Computer weirdness

I had some fun with my E30 lately that I'll tell you all about so that maybe in the future, if (when) it happens to somebody else they might see this and find a little help from it.

I've driven this car ('87 325) almost daily for over nine years and never had any problems with it starting and running - an amazing stretch. About a month ago, I went out of town for the weekend. I parked the car in the driveway on Friday and left town. While I was gone it rained like hell. On Monday morning I got in the car and it refused to start for the first time ever, so I drove my '65 VW Crewcab truck with the bad voltage regulator to work. That evening after work I started testing relays and senders and so forth and couldn't find anything that checked bad. I took a day off Tuesday from trouble shooting to ride my bicycle (I'm addicted). Wednesday I tested some more, still no problems found, so I finally took the computer out so I could look inside it to see if there was anything obviously wrong like a broken solder joint. Nothing looked bad, so I closed it up, put it back in the car, and then the car started. It ran a little weird, but that cleared up after a few minutes. I drove it the rest of the week with no problems.

The next weekend it rained like hell again. I went out in the car and it started missing and acting like it was going to die, but I was able to keep it running and it finally cleared up and ran fine. Drove it to work all the next week and it did fine.

The next weekend - out of town again and it rained like hell again while I was gone. Monday morning the car wouldn't start so it was back to the truck to go to work for the week. During the week I did some reading, including the sticky right here on this site about the Motronic unit. I decided the rain had nothing to do with all of this and that the computer must be dead, so I bought one off ebay for $50 including shipping.

When the new (used) computer arrived, I plugged it in and the car hasn't missed a beat since (that's one week so far). It also hasn't rained. So I guess I could still be in for a surprise, but so far I think the computer was the problem.

More details from the trouble shooting: There were two problems keeping it from starting - no spark and no injector pulse. Those are two things controlled by the computer that aren't really related any other way. Those two problems were there both times it wouldn't start.

I'm going out of town again this weekend and rain is predicted. I'll let you know what happens Monday.

John

Still weird.

I got home early enough to try it in daylight. It rained pretty hard while I was gone, just as predicted, and the %$^& car wouldn't start, at least not right away.

Here are the details:

I tried to start it and it cranked and cranked, but wouldn't start. It hit a couple of times but wouldn't keep going. The next thing that happened was that I accidentally left it in gear and cranked it. It lurched forward and tried to start, so I tried again and it started. It sputterd around and ran like crap for a few minutes and then smoothed out. I took it around the neighborhood and it ran fine up to 2K rpm, but sputtered above that (in an old vehicle I would be suspecting the advance in the distributor - the young ones among us can google that). After a few minutes it cleared up and ran fine again, so I headed home. As I drove down the driveway it started missing again and finally died right where I normally park it. It refused to start again and I had other things to do, so it sat for about an hour. When I could get back to it it started right up, revved up with no missing or sputtering around and sounded fine.

I don't know what to make of this. It's almost like the damn thing is mad at me for spending time in the E34 when I go out of town. That makes as much sense as anything I can come up with. Maybe I should print the ad I just saw on Craig's list for the red '99 Z3 for only $2500 and toss it on the passenger seat. Maybe that'll get some results.

I'm open for any ideas you all might have. I'd like to keep driving this car, but I'm not feeling too good about my chances of getting home in it if I take it out right now.

John

Intermittent electrical problems are really difficult to find. Could be some ground issues that cause some sensor to be acting failed for a while. Warming it up and drying it out seems to be the answer. Otherwise, water ingress into someplace like a connector that dries and then allows normal operation. Check for corrosion in all connectors where water or condensation could form.

Bob in Lake Havasu

May be an intermittent vacuum leak on some hose, as well.
Some distributors are controlled by vacuum, can't remember if yours is the case or not.
Again, this kind of problems are the hardest to diagnose, specially for the DIY with limited tools and resources...

Quote
John Yust
Still weird.

I got home early enough to try it in daylight. It rained pretty hard while I was gone, just as predicted, and the %$^& car wouldn't start, at least not right away.

Here are the details:

I tried to start it and it cranked and cranked, but wouldn't start. It hit a couple of times but wouldn't keep going. The next thing that happened was that I accidentally left it in gear and cranked it. It lurched forward and tried to start, so I tried again and it started. It sputterd around and ran like crap for a few minutes and then smoothed out. I took it around the neighborhood and it ran fine up to 2K rpm, but sputtered above that (in an old vehicle I would be suspecting the advance in the distributor - the young ones among us can google that). After a few minutes it cleared up and ran fine again, so I headed home. As I drove down the driveway it started missing again and finally died right where I normally park it. It refused to start again and I had other things to do, so it sat for about an hour. When I could get back to it it started right up, revved up with no missing or sputtering around and sounded fine.

I don't know what to make of this. It's almost like the damn thing is mad at me for spending time in the E34 when I go out of town. That makes as much sense as anything I can come up with. Maybe I should print the ad I just saw on Craig's list for the red '99 Z3 for only $2500 and toss it on the passenger seat. Maybe that'll get some results.

I'm open for any ideas you all might have. I'd like to keep driving this car, but I'm not feeling too good about my chances of getting home in it if I take it out right now.

John

Sounds like the good old connector under the diagnostic plug. This connector runs signals to both the injectors and the distributor. It is a known source of corrosion due to the rubber booth holding water and not draining.
Check it out; that's where my best guess would find the problem.

This is good. I'm going to check inside that boot under the connector this weekend and there's a cluster of ground connections right beside it that I'll loosen up and retorque just to make sure there's good contact. I might even retorque the battery ground strap to the body - just to be sure. I think the only connector I haven't already had apart is the engine temp. I'll get that one this weekend.

Jose, there's no vacuum advance. I've got the cap on the end of the camshaft setup - M20. All the spark timming is done by the computer. There are still potential leaks, but the engine runs so steady (when it will start) that it just doesn't seem like I have a leak. I do have a can of carb cleaner that I can spray on some of the older hoses. If there's a leak the engine will rev up, so I'll do that.

For now it starts and runs just like it always has, so I'm going to see if I can make it to the weekend without having to work on it.

Thanks for the ideas.

John

I like what Peter suggested.

If you loosen the ground connections, might want to clean the contact area and coat them with a dielectric grease to keep corrosion out. Corrosion is the enemy.

Bob in Lake Havasu

I let the car sit Saturday without starting it to see what would happen. Sunday afternoon I tried it and it started - not as quick as if I'd used it the day before, but it did start. I decided to look at the distributor cap again. I remembered thinking it was a little on the cruddy looking side when I checked it a couple of weeks ago. I wanted it to look OK, since they cost so much, and a rotor isn't cheap either. This time I decided to go ahead and replace them. They were in the car when I got it and that was over 9 years and 80,000 miles ago, so it's about time. $93.00 later the new parts were in and the car has been starting fine ever since. Of course, I think it would have been starting anyway. The real test will be when I leave it for a few days again, which will be happening in a couple of weeks.

I didn't get to the ground checks. It started raining before I even got the cap installed. I had to go finish that during a break in the storms. I did check under the boot at the test connector. No plug under it, just a bunch of wires that all looked nice and clean.

More later as I see how it goes.
John

Looks like it was the cap and rotor. I let the car sit for a weekend again and it started right up when I got home. We've also had a lot of rain and I haven't had any trouble with it running rough.

I still can't explain why the injectors weren't pulsing. Maybe I wasn't checking them right, although I didn't smell any gas at the tailpipe either. I also can't explain why I didn't see any spark from the coil wire, but maybe I couldn't see it from where I was standing. I put the wire on the valve cover and stood by the driver's door and looked at it while I cranked the engine. I never saw a spark. I even put the original computer back in a week ago and it didn't make any difference.

I guess the message is: If you're having trouble with starting, don't freak out and buy a computer, check your cap and rotor.

John

Well now it looks like it was not the cap and rotor. The car has refused to start for over a week now. This all started after another huge rain event, which I can't see how it's related, but there you have it. I've been so disgusted I haven't been willing to post about it. Rain does not get into any places under the hood where it should make any difference and the interior of the car is dry. I don't see how the rain can make a difference, but it seems to be related to the problem somehow. Of course, this time the car won't start even after several days, so maybe it's not related - who knows.

The problem is that there is no spark from the coil. The bellhousing sensors check good on resistance. I know the "missing sensor tab syndrom" is one thing that keeps geting mentioned when you search for information on this problem online. Since my car runs sometimes and then won't run and then runs again, that doesn't seem too likely in my case. The "tab" would either be in the flywheel or not and the car would either run or not. I remember from when I had the flywheel off during the clutch replacement two years ago that my flywheel did not have a tab sticking out of the side for the sensor to pick up. I saw a magnet, or something, embedded in the flywheel near the outside edge that looked like it performed the same function as the tab. Also, over the weekend I hooked up my multimeter to the sensor and got a signal on the meter as I cranked the engine on a very low DC voltage scale. That indicates that the "tab" is triggering the sensor. Now that I have two ECUs I can swap them out and the car won't start with either one. There is voltage to the ECU when the key is on. I decided to throw some more money at the problem so I bought a coil and installed it. No joy from that expense.

I guess it's possible that I have two flakey computers. It seems unlikely, but there isn't much else to this system. There are the bellhousing sensors (check good), the flywheel tab (shows voltage spike on meter when cranking), the ECU (who knows), the coil (new) - and that's as far as we need to go because there is no spark from the coil. I have an extra coil wire that I have tried and it makes no difference. It checks good anyway. The only thing I can't really check is the ECU.

I'm about to go on vacation for a week. When I get back I'm going to be willing to spend a couple of weekends on this thing, but my patience is running out. I like the car, and I've put a lot of nice parts in it over the years, but it's looking a lot like a scrap metal sale at this point.

When I get back, I guess I'll look through the sensor mounting hole and reassure myself that my memory of the flywheel is acurate, even though that doesn't make any sense due to the VOM readings and that fact that the problem is intermittant, and I might throw even more moneay at it to buy yet another ECU to see if I really have two bad ones. Other than that, I don't know what to do besides see how much I can recover from selling it as a junker.

Any ideas appreciated besides "burn it." I only have liablity insurance on it, so that won't help.

John

Well, the interesting thing here is everything was fine until it rained. When you crank, is the tac bouncing? Are you getting voltage to the coil? Now it becomes a matter of tracing. I bet there is a connector somewhere in there that is damp and corroded. The art now lies in finding the bad component. Relays?

Quote
alanrw
Well, the interesting thing here is everything was fine until it rained. When you crank, is the tac bouncing? Are you getting voltage to the coil? Now it becomes a matter of tracing. I bet there is a connector somewhere in there that is damp and corroded. The art now lies in finding the bad component. Relays?

I'm with you on the relays; particularly the main relay.

Intermittent problems often come from bad wiring, poor contacts, corroded connectors ant all that...
Those are the most annoying to diagnose.

Try to test both your ECU on another known good car, if you can get your hands on one...Better than replace The ECU without being sure it is not working.
Get the electric wiring schematics of your car, and follow the circuits and check all connectors you can get at...

Good luck!

Quote
alanrw
Well, the interesting thing here is everything was fine until it rained. When you crank, is the tac bouncing? Are you getting voltage to the coil? Now it becomes a matter of tracing. I bet there is a connector somewhere in there that is damp and corroded. The art now lies in finding the bad component. Relays?

No bouncing in the tac when I crank. What does that indicate? I have voltage at the coil. The relays all check OK. I've bench tested them and get circuits where I'm supposed to. I also have extras that test good that I have tried and that didn't make any difference either.

One extremely slight possibility that I've thought of is that maybe I really am supposed to have that tab on the flywheel that I didn't see, and maybe there was just enough of a remnant of it there on my flywheel that the car would run when everything was perfect. Maybe now the bellhousing sensors are getting a little old and tired and they can't pick up what's left of the tab any more. I'm leaving on the vacation trip in the morning, so I won't be able to check this out for a week or so. When I get back, I'll pull the sensor and look through the hole as a helper turns the engine and I'll see what there is on my flywheel, if anything, for a tab.

If I can find anything else to unplug and look for corrosion I'll do that too, but so far I haven't found any and I think I've looked at every plug under the hood.

Isn't there a sensor on the crank snout damper in the front of the engine?

Check this out:

[www.pelicanparts.com]



Edited 1 time(s). Last edit at 07/17/2013 09:44PM by alanrw.

And this:

[www.justanswer.com]

Quote
John Yust

One extremely slight possibility that I've thought of is that maybe I really am supposed to have that tab on the flywheel that I didn't see, and maybe there was just enough of a remnant of it there on my flywheel that the car would run when everything was perfect. Maybe now the bellhousing sensors are getting a little old and tired and they can't pick up what's left of the tab any more. I'm leaving on the vacation trip in the morning, so I won't be able to check this out for a week or so. When I get back, I'll pull the sensor and look through the hole as a helper turns the engine and I'll see what there is on my flywheel, if anything, for a tab.

If I can find anything else to unplug and look for corrosion I'll do that too, but so far I haven't found any and I think I've looked at every plug under the hood.

Good news! It's fixed!

This morning I checked for the sensor tab and sure enough it was there. I guess my memory of the flywheel wasn't all that good. Anyway, once I knew the tab was there I headed to the boneyard to see if I could get another ECU to try. They're only $20 at the local Pull-a-Part, so I figured that would be a reasonable thing to try. After all, everythig checked good, but the car wouldn't start, so that seemed to leave only the ECU as the problem. Pull-a-Part had two '87 325s. The first one I got to had been pretty well cleaned out. There was nothing I could use left on it. The seond one had only been there a few days. The ECU was already gone, but both of the bellhousing sensors were still there and the Mass Air Meter was a nice fresh replacement part, so grabbed all those parts and headed back home. I didn't really think I needed the Air Meter, but to get one for the prices they charge was worth it to have it on hand. One of the sensors was a recent replacement too.

When I got home I put the newer of the two sensors in the hole with the position tab. The car still wouldn't start. I then decided to replace the sensor over the flywheel teeth. I put the other boneyard part in that hole and the car started right up. I tried it with both ECUs and it works fine with both. Apparently, the sensor over the flywheel teeth that sends the signal to the ECU that the engine is turning over wasn't sending the right signal, even though it checked good by the resistance test.

It sure felt good to drive my E30 again. I wasn't sure that was going to happen again for a few days there.

John

...It was the sensor all along... Glad it is running again now!
Plus some parts left to sell on e-bay...

Is there a way to test crank sensors with a DVM?

I'd keep those parts, they'll be getting harder to get a hold off, especially as two of them are quite new.

I would guess you'd need a scope, DVM's are usually to slow to pick up the signals.

Quote
alanrw
Is there a way to test crank sensors with a DVM?

I checked my position sensor on the 2000 millivolt scale on my meter. I got a variety of readings as the sensor tab passed the sensor, but they seemed to all be between 5 and 20. It seemed like it was about as fast as the meter could read. I didn't test voltage output of the other sensor before I replaced it. After all this was done, I though about this question and the question of whether a DVM could read the sensor over the teeth, and I realized that it is possible to test the output of the flywheel motion sensor. You just have to put it in the position sensor hole and check it there. They're the same part, so you can put them in either postion.

Still enjoying the ride.

John