Injectors not firing, intermittent

[Sparky1947]

My 1988 Vanagon GL dies randomly. I've confirmed that while it's dead the injectors do not fire. I've swapped out the ECU with a known good ECU with no effect. I've renewed the ground connection next to the ECU with no effect. Has anyone else had that problem and what was the specific solution in your case? I'm about to embark on a detailed investigation of the wiring harness and all the grounds I can find but I thought if someone else had had the experience it may point me in the right direction sooner. Thanks!
Jim (Sparky1947)

[VicVan]

Hi

How did you test the injectors? From what I understand from your message, youère saying that when the engine is not running, the injectors do not fire. This seems pretty normal!

When does the engine die ? Cold, warm ? How does it happen ? slowly dying, spluttering, or sudden stop wé no lights on the dash?
_________________
'90 Little Blue Truck, 2WD auto, FAS GenV 2.0 NA (AVH)

[mikemtnbike]

I'm guessing that what you're saying is that the one constant variable you've been able to find when your van randomly dies is the fuel injectors aren't firing when tested per Bentley. If that's so...

You've probably done this, but I'd try swapping out the ignition switch, could be intermittently failing.

When the van has died, does the fuel pump prime when key on before cranking?

If it's happening when hot, relays could be quitting, then starting to work when cooling off.

Have you tested the injectors at the ECU plug per Bently/ pro training manual?
_________________
1991 Vanagon GL 2.1 AT Westfauxlia. "Frankie" Totaled https://www.thesamba.com/vw/forum/viewtopic.php?t=764510&highlight=carnage
1995 Eurovan Camper "Marzivan"
2020 GTI SE manual

[DanHoug]

i assume you are testing while cranking the engine under the no-start condition. but are you observing the injectors "not firing" by watching for spray or thru noid testing?

if not spraying, it could be a fuel pressure issue, easy suspect is the right hand #53 relay in the black box on the left side of the engine.. if not lighting a noid, could be as simple as the right hand #53 relay in same box.

critical questions:
- do you have spark during the no-start?
- do you have fuel pressure?
_________________
-dan
60% of what you find on the internet is wrong, including this post.

'87 Westy & '89 Westy both 2.1 4spd

Past projects can be found at--
www.thefixitworkshop.com

[Sparky1947]

While it is in the non-running state if I disconnect an injector and attach an LED tester to the plug in place of the injector and crank the engine, the tester does not blink. When it is in the running condition, which is totally random, I can disconnect an injector, attach the LED tester to the connector and it will happily blink while running on only three injectors. The car may be dead and I'll come back a day later and it will start right up. Or I may come back a day later and it will still not start. This has been going on for months and I see no correlation to temperature. It ran fine for a three week period recently when the temp was often 90 F. While it's been running I've banged the ECU on the car rug, I've banged on the Mass Air Sensor, I've banged on the relay next to the fuel pump relay, and I've banged on the connectors in the box in the left front of the engine compartment and it did not cause it to die. Did I miss anything? I've done the fuel pump volume test and actually replaced the fuel pressure regulator before I figured out that it was the car dies because the injectors are not firing. I did a total tune up last year and when it runs, it runs great. In the non-running state it sparks just fine while cranking the engine.
Jim

[Sparky1947]

I have not "tested the injectors at the ECU plug per the Bentley/pro training manual". I have the green Bentley manual and I didn't come across that test. I could find the right wire, run a pin through it, and look at it with my oscilloscope or hook the LED tester to it but I have not gotten that far in my testing. That's the kind of detail I will have to address if no one has had the same experience.
Jim

[Sparky1947]

I have confirmed that the ECU us still getting power even when it won't run, so my guess is that the problem could not be the ignition switch. Could I be wrong?
Jim

[Sparky1947]

When it dies, it dies completely, just as you would expect when the fuel supply is instantly cut off. All the lights on the dash are normal/ I see no electrical issues other than the injectors failing to fire.
Jim

[crazyvwvanman]

There are a lot of things that I have seen cause the injectors not to fire. Most of them are wiring related. Don't think that swapping parts is the easy answer. Testing for the missing voltage is the answer. With an 88 injection harness the injectors get power from the fuel pump relay, so ONLY when the FP relay is turned on, by the ECU seeing that the engine is turning. For safety reasons the ECU only turns on the FP relay when it thinks the engine is turning. It also does so for a second or 2 each time the key is turned on. After that it wants to think the engine is turning before it will turn on the FP relay (and power the injectors in 88 vans). Firing the injectors is a completely different thing from powering them. The ECU fires the injectors by providing a ground pulse to the other pin of the already powered injector plug. It does that when it sees hall pulses from the distributor.

Mark

[mikemtnbike]

[Sparky1947]

Thanks, Mark, I agree that your explanations are thorough! The fuel pump relay appears to work as advertised. While the injectors don't fire when it's dead I do get fuel pressure. That does give me an idea, though. I should drain off the residual pressure and make sure that it pumps back up even though the injectors are not firing.
Jim

[Sparky1947]

Yes, I've always had spark, even when it won't start.
Jim

[Sparky1947]

Mark (or anyone),
I turned on the ignition, bled off the residual fuel pressure, cranked it a bit, then rechecked the pressure and it had pumped up, which means that the fuel pump was activated while cranking. Next, with the ignition turned on I checked one of the injector connectors and one of the two leads has 12V on it. So, according to my wiring diagram (page 97.124 of my green Bentley manua, which goes up to 1991) that leaves only the wiring from the injectors to pin 12 of the ECU to check. Sound right?
Jim

[crazyvwvanman]

Yes, ECU pin 12 fires the 4 injectors by providing variable length ground pulses. So check for continuity on that harness pin to each injector plug. Unplug all 4 injector plugs first while you test this.

Measuring the 12v at one pin at each injector doesn't completely rule out a supply voltage problem to the injectors. It could be present in a static situation but with such a weak connection that no current flows when the ECU pluses the ground.

Mark

[crazyvwvanman]

Also, be warned that there are some errors in the relay wiring on that page. That is why it is so confusing to try to follow. Also while it says that diagram is for 88 in fact it seems some 88 got the 87 version of the harness. In any case some vans have had harness swaps over the years so it may not be original and could be from another year.

Mark

[onslowb]

I just went through something similar. Van wouldn't start 90% of the time because injectors were not firing. My vanagon syndrome plug had broken and signal wasn't getting to the ecu from the AFM most of the time. Removed the harness and the problem went away....almost. I've noticed that if I fiddle with my ignition key a bit I can cause the no start problem on purpose with the same "injectors not firing" problem. So it could be the ignition switch too.
_________________
1987 Weekender 2.4GW manual
1970 Vw Bug SP 1600

[Sparky1947]

Good point about the 12V, Mark. I'll run a 10 ohm resistor to ground to make sure it can supply some current. Also, thanks, onslowb, for your comments, although I don't know what the "Vanagon Syndrome Plug" is. Also, what does AFM stand for? I hate to sound like a neophyte, but they've never come up as problems for me before. Maybe that's good.
Jim

[mikemtnbike]

Air flow meter. Here's a version of the "vanagon syndrome plug." The factory one does not require splicing but as usual GW claims their method is better.
http://www.gowesty.com/product/fuel-injection/23990/gowesty-signal-filter-for-air-flow-meter-?v=

Here's the testing procedure for FI wiring from protraining manual. Also, I realized my earlier link was not the vanagon specific manual, my fault for not checking google results. here's the vanagaon specific one.

http://www.vanagonauts.com/files/Vanagon%20Digifant%20Fuel%20Injection%20System.pdf
_________________
1991 Vanagon GL 2.1 AT Westfauxlia. "Frankie" Totaled https://www.thesamba.com/vw/forum/viewtopic.php?t=764510&highlight=carnage
1995 Eurovan Camper "Marzivan"
2020 GTI SE manual

[Sparky1947]

Thanks, Mike. Wow! That Pro manual might come in pretty handy. So it's the Air Flow Meter and electrical plug that goes to it. Got it. I've looked into that and I don't think my AFM has had a problem so far. We will see.

Mark, I loaded the 12V pin on one of the injector plugs with a 10 ohm resistor to ground (approx one amp current draw) and the voltage dropped over 3 volts! That sounds fishy , so I'll be looking into that. I've got enough things to check into for a little while now.
Jim

[Sparky1947]

Guys,
I finally figured it out. Thanks to everyone! As it turned out, knowing that the Hall Sensor initiated both the spark plugs' firing and the injectors' firing was fairly key. I hooked up my oscilliscope (TEK 2465) to both pins of one of the injectors, one at a time, while a friend was operating the starter, cranking the engine, and I found out that the signal from pin 12 of the ECU was quite intermittent, never staying around long enough for it to actually start. One thing that was also happening, however, that up to that time was unexplained, was that while cranking sometimes it seemed to "backfire", stopping the starter cold, like it was firing at the wrong time! That's when I zeroed in on the Hall Sensor. Then I hooked up my oscilliscope to the green signal wire of the Hall sensor by sticking a pin through the wire and the signal was beautiful and steady. It has a fairly high duty cycle. It is not a narrow pulse. Obviously it is the leading edge of the pulse that matters, not the duration. At that point I turned the key and it started right up! So, here's what was going on: As long as I've owned it, which is now 23 years, the connector for the hall sensor has been broken off from the distributor. It just sort of laid there with the three wires coming out through the opening all that time I had had no trouble, at least not enough trouble to warrant looking into it, until recently. It turns out that at the rear of the connector that was just hanging there the three connector contacts for the wires are exposed. After all this time, the contact for the red (+) wire had made its way over to randomly short out against the distributor case. That was killing both the spark and the injectors. However, in the case of the spark, imagine that at the time the hall sensor fires the red wire is shorted, but a short time later while the hall pulse was still active, the short went away, allowing the pulse to rise, triggering the spark, but with the timing way off from the intended time, causing what would have been a backfire. That's what was bucking the starter. The reason it started as soon as I hooked up the oscilliscope is because I had moved the connector, eliminating the random short! I'm now holding the hall connector in place on the side of the distributor with two wires wrapped around the connector and then around the body of the distributor. I'm going to look into replacing it with a new one, but for now, problem solved! Thanks again for the help!
Jim