zogster

I do have the original coil, so I'll probably do as you suggest tomorrow when I resume... though I'm also thinking that the TPS switch is worth looking at again. I did open it and clean it while I had everything apart, and while it tested OK that doesn't mean so much in the context of an intermittent problem.


In the course of this afternoon's fiddling about I managed to get it started twice, and noticed two other things that might or might not be relevant:

• on one of the occasions that it started, and ran for 30 seconds or so before dying, the idle seemed to rise and fall slightly, on about a 4-5 second cycle. The cycling seemed quite even, ie. equally spaced peaks and troughs, hitting the same high and low RPMs, as if it was driven by a sine wave.

• while investigating ground connections (I tried connecting the battery neg direct to the ground point on the firewall) I thought I could hear a rapidly opening and closing relay, crackling away in bursts, somewhere in the fuse box. I thought I'd pinpointed it to what turned out to be the sunroof relay, and then realised that the noise was coming from under the fusebox. So I bypassed the immobiliser again, as those are the first relays under there that came to mind. Engine started, then died again. Grrrrrrr…

divil

Hmm...the TPS would probably not prevent the car from starting, but you can rule it out by simply unplugging it for the rest of the troubleshooting process. A stock car will absolutely start and run with the TPS unplugged. I'm not sure if there are any mods people do that would prevent that, but I doubt it.

zogster

Today's results:

Only had time for a brief tinker, but...

Before doing anything else I tried to start the engine - it started, and ran for a minute or so, then died.

I removed the TPS connector. Engine started first time, the idle quickly rose to about 2000rpm, and after it ran for a minute or two I reconnected the TPS switch and the engine died. It then would not re-start.

I disconnected the TPS switch again. Engine started, ran for 30 seconds, then died. It would not re-start.

My conclusion: the TPS switch is not the problem (because it's both starting and not starting with the TPS switch connected, and also with it disconnected. I'm therefore not reading much into the fact that the engine died when I reconnected the TPS switch... I assume that the DME could have a moment of confusion when reconnecting the TPS.).

jeffro951

You said you cleaned all the grounds that you could find, did you get the one from the engine harness to the bell housing? that one being loose or corroded would cause the type of problems you are having. You can always add some temp grounds for testing.

zogster

No, I can't even see that one! I mean, I know it's there, and I think I know roughly where it is, ie. somewhere above the speed and ref sensors, on the engine end of things... but I can't see through the mass of vacuum and cooling hoses, wiring etc

Any tips for getting at it?

divil

If you pull the crank sensor/o2 sensor bracket off the intake manifold, and unplug the 3 sensors, that should be enough to reach it with a socket at least. On LHD models, there are 2 connections from the main harness (DME ground and KLR ground) and one from the battery negative. Not sure if it's any different on the RHD. Removing the heater valve makes it a bit easier if you want to reach in there and clean them.

zogster

Thanks - managed to reach it and clean things up... and things still seem the same. Started first time after cleaning, then died after less than a minute, and hasn't restarted.

There were two ground leads at that point, one heavy braided cable to the firewall, and one other from the wiring harness. It's point 'B' in the pic of two ground points in this thread: https://rennlist.com/forums/924-931-...questions.html

i don't think I can get at 'A' without some disassembly.

Also: as I suspected the HT lead from coil to distributor, I had previously tried gently pulling and manipulating it with the engine running to see if it would prompt the engine to cut out. It did not, which makes me think that the lead is good.

divil

Hmm, that is a little different to the LHD setup. Is the big braided one bolted to the firewall, or does it go through it? On LHD, that cable goes directly to the battery - it sounds like in your case the bellhousing is grounded via the chassis, so I suppose that means more potential trouble spots.

IIRC, you sometimes had to fiddle with the battery->chassis ground to get the car to even crank, is that right? Did you ever get to the bottom of that? Can you make it cut out by doing the same thing? It might not be unrelated after all.

The way I look at it, no matter what is wrong - bad grounds, DME, sensors etc...it has to manifest itself as either a spark or fuel issue, to cause your problem. Once you find out which, you can work backwards from there. That's why I keep mentioning the spark test. But maybe it's time to take a closer look at that ground cable regardless.

zogster

It's bolted to the firewall, but I cleaned that ground point also so I think we can rule that out now, or at least shuffle it lower down the list of likely causes

You're right that I had a problem with the battery to chassis ground but I don't think that's the cause of my problems. Here's why: when I have a problem with that ground, which I can fix by fiddling with the battery to chassis ground lead, I lose all power. But the non-start is happening when I have a healthy 12V. Also that seems to be a much rarer problem than my non-start. Since I don't know what's going on I can't say for sure, but it does seem like that's not likely to be the problem.

But back to basics: I'm sure it's a spark problem, not fuel. I've had the in-line testers on a few times, and every time they've shown no spark when the engine is cranking but not starting.

I'm beginning to wonder if it's the DME itself - it seems like I've tested, prodded or replaced every relevant input and output, so maybe the brain itself is the issue

zogster

Ps. But it's probably worth replacing that battery to chassis lead anyhow.

divil

Could very well be the DME - we know it's producing the base ignition signal, but we don't know if that signal is getting to the coil. But if the only problem was the DME, I'd expect you to be able to get a spark from the coil using the Clark's Garage test. If you can't get a spark from the coil in the engine bay, then I'd say either:

1. the test is somehow invalid
2. the ground point you're using for the test is the problem
3. the power wire to the coil is the problem

Personally I'd be anxious to eliminate #1 - not that I doubt your ability to do the test correctly, but I would want to verify it if it was me. Only when you can produce a spark manually would I worry about the DME. All the DME does as far as spark is concerned is ground the coil negative to charge the coil, then open that ground path again, to fire it.

zogster

Yeah, I think that makes a lot of sense. I'll check the coil again, and double check grounding etc, try another plug, clean the + terminal on the coil.

zogster

I tried the spark test with the coil again, and first time around had no luck.

Then I tested the old coil direct at the battery at the rear of the car. No spark. Which is odd, as it's a simple setup and I think both coils are probably good (reasoning: changing to a new coil seems to have made no difference, the new coil should be working, therefore the old coil is probably working). So I slightly reduced the spark plug gap and tried again. This time I got a spark.

Then I tried the test again on the installed coil, and got sparks OK.

I'm not quite sure what to make of this... it could imply that the spark is weak (because I had to reduce the gap to get things working), or that I'm somehow not doing something right in this test (because I had to reduce the gap to get it to work with the simple setup and old coil), or that the coil is totally OK (because I'm seeing a spark when I break the connection to ground).

In any case, after doing the test I tried starting again. Started first time, ran for a couple of minutes, then I turned off. Started it again 2 or 3 times without trouble. Left it for 15 minutes, then tried again... and it won't start.

zogster

...so I turned my attention to the DME. Cleaned the contacts (there was some corrosion/salts, possibly copper going by the blue colour, at one end of the the two rows of connectors on the DME), sprayed contact cleaner on the multiway connector, and opened up the DME to take a look. No obvious corrosion, damage or cracks, but I sprayed the circuit boards with contact cleaner anyway (there were some brownish deposits) and brushed them clean.

On reconnecting the DME things seem much, much better. Started first time, ran OK, and started again several times after shutting off the ignition. I even ventured a quick ride around the block, and it made it OK. It did, however, die once - so I haven't solved the problem, but feel I'm getting closer.

One other small thing: when the engine is running it's doing a slightly odd thing at low throttle. Idle is smooth, and if I open the throttle up to above 2,500 rpm it's smooth. But in between it seems to suffer a slight dip in revs every 4 to 5 seconds. I'm not going to worry about that too much until the basic starting and running thing is sorted.

jeffro951

Good to hear you have made some progress!