rusnak

I'm trying to figure out why the fuel pump runs sometimes with the key in the "run" position and the engine not running.

I understand the DME will not send power to the DME relay, thus energizing the fuel pump unless it receives a signal from the reference sensor (engine cranking). But at times, the fuel pump runs with the key on. Also, I noticed that sometimes it takes a few seconds of cranking before the engine starts.

I had a problem with "no start" until I started the engine by pulsing an old reference sensor against the sheetmetal at the back of the engine bay. Now, that problem goes away and the car starts, but there is this intermittant problem of the fuel pump running with the engine off and key on.

I'm wondering if the reference sensor gap is off.

I have the gap to the speed sensor set at .8mm (as per Bentley), but did not check the reference sensor gap.

All of this started when I changed the speed and reference sensors.

Otherwise, the engine runs perfectly. Timing seems right, and the AFR is perfect at .97 to 1.03 lambda. The car has smooth power, and otherwise seems ok. I occasionally have a little idle bounce, but that might also be a vacuum leak somewhere. When the idle is bouncing, it is doing so rythmically, and the timing is impossible to read with a timing strobe.

rusnak

thinking about this some more, thought I'd add some more background:

I have a new DME spare. Changed DME brains, and no difference.

I have one or two spare DME relays. No difference. The German made DME relay has two contacts. Both were energized and closed with the key on. I could hear the fuel pump run, and hear fuel flowing in the engine compartment. Not good. I opened either contact, and the fuel pump stopped.

I think I'll mark the position of the reference and speed sensor holder, and then try a larger gap. If the car runs better and the fuel pump does not run with the key on, then I'll know that a larger gap is required.

whalebird

I'm thinking rus. For now though, the sensors should not be sending signal without the engine turning...just being 'close' will not generate the alternating current wave needed for the DME.

whalebird

I hope those 'budget' sensors aren't the problem.....

User 4221

Funny you mention that, WB. I read a post on Pelican about a guy who installed a new flywheel sourced from Worldpac. He installed and got things back together and the car wouldn't start.

He could get the car to start by waving a wrench in front of the sensor. After several hours of troubleshooting, he found the reference pin in the FW was aluminum and not ferrous..... Thus, the reference sensor couldn't pickup the signal.

I've heard other stories about Worldpac but can't comment on Rus' sensors...

whalebird

How about the fuel pump relay...the red one. I have seen some crusty fuse box situations where the wiring BEHIND the fuse/relay is bad. Check the fuse holder itself for weak/overheated tabs and tight setscrews. just a thought.

rusnak

WB, no, the sensors are Bosch sensors, and are an exact match for the original ones. They ohm out perfectly, so I think they're both good. Plus, the car has been starting and running fine. I think there might be a vacuum leak, because the idle pulses if I rev the motor. I have to shut it off, then re-start, and the idle is rock steady. AFR is .97 to 1.03 lambda. The engine runs very well, and the clutch/flywheel are perfect.

I think I'll remove and clean the fuse holders and replace the fuel pump fuse in the front trunk, and also re-clean my grounds. There could be some weird static too. I'm going to take the battery out, clean the clamps again, and just go over everything again.

I took the cover off of the dme relay, and I could see both connections close when the key is on.

But, sometimes the fuel pump DOES NOT come on, and the relay acts normally. That is, the relay is open until I start cranking, then the fuel pump is energized and the engine runs normally. This is sort of weird.

Scott: I saw that same thread and wondered if my reference pin is stainless steel. But if that was the case, then the car wouldn't start with the sensor in the engine. If this keeps up, then I'm going to remove my sensors, and bump the engine until I can see the reference pin, and check it with a magnet.

The sensors are both magnets, and they pick up the steel in the pin and in the flywheel, which is why the gap must be correct. I'm wondering if in my case the gap should be set wider. Or possibly the reference sensor gap needs to be the same .8mm as the speed sensor gap. I noticed that the gap between the reference sensor and the pin is much larger than .8mm. But I reasoned that it needed to be in order to not send a false signal to the dme.

rusnak

Here's a pic of the sensors: (from the clutch thread)

I just got an answer from the PP 924/944 guys. They said the reference pin should be 5mm in height, and the gap should be correct when the speed sensor gap is .8mm.

My gap for the old speed sensor was probably more like 1.5mm and the car was running pretty good. That's why I'm wondering about trying a larger speed sensor gap.

rusnak

Today:

Engine did not start. I got it to start again by pulsing the old reference sensor against the rear sheetmetal. To me, that seals it. I have an issue with either the flywheel bolt or the sensor (not likely).

I'm going to remove the sensors and sensor holder, giving me a huge hole to find the flywheel bolt. I'll use the remote starter button to bump the flywheel around, and I can then remove the reference bolt. I will have to check it to see if it's magnetic or not (could be stainless). I was checking PET for any additional spacers. I think I'm going to find that there is an issue with the new flywheel or flywheel bolt.

There was a paint mark on my old flywheel. Anyone know what does? Why it's there?

****And now for the dumb question:*****
(1) remember the flywheel timing mark sensor? How does this work? Do I need a separate mark or paint or a bolt to trigger it?

It's #12 on the diagram, thanks


---Edit:---
Nevermind, it's capped off, and not used for anything. I should sell it to the 928 guys.



------------------------------------

rusnak

You guys might get a kick out of this:

The reference sensor bolt on the new flywheel is stainless steel (non-ferrous), so that's why I was having trouble with starting the motor. I got to it by removing the left rear shock absorber, then the sensors. I bumped the flywheel around with the remote starter until I could see the bolt through the sensor hole.

Comparing the old and new is very interesting. I don't think the new bolt has any chance of working right given the height of 4mm. The old one is the correct height, as well as being made of steel. So, I'm going to re-install the old bolt after cleaning the flywheel hole out and putting some Loctite on the threads. I'm too tired to try putting the bolt in right now - I'd end up dropping it into the trans bell housing and it would get jammed in the pressure plate. Time for a Stella Artois and an Avo Uvevian.

User 4221

Dammit Janet!!! It looks like you got one of the FWs like John Walker..... Where did you source it?

rusnak

It's in the World Pac parts network as a German made flywheel (oem). I got mine from The Raceline guys. They probably know nothing about these problems, but I bet a ton of people bought them.

User 4221

I just went and checked my new OEM FW and it has a ferrous internal hex set screw for the reference pin.
It sits 5mm exposed from the FW surface..... I'm not gonna mess with it then cross fingers when the sensors are gapped properly.

Like this:

rusnak

If you look at my original bolt, you can see the 5mm is exposed, and 1mm of the bolt head is recessed. I think this is what you might be seeing?

User 4221

I will snap a photo of mine later for posterity.

I see where you had 5mm of the 6mm head exposed so that makes sense.