'97 GLT - Dash warning lights staying on after key is turned off. Topic is solved

[Chuck W]

So, this just popped up today. Just finished up the COP conversion on the car yesterday and everything was fine. She drove it with no issues. Today, I went to move it today after futzing with some exhaust stuff. Went to turn it off, and this is what's happening.
VIDEO 1
VIDEO 2

I have checked the ignition switch. The lights stay on.
I have swapped out the Overload relays, still the same.
Checked the alternator. No change.
Swapped clusters (was needing to do this anyway). No change.

Basically what happens is you turn the car off, and either the lights and gauges (note the fuel gauge) stay on and you have to disconnect the battery to turn them off. Or, the second video shows, if you cycle the key to ON and turn it off, the lights/gauges will hang for a sec, then shut off. Also at this time, you can also hear the main relay on the radiator cycle off.
You note that when I'm just cycling the key, when it's turned off, the lights dim, and then eventually cycle off.

It seems to me like a relay is hanging up, or a feed to said relay is causing the issue, but I can't seem to locate the source.

Any ideas?

[misha]

I would say that ignition switch is bad.

[Chuck W]

I would say that ignition switch is bad.

It's less than 2 years old, and when I unplug it, nothing changes. The lights and gauges stay on.

[misha]

Nothing else shouldn't make that kind of problem.

Let's see what others have to say.

[Chuck W]

Nothing else shouldn't make that kind of problem.

Let's see what others have to say.

Well, when I disconnected the electrical plug from the switch and the lights stayed on, that kind of rules it out entirely, in my opinion.

I checked a couple more things, but the result is the same. The lights and gauges all hang up for about half a second, then cycle off, when the key is turned off. They're not staying on now, but something is still slight off. At least compared to my car and from what I remember how this one was.

I even tossed the old ECU in to make sure there wasn't something weird happening in the one I modded for COP, but it didn't change anything when I cycled the key. Things still delayed shutting down by that half second.

[scot850]

This is weird. I had an issue recently but it was the headlight switch that was the issue. Swapping for a spare I had fixed it. But having said that, it does not quite sound like your issue. Like you say it almost sounds like a relay getting hung up but as to which one I am not sure, but would look at the lights. Try pulling the headlight fuse and see if the issue is the same.

I am not the worlds wisest when it comes to the electricary side of things so sorry can't help more. If you can get a wiring diagram, you should be able to see if there is a common hot wire that runs through a relay related to the ignition switch and the cluster.

When you changed your ignition switch 2 years ago did you throw out the old one? Don't know what issue it had but it may help eliminate it as being a culprit if all else fails.

Neil.

[Chuck W]

The "lights" are not the cluster illumination lights, but the warning lights along the bottom.

Take a look at the wiring diagram HERE

The key ON power to the cluster runs through the Overload relay 15+ supply (2/31) on the upper left. When the key is ON, it runs into the cluster at pin 18 and powers the microprocessor (4/36) which in turn feeds the exciter for the gauges (19/7). Also note, that feed on pin 18 also powers the warning lights.

I didn't save the old ignition switch evidently (just was out looing for it earlier), but when the issue started, and ALL the warning lights and the gauges were on, I pulled the plug on the ignition switch, which should TOTALLY negate that as an issue. (Unless the plug itself is bad) everything stayed on. To me, it looks like that Overload relay is sticking. But no matter what relay I put in there, the result is the same.

I suppose I could probe the connections for that relay (It's "106", or the middle "J" relay) and see if it's seeing power when the switch is supposed to be OFF. However, no jiggling or other manipulation of the ignition switch changed anything.

[scot850]

Like you not convinced about the ignition switch. The issue I had affected the dash cluster lights when the light switch went back. All the dash lights came on but not always bright, even the turn signal arrows both came on. Originally I thought it might be the ignition switch but again like in your case it was replaced a few years back. But it was an issue only once the ignition switch was turned on. Everything went off when the ignition went off.

I can't think why it happens once the ignition is turned off. If it was a dead short the lights would never go off, but if the lights are not on when you start the car then a hard short would make no difference. Like you say, when you reset the battery, the issue fixes itself, so something that is energized is staying life which like your logic sounds like a relay.

Short of pulling relays I can't think what else to suggest.

Neil.

[Chuck W]

Ok, got back to it today.

Triple-checked the ignition switch. No oddities there, and all contacts make/break as they should. No bad connections with wiggling, etc.

Pulled the 106 overload relay. It triggers fine and the energizing circuit is clean ON/OFF.

I did notice that when I pulled the main fuel system relay on the radiator (2/32) the problem went away. I'd swapped this relay with the one in my car, and nothing changed, so it's something in that circuitry. Need to investigate it a bit more.

[Chuck W]

Dug into it a bit more and thought I had a culprit, but instead it just gets more weird.

I rechecked my (recent) COP conversion wiring under the ECU to make sure nothing had gotten pinched, etc. I'd made jumpers to run between the "B" side ECU connector and the '24/15" connector that runs out with the engine harness. Didn't cut into anything, but tracing some leads for this issue ran me back to the 24/15.

As mentioned, when I pulled the main fuel system relay on the radiator, the hang-up issue didn't happen, so I dug into that wiring a bit.
As seen HERE there's not much to it. (It's the relay labeled "2/32")

First, I investigated the GN ground wire at pin #3, but that all checked out fine. The only other options were to power sources and/or the ECU. Pin #1 goes to the ECU (Pin A41 - main relay control signal) and Pins #2 and #4 are connected to battery power, but also Pin A26 on the ECU which is also battery power to the ECU.

Now here is where the weirdness starts. I start probing the A-side harness connector (all power was off and disconnected). I had one lead on A12 (Ignition switch power to ECU) and was just checking other pins to see if anything was showing continuity. I happened to touch on A26...and it beeped at me for about a half second. I checked it again, and nothing. No beep. Just randomly messing around, I probed A26 and did the same thing. When I touched on A24 and A25 (Control signals to injector #3 and the IAC) it would beep. After messing a bit, if I kept one lead on A26 and touched A24 or A25 and had it beep, and then touched A12...it would beep for half a second. As a reminder, the duration of the time the cluster is being held on is about half a second.

If you follow A12, you'll see that it also connects to the coil side of the "106" Overload relay, which does what? Feeds the cluster power.

So....


Somehow, there is apparently a half second connection between A26 and A12 when the key is cycled off. That is then back-feeding to the "106" overload relay AND keeping the main control relay signal on A41 active (Which is active when the ignition switch power is received at A12).

What?

I already tested the bad ECU theory yesterday when I swapped in her old (non-COP modded) flashed ECU we'd been using for the past 2 years. It did the same thing.

So, is there something wrong with injector #3 or the IAC? I suppose I could unplug those and see if the problem goes away.

Aside from this weirdness, the car runs great.