1992 Allante Misfire - Help!

[Jeff Pendleton #22056]

Three weeks ago I decided to finally fix that pesky oil leak at the oil cooler line near the filter by replacing the o-ring seal, per my factory suggestion. The repair went flawlessly. However, I had to remove the top of the air cleaner and tube that goes to the throttle body, the cruise cable and, I believe idle speed electrical conncter, then the outer cooler line and the oil pressure sensor. I think that's all I removed or disconnected.

While I could see inside the throttle body, I noticed the two EFI butterflies had a small bit of dirt around the outside edges, so I used some Gumout carburetor cleaner I had on hand to knock-off that little bit of dirt. I could not have used much more than an ounce or so, as my can was almost empty and I had trouble getting hardly any to come out of the can. I did this with the engine running and it nearly killed the engine several times as I sprayed the cleaner. When I was done and tried to drive the car it ran Horrible! I had never had a single drivability issue before and was quite disturbed. At first I had an E017 code (left 02 sensor not ready). I clerared that code and it came back one time. I thought maybe I washed the dirt into an injector, so I got a can of seafoam and put it in the fuel tank and ran it down to 1/4 tank, then filled with fresh 91-Octane gas. No help whatsoever. The E017 code has not come back again, but because I got it twice, I replaced both O2 sensors, again, no help. One of the tests points to the PCM, but that's ridiculous, unless all this is just a great coincidence, which I doubt. I do get random codes once in a while on the tailights and other sensors, and I just clear them and they stay away causing no problems.

A little more detail on my drivability issue: The engine misfires, seemingly on one cylinder. It will cut in and out, running fine for a while then drop back out to a misfire. I ususally notice this change while idling, but it does miss when driving some. I am about out of ideas short of replacing the PCM. The O2 sensors were relatively inexpensive, but the PCM is more and more work as well. Given your collective level of genius, I thought I might ask for even a guess that might help.
Jeff Pendleton #22056

[76eldo]

Jeff

Check the connectors that bring the individual wires to the injectors.
There is a plug on each side. I'd take a look there and also make sure the connectors are tight on each injector.

You can't get anything in the injector by squirting cleaner in the throttle body.
Look for anything loose.

Brian

[Dave Shepherd]

Agreed on the injector issue, you could disconnect one at a time to see if there is an rpm drop on each one, when one you shows no rpm drop this is the suspect cylinder. Hence a  wire, plug, injector issue.

[76eldo]

If you have the shop manual you can do the injector test right from the DIC sitting in the car.

Brian

[Jeff Pendleton #22056]

Brain gave me good advice.  I did the injector cutout test and found that the #4 cylinder was unchanged with I turned-off that particular injector...if the miss was present at the time.  When it is not missing (about half the time) it made a similar difference as other cylinders.  All other cylinders seemed about equal.  So... to see if the injector was stuck I borrowed a fuel pressure gauge from O'Reilly's.  It was a little confusing as I think the gauge was sensitive to sharp bends in the tube (which made it so I could see it from the driver's seat, but it did not seem to work well that way).  Anyway, In the end the fuel pressure seemed good, I guess, unless it is that intermittently sticking injector that is causing the problem and not a kinked hose.  And the #4 injector did fire..at least when I pressed the button to do so.  Its actually kind of hard to feel or hear the difference of whether it is missing or not at idle, but I can tell if it switches in the middle, like at a traffic light.  During this testing though, it is hard to tell when its in and out of the problem.  You can really tell under load though.  Supposedly all the injectors were changed about 11K miles ago and the plugs and wire were previously changed also.

Okay, so tomorrow when the engine is cool I am going to check the #4 Plug wire and plug.  It's really difficult to get to those back plugs and wires. I don't know if a lift would help or not.  I do have a set of ramps, I might try, but I hate to start it and get the exhaust hot!

I am getting really tired of the intermittent nature of this problem. And the manual is not as well-written as the '76 Manual I used to have.  Things are covered in multiple places, with slightly different procedures, and I believe at least a couple of errors (e.g. The ES10 override procedure is described completely differently in two places, and it says to unplug the right O2 sensor when checking both the left and right O2 sensors, but never to unplug the left).  That does not help!

Lets start voting on what I should try after that!

[76eldo]

You might want to go on some of the Allante sites and read up on the engine tilt forward procedure.
I could never get to the back plugs without lowering the front of the cradle.
Watch the power steering pump line if you do it.

When I did mine I used the Delco wires and AC platinum plugs.
Made a big difference but I still had to replace 2 injectors while I owned my 92.

Good luck

Brian

[Dave Shepherd]

You could also test the resistance of that injector at the connector terminals. 

[76eldo]

That's true but the coil could be ok and the injector still clogged or spewing fuel.
If it goes on long enough the computer will throw a code for imbalanced readings between bank a and bank b through the O2 readings.

Personally I'd change all of the injectors and plugs/wires if you are not clear on if and when they were done.

Probably never have to touch them again.

Brian

[Dave Shepherd]

Before OBD11 there were no misfire codes in any car, if it was dumping fuel, yes you would get a rich o2 sensor code.

[Jeff Pendleton #22056]

Please don't forget, this all started precisely with spraying that carb cleaner...I had zero issues before that.  It clearly played havoc when I sprayed it in the throttle body with the engine running.  Far worse than an old carbureted engine.  Plus, the injectors, wires, plugs, etc would not be affected.  I feared about the 02 sensors, but I replaced them with no effect.  I think that's barking up the wrong tree, and I did all the tests.  And I do know when the plugs, wires and injectors were replaced and it has not been that long.  and its not exactly cheap to replace them, nor easy.  I'm really thinking I have done the work to eliminate these problems.

No one has given me any ideas directly related to spraying the cleaner.   Anyone?  Anything?

Here are three questions: 1) Why does the manual warn not to clean the secondary air butterfly?  2) What are the consequences if you do (like I did in ignorance)?  2) Why is this intermittant?  The car runs perfect a good part of the time. 

Actually, according to the manual I am now heading toward the PCM check.  I have eliminated the O2 sensors, plug (the plug on #4 looked perfect, I wrestled it out and back in with the engine in place), wires, fuel pressure, and injector leakage.  I have been getting some ghost codes associated with the power steering sensor.  Maybe something there shorted out the PCM?

The only thing about the PCM that does not make sense is why this started when I sprayed the throttle body butterflies with carb cleaner (yes carb cleaner not throttle body cleaner, but it contained no prohibited chemicals per the manual)?  I just can't believe this is not a direct result of this.  If I had taken this to a shop where they did this, I would accuse them of causing this for sure.

[Jeff Pendleton #22056]

Okay gang!  It took me over two months but I finally fixed my '92 Allante.  It was the Engine Control Module.  I knew there was a good chance it was the ECM, but could never prove it.  Finally I saw a guy on youtube use a Noid light to test injector pulse in his 1992 Deville. He said no flashing light meant no signal from the computer so most likely a bad ECM, which he claimed did fix his car.  He also had an intermittent like I did.  So I rented a Noid light and it was on, then it was off, then it was on again!  That was it for me.  $100 for a reman ECM and I'm back cruising the Caddy.

Noid light!  Who knew?

Oh yeah, about my claim that this all started with spraying the carb cleaner.  Well, it kind of did, although I remember now I had seen some check engine flashes, but had no drivability problems.  I wonder if spraying the cleaner with the engine running made so many parameters out-of-whack at once that the ECM overheated trying to compensate everything?  It was on the verge and kablooey!  I may never know for sure, but I like the theory.

Thanks for all the posts and thoughts about this.  It was a challenge for me to say the least.

Thanks,
Jeff  22056

[dplotkin]

More likely was that you wiped out the 02 sensors with the carb spray and that for whatever unknown reason caused the ECM to fail, maybe software corruption, maybe just a cold solder joint that acted up by coincidence.

Dan