GaryDoug

I wonder if it's possible one of the other coils has an intermittent short to a mounting screw and by tightening the other coil, that causes the short to appear? Have you inspected the bottom of all 3 coils? It looks like the bottom mounting screw is very close to the two contacts for the coil. Also check all 3 coil seals.

hungryhobo

I tried a million variations tonight. I have a brand new spare coil. So I've swapped out each coil in place of the new.
I have looked over the bottoms very carefully.
I've had each coil in a different spot on the ICM. So every coil has been in a different spot at least once, if not more.
I also tried to shim up the problem coil area by putting a rubber o ring under both mounting screws, then under just one screw on each side
It makes it so tedious to have to remove the throttle body each time I want to play "musical coils". I've done it so much now that the throttle body gasket is pretty much done for. It's a two piece gasket now. It broke :/.
I put in the new rubber seal that slides over the ICM contacts that came with new seal.
I haven't tried using the new seal in any other spots however, so that's an idea I'll mull over.
I wish I could just get it back to stalling out at 5k and above. The best I got tonight was just above 3k before it hit the wall, stalled, and tried like hell to find its place to idle again. Sometimes it does. Other times it flat out dies. Then I get under there and play with the tightness of the screws to get it to fire again. Screw it too tight and it won't fire. Too loose and it won't fire
I even went so far as to Daisy chain a ground to each coil tonight as well with no improvement.
I guess swapping out the rubber seals is the only idea I haven't tried at this point. That, and inspecting the secondary sides of the coils where the mounting screws go to see if there is any metal showing. As far as I've noticed there is only a metal ring that's supposed to contact the mounting screw on the side that plugs into the ICM spades and the other side should be just plastic.
I also tried slightly bending the ICM spades to either direction to create a better contact
Maybe I should shim all 3 coils underneath with rubber o rings directly under the mounting holes...I already tried shimming the entire plate last week but was afraid of ICM over heating.
I appreciate your reply and ideas. It's nice to have fresh input from a different perspective

Injuneer

Do you know anyone that has a similar car, and might let you swap the entire ICM/coil assembly to test it? Or maybe a complete assembly from a salvage yard?

hungryhobo

I do not. A salvage yard isn't a terrible idea but I'm not to the point of dropping money I don't have yet..

Tonight..
I pulled the tray the ICM sits on top off. Also the same tray the coils bolt to.

I had never done it before...so when I saw that it was missing the third bolt..down on the bottom..underneath the coil I can push on to kill it..I thought I would have it licked.

I shot blasted the whole tray. I hit the mounting area the tray sits on with a 90 grinder. To Clean and shine it all up. I thought.. "This is the reason why when I push on it, it dies..intermittent ground!" It all made sense.. The tray was missing a bolt down there (which..brw..is impossible to get to without moving the power steering pump out of the way)

Well I couldn't get a nut onto the stud in that area so I wrapped a ground wire around the stud and torques down real hard on the top to mounting nuts. No good..same problem

So then I pull the tray all the way off.. It starts and revs to redline! (I think..couldn't see tach as I was revving via the throttle body) it responded better than ever with the tray being lifted away from its mounting spot.

I accidentally touched one of the tray studs to its mounting location and instantly killed it! So it turns out the coils don't need an external ground. I suppose they get all the ground they need from the power wires going into the top of the ICM because the car ran damn fi e without it.

The weird part being ad soon as I touched the tray to its mounting location it KILLED the car.

I then though I could just shim the tray up with rubber o rings to keep it from ground g out on its lint location.. No go.. It still acts the same..
Thought I was making progress but back to the drawing board...

Something is making it short out when it contacts the mounting location.. It it could all just be anecdotal I suppose..

I'm now planning to pull the fuel rail to see if I can find any bad wires going to the injectors.. No other wires are touching the mounting location so I'm at a loss there

hungryhobo

Last night I got another DTC 82, as well as the " injector fault this run " flag lit up.

I went through the injector wiring harness. Checked continuity between all the pins and connectors.

At one point it after it stumbled.. When it caught itself and stayed running, it was so choppy running I swear it was only running on one bank of injectors. I still think maybe that's where the code 82 is coming from

I believe something is shorted on one of the injector banks causing the PCM to disable it for a time period. It seems like it won't start back up and when I go back to it hours later it fires right up.

I guess I have to procure some test equipment to see what they're doing while running

hungryhobo

Continuing the saga here..

I messed with it throughout the weekend. I think my DTC 82 came from a failed butt splice I put into the harness a few weeks back. I moved, and articulated the ICM tray around so much that it lost its grip on the wire. I soldered it up right like I should have done to begin with.

My latest theory is still based around a fuel injector that is shorting internally after a while of running

I've seen the injector fault flag more than twice now. That coupled with the fact that after it dies a lot of the time, it won't restart until some hours later. On many occasions I'll leave the car frustrated because it won't start back up..I'll come back the next day and it fires right up.

Nonetheless, I couldn't get any type of signal out of the ICM 3x reference yesterday..but it wouldn't start. Today, I went tobstart it and it fired right up. I don't know of anything beyond a missing signal from the sensor itself that would cause the ICM to not send the reference signal to the pcm ..its 5 volts DC if I remember right. So I'll try to start it again this afternoon and hook up a meter to the red/black and blue wires exiting the ICM bound for the pcm. If anyone knows why the ICM wouldn't send the signal out beyond no sensor input I'm all ears.

When it comes to the injector theory..I don't know how I'll approach that just yet. I thought about starting the engine and unplugging one connector at a time and seeing if any difference happens. Its either that, or use a needle to pierce the wires or harness going to the injectors and meter them one at a time. I'm unsure as to what sort of signal they get..if I'll even be able to read them with a meter. Be it DC voltage or ac voltage.. I imagine it switches off and on many times per second so it would be kind of hard to get a reading. Another venue is to use a makeshift mechanics stethoscope to listen to each one.. Any ideas?

Injuneer

Each injector is supplied with constant 12V power with the key on. Connector off the injector, key on, check for 12V between the pink wire in each connector and a chassis ground.

With the connector off the injector measure the resistance across the two injector pins. Should be 12-18 ohms.

The PCM supplies a ground to fire the injector. The pulse width is so short at idle (3 to 4 1/1,000's of a second) that is isn't going to show up on a volt meter. That's why you need a 'noid light. They are inexpensive. Failure of an entire bank of the injectors would point to the 12V supply wire from fuse #9 or #10. Each injector pulses once for every 2 rotations of the crankshaft. At 800 RPM, it takes 0.15 seconds to complete 2 crank revolutions.

DTC 82 can only set during cranking, when no 3X reference pulses are detected. Per the diagnostic description, the problem lies in the ICM, the power and ground supplied to the ICM or the 3X sensor.

GaryDoug

It seems fairly obvious to me that something is connected/shorted to the tray that should not be. Have you examined the ICM carefully where it connects to the tray? Did you ever check those seals?

hungryhobo

Hey you two. I haven't totally abandoned the car or this thread just yet. I appreciate you guys still taking the time to reply! It makes me feel not so alone in this venture.

I had to take a break from beating my head against a brick wall for the past 4 days or so. I've been running around in circles. Thank goodness I'm fortunate enough to have access to another car to get back and forth when I need. It's been non stop cold and rainy for days now, but the sun is out today finally.. So I plan to dive in for a little bit this afternoon, after work.

I'm certain I've tried many more things since I last updated. I had the mindset to stop focusing on the ICM and coils and test the injectors the lazy way by starting the car and unplugging one injector and then revving to reproduce the problem. The plan was to eliminate each injector that way. I never got to carry it out though. I tested the number 6 injector the last time I messed with the car, utilizing the crude method I outlined above. I revved with it unplugged. The car still died. I came back the next day to"test"another injector... It wouldn't start. Wouldn't start the next day either. The day after that I jumped in and tried to start, and it fired right up! I left it running to charge the near dead battery while I went inside to do work things (that's where the car resides). I popped back out twenty minutes later and the car had died. That took place on this past Saturday... I haven't had the heart to even bother starting it since then. The battery charge is waning, so I need to be sure that I have a chance of it starting before I go cranking it. I COULD run an extension cord outside and use the battery charger in the shop but I'd rather not if I don't have to.

So that leads me to today. I came outside on lunch and tried to start it. Nothing. So here I am sitting in it while I update this thread.

I DO know it's for sure not putting out spark. I made a spark tester from am old plug in my toolbox and wrapped some thick gage wire around the threaded portion of the plug and grounded that wire.. That told me it's definitely a spark/ICM related issue.

I had this crazy theory last week sometime about pushin down on the coil and killing the car. I posited that since the actual ICM guts reside directly beneath the coil in question, that what was happening is when the coil is pressed down upon.. It's actually pushing the aluminum casing of the ICM down and shorting the ICM out on something exposed inside of it (even though everything SHOULD be covered by epoxy inside the ICM case.

I tested this theory by utilizing a hole that was already in the top of the aluminum case..that had scotch tape over it. I took the tape off and blew compressed air inside to balloon out the top of the aluminum case. Well obviously my theory was wrong or I wouldn't be writing this novel right now!

Anyway... I'm sort of lost on what my plan of attack will be for this afternoon. I feel like I've combed through everything about five times now! I've followed the crank sensor wiring from the brand new sensor, all the way up to the ICM. I've popped the pins out of the connector to give them a little squeeze to ensure good contact with their male counterparts. I've run new wires AT LEAST twice from the female ICM connector where I originally found the crank reference wires had melted together.. And spliced them in further up the line. I've checked continuity from the pins up to my splice.

I HAVE checked the seals under the coils in response to your question GaryDoug. I even threw some thick rubber o rings under the mounting bolts for the problem coil. I've spaced the ICM tray away from its mounting location with rubber rings.

I HAVEN'T pulled the pcm down yet in order to check to see if the signal is making it all the way back from the ICM (for crank reference) I have followed those reference wires as best I could without ripping off heat shield tape and short of just ripping the wires out all together.

The constant variable has been how tight I torque down the coil mounting bolts on that one particular coil above the ICM brains. I imagine if I take the tray off today and pay around with the tightness of those two bolts it may start.

I guess the next step is to actually start probing connections at the pcm. I already have that part of the dash dropped where it should just be a couple bolts to drop the pcm... Beyond that I'm toying with the idea to run my own set of wires from the crank sensor to the ICM (twisted, of course) and reusing the existing connectors on either side. Running new reference signal wires to the pcm itself however would not be such an easy task.

Anyway, lunch is over...I appreciate all those who are following along with me for the ride. Hopefully someday soon I can post my resolution and possibly help someone else out through my own trials, and strife!

hungryhobo

I'm looking at the trouble shooting charts you supplied right now. I appreciate that, and the injector info.

I realize that you both have supplied me with enough information at this point that there isn't much else to do, but sit back and watch! So sit back and enjoy the show! It's gonna get fixed one of these years. I just need to find the motivation to get to it

hungryhobo

EDIT: NEVERMIND THIS POST.

I looked at the diagram more closely and the circuit 239 it refers to is the actual power wires that supply the ICM. my bad. Jumped the gun


I don't quite understand the first part of the trouble shooting diagram...
It says connect a test light to the 2 pin connector (the actual connector that runs straight from sensor to ICM in a twisted pair)

The next step says "ignition on"...

I interpret that as KEY ON... Not cranking.

So then it asks.."light on" or "light off"

What doesn't make sense is that the light would ALWAYS be off unless you're actually cranking the engine! The sensor itself produces an AC voltage. (550 to 650mv in my case). So the light wouldn't ever be on in any situation I can imagine.

Am I not understanding correctly?

hungryhobo

That seemed pretty apparent to me as well.. But then nothing else is touching the tray. The tray sits on top of a bracket that's bolted to the block.

Maybe it's all anecdotal coincidence.

Maybe it's just the articulation of a bad wire or two causing it. I've checked them all thrice over though at the least...

Injuneer

My "issue" if you want to call it that, is just my serious unfamiliarity with the details of the 3.4L V6. Trying to help someone else, I didn't realize the cylinders were numbered differently than the SBC V8, and I provided bad info. Apparently it doesn't matter if the V6 encounters injector faults (why isn't there a code?),or loses the crank sensor. Now I see from your post above:

Is the 3.4L V6 PCM under the dash? I've always assumed it is right where the one on my 94 V8 is, under the hood?????

It's just a huge effort to page through the factory manual looking for info on where parts are located, wiring harness details, etc. And at this point, I don't see what more I can offer, because it's beyond just using logic. Seems to be involved with the physical construction of an ICM I have never seen.

hungryhobo

I never looked specifically where the pcm was in a manual tbh. I though that looked like a PCM to me as well..under the hood on the passenger side, up against the firewall right? It didn't seem to have enough connections going into it though, so I assumed it was in the same spot as it was on my 85 firebird (r.i.p. trans am.. :'(..uninsured driver took it to an early grave) which is inside cabin, passenger side under the dash. The cover is already off because someone installed amp/speaker wires and ran the RCA cables through that way and never bothered to screw it back in..so I only had to pull it down..there is SOMETHING up there with a bunch of wires..

Hopefully I may NOT have to go that route, as I MAY have had a breakthrough...

I followed the chart that you posted earlier..finally.I found it strange that there was no test for the 5 pin connector, but whatever..

According to the chart, I end up at "replace ICM" Well, I don't like that resolution very much because that is pretty much THE FIRST thing I did over a month ago! Which then led me to buy a new coil that I'm now stuck with..and so on..and so on..

When I put the new ICM on over a month ago, the first thing I did was press down on the problem coil area and the car died. I didn't try to rev it, or drive it, or anything. Because the car died when I pressed on the problem area..I immediately dismissed the idea of a bad ICM. I also immediately took it back off because a hundred dollars is a lot of money to lose on a part I don't need! I successfully returned it and that's when I bought the new coil. One thing I noticed with the ICM I had (from Napa) is that the box had greasy finger prints all over it. It had obviously been bought and returned before I got my hands on it. Which is why I didn't feel so bad returning it myself!

Okay, with the back story out of the way..back to the chart telling me to replace ICM. I found this hard to believe..so I decide it's time to carefully dissect my ICM. I wanted to take the aluminum top off and see if anything sticks out as the reason why I can push down and kill the car... I carefully scrape the the glue bond with a razor knife and coax the top off after about ten minutes.. I DID see something amiss. It may explain it all. I'm kind of afraid to call it a done deal though after all this trouble.. especially knowing that replacing the ICM was the very first thing I did!

I'll try to post a picture.. But what I saw was a burnt spot. It's pretty small. It looks to be originating right from one of the pins. It's a metal strip with many connections to it. But it's definitely gotten way to hot at some point..it's melted the potting compound around it a little.

I should also note that I've had this ICM tested 3 times now. They put it through the paces. Tested it at least ten times in a row..every time it passed. However they never tested it with the coils bolted down on top of it, which would flex a bad connection..it would explain why pushing down kills it..it would explain why how tight I torqued the coil down mattered so much..

I'm gonna buy ANOTHER new ICM tomorrow with my last hundred bucks..this time NOT from Napa. I guess then I'll know for sure..

That's all I've got for today. I'll post up a picture below if I'm capable.

GaryDoug

I assume that burnt piece of metal goes to the connector pins. Is that right? Which pin number is the one that is burnt? Or the wire color for it? If it is E or F, that could explain the DTC82 setting.

If so, here is what I would do: Get some alligator clips (Walmart, Radio Shack, etc) and make a jumper with a piece of wire by connecting the wire ends to each clip. Then jumper that piece of metal to the wire in the cable going to the connector. If that fixes the issue with the pressure-sensitivity, then you have your smoking gun. Of course, that may not be possible if you can't test it with the clips in place.

Tyco Electronics 1.25 in. Non-Insulated Alligator Clips (4-Pack)-CPGI-ALLIGATOR-1-1/4 - The Home Depot

3M Alligator Test Clips - Walmart.com

If that fixes it, I would go even farther and solder a wire to that metal and run it out through the case to splice it to the cable's wire. But that requires some soldering.

If none of this is possible due to the coil being too tight against the ICM, then I would cut away the plastic over the burnt area and see if a break is present. Then apply the soldered jumper wire fix.