SaabCentral Forums banner
1 - 20 of 39 Posts

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #1 ·
So I have a problem with an erratic idle that is not explained by vacuum leaks: it can hold steady at 800-900 RPM, but the behaviour is fractured and may suddenly and immediately stall and compensate, then hold steady at 800-900RPM again for a few seconds, before doing this again. When the car is warm enough, it might do this and hold at 1500+ RPM for a while before it's "comfortable" with the idle and settles back down to 800-900 before fracturing again. The behaviour certainly feels very electronic, like it's a sensor, instead of mechanical like vacuum. I'm running a -538 ECU at the moment ('88 LH2.2 unit) but may consider moving back to a -515 ECU ('85 LH2.2 unit), though I think it may not be the problem. I've uploaded a video to show how it's behaving since it's hard to explain.

The next thing you'll notice is when I'm trying to shut off the car, it doesn't shut off, the instrumentation is still live, and either the fuel or system or both relays are clickling steadily, bringing some of the instrumentation with it (ie. start up lights in the panel such as alternator, engine, etc). Fuel continues to run, and the car (as is the case in this video) stalls; at other times the car won't shut down, as the RPMs rise until around 2500RPM, and I have to pull fuse 22 to shut everything down. If I don't start the car but put it in the run position, the relays may also click, and the fuel pump is going (the duration of this and the chance of this happening in the first place is random). I suspect the ignition switch, which I may look at today, but welcome any other ideas.

li Arc
 

·
Registered
Joined
·
18,382 Posts
I like your odometer.

Make sure about grounds like rear tail lights and such, I remember a thing where I had if I turned the car off with the headlights on, and pressed the brake pedal, the oil pressure light come on and the headlights :eek:
 

·
Registered
Joined
·
664 Posts
Ignition switch sounds plausible, bad grounding, short to ground (maybe in the ignition circuit). Several things could cause the erratic idle, but the other symptoms are definitely wiring related, and I'm betting whatever the electrical gremlin(s), it's also responsible for the idle issue.
 

·
Registered
Joined
·
4,203 Posts
Wasn't the 85 model year the year of the self destructing wire insulation?
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #6 ·
Thanks for the suggestions guys; the '85 is notoriously known for bad wiring, and I've been lucky so far. What I'll probably do is do some voltage drop measurements around the chassis and engine to see if anything is amiss, definitely check the ignition switch and wiring close to the switch (since that was replaced over the winter). I wouldn't mind having an entire harness close at hand for whatever may happen in the future (engine harness, APC harness), but seeing as the '85 and perhaps '86 are the only ones with the 22-pin (?) APC in the back, this might present some problems. I also really don't feel like grabbing one from the yard; might check eBay. We'll see how things go, I'm still confident enough in the existing wiring.

Heat shield should still be good, but I'm wondering if I can add some protection by using exhaust tape in the cabin underneath the ignition switch.

Odo is certainly a bit wonky, might fix that soon!

li Arc
 

·
Registered
Joined
·
7,091 Posts
but I'm wondering if I can add some protection by using exhaust tape in the cabin underneath the ignition switch.

anything that stops direct heat will help, but have seen heatsheilds rotted away above exhaust boxes but rest ok
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #8 ·
I took at gander at the problem today, and found some strangeness to it. First, I pulled apart the centre console, and checked the wiring (I didn't pull the ignition switch housing out, but I was able to check the wiring loom nearby). I didn't find any signs of degraded wiring or heat damage, so I left the ignition switch housing. Instead, I grabbed my good ignition switch and simply hooked the wires up to that instead. No change; it wouldn't shut down, even if I pulled all the wires off the switch. So it wasn't the switch or the wiring.

I noticed something else, however. I went through my mind trying to determine what I touched last that started this whole problem, and I recall pulling the AMM to check the resistance between terminals 3 and 6, and tweaked it only slightly from 383 ohms to 380 ohms. I'm quite sure that wasn't the problem, however it's possible something changed in the AMM so I decided to test the limp-home mode. Under that, without the AMM connected, it appeared to shut off without any fuss (although I didn't run enough tests this way to determine if it wasn't simply a stall and if the intermittent relay clicking went away). So I tried connecting it up again, trying (unsuccessfully) to shut off the engine, then pulled the connector to the AMM: the car shut itself off. I'm not sure what this could mean, but it seemed out of the ordinary. I had a spare AMM that put in to try, and while it seemed marginally better, the problem mostly didn't go away. However, it seemed possible to try shutting it off (unsuccessfully), then put back into "Run" position, then shut off again and it just might. This all sounds very confusing, so I uploaded another video.

Lastly, I did a voltage drop test and measured 0.6V max drop between the intake manifold and the negative terminal on the battery, and more like 0.3V on the block and 0.1V from the fusebox or chassis. So I don't think it's a grounding issue. When the engine is running, the positive is around 13.5V, so it's healthy.

li Arc
 

·
Registered
Joined
·
2,986 Posts
I would disconnect the bat. and very carefully check the continuity of all the ignition switch wiring. Not only making sure continuity goes where it should but also where it shouldn't in each switch position. You're still drawing power from somewhere to maintain spark and fuel--with the ignition switch in the 'off' position this shouldn't be happening...

Check if the LH is receiving 12v when the ignition is off.
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #11 ·
Could the dizzy or ignition amp cause a non-shutdown condition? Where do the ignition wiring go, and what would I do to check they are good (ie. do we know that the wiring doesn't go through a circuit back to itself in some way, so that I can simply do a 0 ohm short test on each wire at the switch itself)? I don't have the Bentley on me, so I will have to trace through the wiring later.

However, I'm not convinced it's shorted wiring, and more convinced it's failed electronics. There are several things that don't add up to shorted wiring, namely 1) if wiring was shorted, the clicking fuses would either not click at all or not stop clicking, 2) it wouldn't cause intermittency in shut down behaviour (that it might stall in one instance but raise RPMs in another), and 3) it wouldn't explain why limp home mode works without AMM connected when switched on but it doesn't work when switched off. If the ECU is still running when switched off, then limp home mode should still continue even after RPMs raise but AMM is unplugged.

I'll run a check of the voltage at the ECU when shut off, but I have a feeling it's something else electronic, perhaps a relay or something; if there is still voltage at the ECU when shut off, what relays could cause this? I assume this is the only way the injectors are still going (as well as the fuel pump).

li Arc
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #12 ·
Well, it is not the ignition amplifier or hall sensor connector. I went through the wiring on the ignition switch, but nothing really stands out, and I can't make heads or tails of it vs the diagram in the Bentley (the ignition switch pinouts there are numbered 1-6, but the switch itself has numbers 30, 50, 15, 54, X, and S). Colours are as follows: 30 (grey/red), 15 (green/white), 50 (yellow/red), S (green), X (red), 54 (white/grey). Here are my measurements (all in ohms, or OL for open-loop):

30____50____15____54____X____S
30___0.7M__0.7M___OL___OL___OL
50__________8_____50__3221__OL
15________________57__3228__OL
54____________________3269__OL
X___________________________OL

If someone else has values to compare to, I'd appreciate it. But from what I can tell, unless 15 isn't supposed to be shorted against 50, there isn't much else that looks suspicious...

Edit: oh for crying out loud, the dumb SC forum software isn't going to allow me to put in multiple spaces, so it's ugly I agree, but it gets the point across.

li Arc
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #13 ·
Okay, more information. To add to the above table, I've made the following list, where the NE corner of the table = resistances and shorts in ohms, while the SW corner of the table = voltages in V.

Ignition wires, with car off:

_____30_____50_____15_____54_____X______S
30___X_____0.7M___0.7M____OL____OL_____OL
50_12.86_____X______8_____50____3221____OL
15_12.77_____0______X_____57____3228____OL
54_12.72_____0______0_____X_____3269____OL
X__12.69_____0______0_____0______X______OL
S__12.14_____0______0_____0______0______X

Ignition wires, with car running (which is the anomaly here because it's possible that unplugging wires from the switch will shut off a car under normal circumstances, so I'm not sure if anyone else can come up with this):

_____30_____50_____15_____54_____X______S
30___X_____0.7M___10.7M___OL___0.7M___0.7M
50___14_____X______0.7M___61____37k____10M
15__0.66___13.5______X____0.7M___0.7M____OL
54___14____0.03____13.5____X_____3.3k____OL
X____14____0.04____13.5____0______X______OL
S___13.5_____0_____12.8____0______0______X

Also, I tried pulling the ignition switch relay ('E' in the main fusebox), but it didn't seem to do anything: I could still start the car with it out, and removing it while the car is running (I know, not the brightest thing to do, but...) didn't do anything.

li Arc
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #14 ·
I'm still not sure what the ignition relay is supposed to do, but I think I'm starting to understand what's going on after finding some definitions, and the difference between the two maps I put up. More importantly, when the switch is in the off position, it should look like the first table, voltage-wise. From some research:

30 constant feed battery voltage (+12-14V)
15 accessory
54 ignition switched accessory
50 starter
S warning relay
X radio

Because the position of the switch dictates which connections are bridged, when in the off position nothing is bridged and everything but 30 should show 0V. This is the case in the first table. However, in the second table, we see that 15 still contains voltage (with an expected slight voltage drop from 30) when everything else is off. So whatever is at the upstream end of 15 may be the problem.

This wasn't a problem before I solved my radiator fan problem, so I'm wondering if I disturbed something along the way. Townsend's diagram indicates that the ignition wiring goes through component 152A, a 29-pole white connector in the "electrical distribution box". I'm not sure what this is, but if it's the bulkhead connector at the firewall between the fusebox and the cabin, where my central lock wiring is broken and I pulled the fusebox the other day to see if I can spot anything, then it could explain something. Only, if that were the case, it would more likely be a connector or wire stretched out of place and disconnected, rather than a short. The diagram also lists the red 152B 29-pole connector that sits after fuse 22, which supplies power to the instrument cluster (and which explains why it's still active after shut off). Incidentally, fuse 22 is what I pull to get the system to shut down, since I know the car won't run without the instrument panel (a lesson learned years ago). This could also have been disturbed, and lastly the diagram lists 26 radiator fan timer delay, which I was troubleshooting before all this started, as I mentioned (although in the diagram they say it's for the '89, but I've got one anyhow). In this case, somehow everything started to work with the radiator fan in the middle of testing, so I couldn't get to the bottom of the original radiator fan problem and it's the most suspicious to me at the moment.

I'm still convinced this is a relay issue or something, since the car won't go on forever. In it's pseudo-shutdown state, the RPMs did go up to 1500RPM, but after replacing the ignition control module, it appears that the RPMs continue idle fairly well (or as well as it was idling before). After some time however (something like 1-10 minutes), it'll shut itself down: I had this happen to me several times while I was trying to get voltages from the ignition wiring. Sometimes it was a stall, sometimes it was a clean shutdown. If it were shorted wires that can hold for more than 10s, it should be able to hold indefinitely. So we know that in the pseudo-shutdown state, it's NOT the same as having everything running. Apart from S, X, and 50 (which don't factor into the equation much at the moment), we have 54, which seems to be operating normally. So running 15 without running 54 may or may not explain a bit of why the car can't continue running indefinitely in the pseudo-shutdown state. But I'm not entirely convinced of this.

Besides which, according to Townsend, 54 energizes that mysterious ignition relay I mentioned in the last post, which again links back to relay 26, the radiator fan relay. I'm not sure how this relay didn't do anything whether it was connected or removed last night, but I will try doing a test with this and the radiator fan relay and seeing if there's anything to it. Maybe I can isolate a relay and get to the bottom of this.

All this electrical stuff seems needlessly complicated. :(

li Arc
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #16 ·
You're saying the Hall sensor kept you from shutting down your car? Did it give you almost the same symptoms as in my video? I have an extra distributor lying around that I can probably hook up if needed. But it doesn't seem to be supported by any of the research I just did. Of course, all that research could also be leading me away from the true problem into diagnosing symptoms instead, but the Hall sensor should be an easy enough fix if it is.

li Arc
 

·
Registered
Joined
·
7,091 Posts
Has the car got an immobiliser at all and it's that that the power being routed through?? No idea but just another thought
 

·
Registered
Joined
·
780 Posts
This is what happened to my 900. It was the hall sensor.
I don't think that is hall sensor related.

If is not the grounds I will say that maybe the Ignition control module.

Do you have another one ? ( I have plenty I can send you one for testing).

There is a brown wire in the ICM that it came off in my car and there is another wire there that it's a two wire one with insolation the the other one doesn't have any isolation so maybe your brown cable came off the connector and is touching that one ?

I dunno if that is the issue but it is possible to happen.
 

·
Registered
Joined
·
2,186 Posts
Discussion Starter · #19 ·
No, I don't have an immobilizer installed. In my last post I did mention that I had an ICM that I swapped out (the one I swapped out is a ScanTech, the one I swapped in was the original Bosch I had in the car when I had other ignition problems many years ago...seems to be working fine now though), which lowered the RPMs back to normal instead of shooting up to 1500RPM when the ignition switch is set to "OFF". The swap didn't change much else, though.

I still don't really think it's a hall sensor or something that occurs when the crank is spinning, because (if we look at it as a single problem) somehow when I turn the ignition to "Run" position (ie. before we hit "Start") without the engine running, the fuel pump will either go pumping outright or the relay will click for a few seconds then die off. It can't be a wiring issue since this behaviour is 'resettable': in other words, if it were a wiring short, it either would perpetuate and not end, or if it did end it wouldn't cycle the behaviour again when the car does shut off and we set it back to "Run" again. This is most likely the behaviour of a timed relay or something, but what I don't get is doesn't the ECU need a crank signal to start the fuel pump? How is it that it's able to get this signal before the crank even starts spinning? I did replace the ECU with another one, but no change, so it's not the computer.

I'll try doing some stuff with it tomorrow, including replacing the dizzy, but I don't think the hall sensor has anything to do with the pre-ignition fuel pump start. Unless it was just *that* busted, I suppose. :roll: I might do some elementary wire testing according to the Bentley on the ICM wiring as well, but I don't think I'll expect to find any problems.

li Arc
 

·
Registered
Joined
·
664 Posts
Isn't this the same car that had the elusive vibration problems awhile back?

I'm not at all as well versed at diagnosing electrical faults as you are li Arc. I wish (like I did when you were having those vibrations) that I could help...

The more you do toward solving this problem, the more I am inclined to agree that it is a component fault rather than a short. I just can't get how current can "stay" in action when, what appears to be a functioning ignition switch is turned off. I had a 57 Chevy that did that once shortly before the whole engine harness melted down due to a bad ballast resistor... I know there's no correlation here, just illustrating the extent of my experience with anything minutely similar to your symptoms...

Maybe you could call an Excorcist???:evil:
 
1 - 20 of 39 Posts
Top