite sure I'll check the forums, but I probably won't be able to get anything done on the car during that time.

Happy Thanksgiving

 

Yeah, its seems that something is up with the injectors at this point. Hopefully cleaning them solves your issue.

One thing you might try as a last ditch diagnostic...

Pressurize the system, then go apply 12V by hand to each injector for as close to the same amount of time as possible (maybe like 1 second) and see if the rail pressure drops by the same amount each time, looking for any injector that's out of family. With the primary suspects being on the passenger side. If you find one where the pressure barely drops, that could be a smoking gun.

Happy Thanksgiving to you as well.

 

The thread on thirdgen.org is getting some momentum. The guys looking at your data now are two of the more knowledgeable guys in the DIY PROM section. It seems like they've already read through this thread.

Feel free to log in and join in the conversation. I'm sure they'll have questions and it'll be more efficient to cut out the "middle man" (me) bridging across two forums to get answers.