Started running on 2 cylinders (but not the two you'd think)

[ksgrimsr]

Hi guys,

I was enjoying a nice spirited drive yesterday and after a brief stop for an errand I fired my car back up to find that it was then running on 2 cylinders.

I thought, that's simple, one of the coil packs is kaput. However, on closer inspection I find that the two cylinders that are not firing are 2 and 4, not 1/4 or 2/3 like you would expect if a coilpack failed. So the misfiring cylinders seem to correspond to the second tap on both coils.

No codes are showing.

Any thoughts?

Knut

[rip]

I'm pretty sure the injectors are paired 1&3, 2&4.

[ksgrimsr]

I'm pretty sure the injectors are paired 1&3, 2&4.

Hi Rip,

That's an interesting lead.

To attempt to isolate this a little, I just got back from doing another quick test (before I saw your note). Since cyl 1 & 4 share the same coilpack and it's 2 & 4 that are dead, I took the lead from cyl 1 and hooked it to cyl 4 (the waste spark goes from the edge to the electrode rather than the other way around, but I figured it would still fire an idling cylinder) and cyl 4 was still dead. So I'm thinking it might not be spark and maybe instead it's fuel.

A quick look at the wiring diagram confirms that 2 and 4 do indeed share a common port on the ECM. I'll start by seeing if perhaps there is some bad connection to the ECM on the UW lead going to the INJA contact (Sheet 4 in the MJ.25 circuit diagrams in the electrical manual).

Thanks for the nice lead. In the event a connector wiggle at the ECM doesn't address this, do you have any suggestion for measuring/confirming a hypothesis that 2/4 injectors are not getting pulsed?

Knut

[GeoffSmith]

Wiki dead cylinders.

[ksgrimsr]

Wiki dead cylinders.

I already re-seated the IHC connector and the ECU connector with no impact, so it's time to break out some real measurement tools and trace things back starting at the endpoint (injectors). In digging into my likely injector situation, what is the ballpark expected DC impedance of the injectors? Can I assume the ElanScan parameter named "Injector Pulse Width" is exactly what it sounds like and represents the duration that the injectors are active so I should be able to readily measure this on my oscilloscope to determine whether the dead injectors are getting the pulses expected?

Knut

[ksgrimsr]

Thanks to the helpful and timely advice of the group, I believe I have a diagnosis (that took only a couple hours from first post to diagnosis). Injector #2 appears to be kaput and when it is connected it also takes out injector #4. Disconnecting injector #2 gives me operation on 3 cylinders (makes #4 live). The DC impeadance of injector #2 is also a bit less than the impedance of injector #4 (I didn't measure the others), so it seems logical that it could draw excess current.

I took some oscilloscope shots of the injector waveforms, but for some reason my scope didn't properly save them, so I'll have to take the screenshots when I go install the new injector and post it for posterity.

Thanks for the help.

Knut

[tweetdriver63]

Thanks to the helpful and timely advice of the group, I believe I have a diagnosis (that took only a couple hours from first post to diagnosis). Injector #2 appears to be kaput and when it is connected it also takes out injector #4. Disconnecting injector #2 gives me operation on 3 cylinders (makes #4 live). The DC impeadance of injector #2 is also a bit less than the impedance of injector #4 (I didn't measure the others), so it seems logical that it could draw excess current.

I took some oscilloscope shots of the injector waveforms, but for some reason my scope didn't properly save them, so I'll have to take the screenshots when I go install the new injector and post it for posterity.

Thanks for the help.

Knut

I had the same problem on my 1 and 3 cylinders, but replacing all the injectors didn't fix my car. It's been laid up in the shop for almost 2 months, and my mechanic is getting help from the folks at GST via email. They're about to start running all the wires to see if they can find any bad circuits, and they think I may need to replace the ECU. On that subject, if anybody has a turbo-catalyst type ECU they could spare, even just for me to use it as a test and return it (that could save me buying an $800+USD part I might not need), I would really appreciate it.

[ksgrimsr]

I had the same problem on my 1 and 3 cylinders, but replacing all the injectors didn't fix my car. It's been laid up in the shop for almost 2 months, and my mechanic is getting help from the folks at GST via email. They're about to start running all the wires to see if they can find any bad circuits, and they think I may need to replace the ECU. On that subject, if anybody has a turbo-catalyst type ECU they could spare, even just for me to use it as a test and return it (that could save me buying an $800+USD part I might not need), I would really appreciate it.

Your description sounds like your shop isn't measuring/diagnosing your issue in a systematic way. You can confirm whether your injectors are getting the signal they are supposed to get by pulling the dead injector plugs off, inserting a suitable dummy load, and hooking up an oscilloscope to confirm that the proper pulses are getting to the plug (and dummy load). One contact on the plug will be at +12V constant (while running) and the other side will get pulled to ground with a pulse width corresponding to the value ElanScan indicates (in the neighborhood of 5ms). Measuring with the plug connected would probably require gaining access to the appropriate contact at the IHC harness connector, but for me I got the info I needed with a simple dummy load and a measurement with a scope.

Hope you're able to get your issue diagnosed and fixed.

Knut

[tweetdriver63]

I had the same problem on my 1 and 3 cylinders, but replacing all the injectors didn't fix my car. It's been laid up in the shop for almost 2 months, and my mechanic is getting help from the folks at GST via email. They're about to start running all the wires to see if they can find any bad circuits, and they think I may need to replace the ECU. On that subject, if anybody has a turbo-catalyst type ECU they could spare, even just for me to use it as a test and return it (that could save me buying an $800+USD part I might not need), I would really appreciate it.

Your description sounds like your shop isn't measuring/diagnosing your issue in a systematic way. You can confirm whether your injectors are getting the signal they are supposed to get by pulling the dead injector plugs off, inserting a suitable dummy load, and hooking up an oscilloscope to confirm that the proper pulses are getting to the plug (and dummy load). One contact on the plug will be at +12V constant (while running) and the other side will get pulled to ground with a pulse width corresponding to the value ElanScan indicates (in the neighborhood of 5ms). Measuring with the plug connected would probably require gaining access to the appropriate contact at the IHC harness connector, but for me I got the info I needed with a simple dummy load and a measurement with a scope.

Hope you're able to get your issue diagnosed and fixed.

Knut

Yes. They are doing all that. They are now to the point I stated above. My car has had a lot of bad things turn up during this process, and it all goes back to a power surge last year that caused a lot of damage to various electrical components. Right now, they're at the point where they've checked everything that can be checked, inculding a long diagnostic list sent to them by GST. They're pretty sure my remaining issue is either some damaged wiring between the ECU and some component, or the ECU itself.

[steve matthews]

Chris, I believe my old ECU is still on the shelf at Knox.
I don't think it would be a problem as a loaner.
Let me know your decision.
Steve-w-new lady

[rip]

my mechanic is getting help from the folks at GST via email.

Gerald is providing support for the US???

[tweetdriver63]

my mechanic is getting help from the folks at GST via email.

Gerald is providing support for the US???

I'd say, as a last resort, he is willing (time permitting) to try to help solve a problem that local mechanics are having a lot of difficulty with. It's not an easy task to find somebody who knows that much about Elans here. I may have made a mistake mentioning it though.

[rip]

I'd say, as a last resort, he is willing (time permitting) to try to help solve a problem that local mechanics are having a lot of difficulty with. It's not an easy task to find somebody who knows that much about Elans here. I may have made a mistake mentioning it though.

Mistake, nah. I doubt anyone would get upset about that.
I shouldn't be surprised though. He's just about one of the most helpful people I've met.
He really does know the Elan well though. He had some involvement during the production stages & has been working with them ever since, so there are not many faults he has not seen.

[tweetdriver63]

Chris, I believe my old ECU is still on the shelf at Knox.
I don't think it would be a problem as a loaner.
Let me know your decision.
Steve-w-new lady

Steve, you have a PM.

[steve matthews]

Chris, you have pm

[Doug]

I'm not going to claim any knowledge here except a fleeting memory of someone having this type problem and solving it thru the injectors. Who was it?
was it a grounding problem of one injector circuit...?
Can't remember

[ksgrimsr]

I thought I'd update the gang on my status.

I installed a replacement injector for the bad one today and the car purrs like a kitten again (I'll be taking it for a vigorous drive later today). I got the new injector from my local NAPA parts store and I'll post the NAPA part number later today (I left the packaging with the part numbers on it down in my shop).

I noticed a couple things when replacing the injector that I thought I'd share with others. Although the info on the Wiki indicates that the plastic wiring guard should be unscrewed, I did not do so and don't see why it would be necessary. Also, the Wiki info indicates that the fuel pressure regulator should be removed from the end of the fuel rail. Although I did so, after doing the work I don't see why it needed to be removed and if I did it again I'd probably just leave it attached. The only thing I can figure is that soaking up the spilled fuel from the rail might be easier off the end of the fuel rail where the regulator is attached rather than spilling the fuel on top of the intake (and likely into the intake for the port where the injector is pulled off the rail). The hardest part of the procedure was releasing the retaining clip on the electrical connector.

Knut

UPDATE: The vigorous test drive was immensely satisfying. The NAPA part number for the injector is 2-18354 (the package is marked with both the brand names Echlin and NAPA).

[DeanG]

I had a fuel injection problem that I chases for over a year. One injector had an intermittent short. If it shorted long enough it overheated the driver and killed the 2nd injector. The problem finally got bad enough for me to confirm it was due to fuel and get it fixed.

[tigerdog]

Congratulations on finding the NAPA replacement and especially on solving the problem.