OK, this is starting to chap my ass.

[psychlow]

So I have a random misfire at idle.

Replaced/new
- Plug wires
- Spark plugs
- Transistor pack (ignitor)
- MPI Relay

Checked:
- Compression - 120-120-120-130 (I'm at 6,000 ft... service limit is 101PSI, standard compression is 136.)
- CAS harness voltage
- Fuel pressure
- Base timing has been adjusted
- ECU has no leakage whatsoever, and communicates very well with my logger
- Valve timing verified correct (and confirmed by compression results)
- 2G MAS is logging airflow, barometric pressure and temperature correctly
- Coil pack primary, secondary, and insulatory resistance was checked. Replaced with a known good unit anyway, and the problem still occurs.
- Resistance from battery to the spark plug threads is 2.0 Ohms.

I know I have 2 major boost leaks - 1 in the throttle body shaft seals, and another in the compressor housing of my T04B - (I need to pull it and put some high-temp RTV in it). The leaks only affect air/fuel, however - and I can adjust that by tweaking fuel pressure on my AFPR. I can't find a single fuel pressure setting that keeps the engine from misfiring.

So, what else could this be? Has anyone ever known of a random misfire resulting from a bad CAS?

 

[ddavisaf]

Fix the boost leaks then post back with the problem

 

[slightlyboosted]

What the first guy said, boost leaks do not affect fuel pressure or flow, they let unmetered air enter the system which causes a lean condition. Upping fuel pressure will not solve this problem, air leaking through the thottlebody shafts will cause a rough or random idle. Repair the boost leaks then go from there.

 

[92 GSR-4]

Psychlow,

I am very impressed with the research done before posting! Very nice.

A major vacuum leak will cause the car to idle bad, but I dont think that is your problem. Before everyone jumps me:

1) A boost leak is a vacuum leak at idle. I feel ashamed I have to say this here, but..

2) I have had 5 or 6 cars run fine with leaky TB shaft seals, before I had time to rebuild them.



If your TOB housing is leaking THAT much, then it shouldn't be a random misfire, it should run bad all the time.

My comments are based on the asumption that this is a random misfire - no rhyme or pattern to it, purely random, meaning it may run fine for a minute and then misfire.

You have exhausted just about all I would do. You may look at the CAS getting worn out. I have taken a few apart. I dont know what year model we are dealing with, but the CAS wires can get old and break on the plug on a 91+ and DEFINATELY on a 90 CAS, those suckers will crack and start shorting out.

Next I would look at my ECU harness plug, and ECU pins. I've seen cars do some crazy stuff if one of the pins is bent slightly causing an itermittent break in contact, ignition problems on one occassion (but the car was running on 2 cylinders permanently).

If that doesn't fix it, maybe its time to check the strength of your spark. Alot of guys have weak spark and don't even know it, but I highly doubt it would cause a problem at idle.

Your ECU might be going bad slowly. I know you said you didn't see any damage, but if you run out of ideas, Dave Mertz @ DSMlink can test your computer and make any repairs. I have seen plenty of ECU's causing idle problems.


If none of those help. I apologize. The only other things I can think of is the elevation and maybe some bad gas.

 

[92 GSR-4]

Talked to a friend today about the problem, he said you might possibly test the Injector resistor pack. Sounds like you've cleared MOST of your ignition, so maybe it's fuel related, like an injector "misfire"?

I have seen some injectors go bad, too. A stock injector should have like 3 or 4 ohms of resistance, maybe one of yours is going bad?

Please post what you find, I'm interested!

 

[ddavisaf]

92 GSR-4: Your advice is sound however, I would not even touch the injectors yet. The boost leaks need to get fixed first and foremost. Major leaks are enough to cause a "misfire" at idle. Oh, and if you use RTV on the compressor housing you're asking for problems. Mach V sells replacement O-Rings for your turbo.
Fix the boost leaks, post back, and we'll go from there.
Good job 92 GSR-4

 

[92 GSR-4]

ddavisaf,

Thanks for the kind words, we're all here to help.

I think what he is talking about on his compressor housing isnt the O-ring (viton) that seals it to the center cartridge, but the bolts that hold it on. Regardless of what my profile says, I'm actually using a SBR GT-12 turbo, and I used white teflon tape to seal up the bolts so they would stop leaking. I noticed the leak on a boost test. Based on that, I didn't think it would be a major leak causing misfires. If anything I would think it would cause a high/ possibly surging idle.

What do you think?

 

[92 GSR-4]

By the way, please don't think I'm trying to say "don't fix those leaks" !!!

 

[ddavisaf]

92 GSR-4:
An actual "misfire" no. However, a boost leak will cause a "misfire condition". It will sound like a misfire, but it will be a rich/lean rich/lean condition that it's trying to compensate for. The major boost leaks will affect you more while driving, and will have a rough, misfire like idle.

Fixing the boost leaks may or may not fix the idling problem. Chances are the problem will lessen. However it gives us a fresh place to start with instead of trying to diagnose around one problem to fix another.

 

[Talesin]

Said that you have a logger... have you tried logging at idle when it's showing the 'misfire' behaviour? Particularly the O2 sensor voltage, timing and knock?

 

[92 GSR-4]

Said that you have a logger... have you tried logging at idle when it's showing the 'misfire' behaviour? Particularly the O2 sensor voltage, timing and knock?

I don't see where he said he had a logger, but it would help if he logged the right parameters.

Logging knock will do nothing at idle. Logging O2 voltage will do nothing, unless you just want to see if it cycles.

Log airflow, and your low fuel trim (once again assuming it's a 1G). If the fuel trim is 100% +- 10%, your fine. Since you say you definately have leaks, I assume your trim will sit WAY upwards of 100%, which means you will have to fix the leak in order for it to stop leaning out.

ddavisaf, I still can't ever remember any of my cars misfiring because of a vacuume leak, just idle problems like unadjustibility, and surge. I may go and "cause" one on my mothers car just to see.

 

[ddavisaf]

Perhaps sputter would be a better word?

 

[92 GSR-4]

I think this thread is dead anyway. This guy must have fixed the problem or something. Unless he responds we will sit here guessing.

 

[Talesin]

First post...
- ECU has no leakage whatsoever, and communicates very well with my logger

Also, why would logging knock do nothing at idle? AFAIK, you *can* still knock just sitting there, if your engine is in fairly poor shape.
And I've never gotten the low trim logging to work, O2 would be a workaround to get a rough idea if it goes super rich or super lean at the misfire.

 

[92 GSR-4]

First post...
- ECU has no leakage whatsoever, and communicates very well with my logger.

Oops. I guess I missed that. It's great he can log, that will definately help him solve his problem.

Also, why would logging knock do nothing at idle? AFAIK, you *can* still knock just sitting there, if your engine is in fairly poor shape.

2 things:

1) I have no idea what AFAIK means

2) Logging knock will do NOTHING at idle. Let me repeat it, logging knock will do NOTHING at idle. The reason, Talesin, is that the ECU looks at the knock sensor ONLY under certain conditions such as throttle position and load. This is why if you are plagued with phantom knock and get knock while cruising, let go of the throttle, and then continue cruising the knock will stop.

The ECU will not show any knock at idle, even if you have a loose rod thats banging about in there and the knock sensor can "hear" it, the ECU will never take a reading from it anyway, and will show 0 counts on the logger. An engine will not detonate at idle anyway, even from super lean conditions, the exhaust may pop a little, but it will not detonate. Detonation only occurs under load.

And I've never gotten the low trim logging to work, O2 would be a workaround to get a rough idea if it goes super rich or super lean at the misfire.

The reason your low fuel trims aren't working is becuase you dont have your car baselined right. This may be because you have something like a vacuume leak (in which your trim will be above 100%), or a fuel oversupply caused by something like a high flow fuel pump and no adjustable fuel pressure regulator (in which your trim is under 100%). Or you may have all the correct parts and just havent richened/leaned it enough at idle.

 

[DSM_Aruba]

I'm having the same problem, but I did a head swap on my 2G and used a 1G head and '91 CAS. Since you're using a 2G MAS, maybe you may try hooking a potentiometer on one of the ECU wires (can't remember which one right now) and adjust the voltage so it reads less than 3 volts I think. I suggest you to check on the other threads for 2G head swap with misfire issues