Misfires at idle, dies when the throttle is opened

[gull]

My '89 just started running like garbage, so bad it's pretty much undrivable. It misfires badly at idle, and tries to die when I open the throttle. If I'm slow and patient I can nurse the revs up, but it has no power.

Here's what I've done so far:
- Checked for vacuum leaks, replacing any suspect hose
- Cleaned the airflow sensor and recentered the plate
- Replaced the distributor cap and rotor, which were pretty worn out

It was very sudden. The car went from running pretty well, with the occasional misfire or bit of hesitation if I opened the throttle quickly, to barely running at all.

One odd thing I noticed, that may be unrelated: Normally, with the engine running, I can hear a humming or buzzing coming from the driver's side of the engine bay, in the general vicinity of the fuel distributor. That sound is absent now.

[Briano1234]

Not a Cis guy, but I would suspect a wonky fuel supply or timing...

[gull]

I'm leaning towards thinking it's a fuel delivery problem, but I'm not sure yet. I filled up the tank just before this happened, so it could be bad fuel, too, but the car behaved fine for the ten miles or so I drove it home after filling it. It wasn't until the next time I tried to drive it, a few days later, that this came up.

My gut says the total lack of throttle response suggests it's going way, way lean, for some unknown reason.

[gull]

After a bit more reading online (my manual doesn't cover the Lambda part of CIS well at all) it looks like the missing noise is probably the frequency valve. It's getting dark out now, but tomorrow I figure I'll get out my DVM and start troubleshooting that part of the system.

[Calimus]

Frequency valve sounds like the culprit as that is what makes the buzzine sound. Not overly rare that they crap out, but it doesn't happen very often. However, when it does, fuel delivery goes right to the crapper.

[PDX Cabby]

As mentioned, yes, the frequency valve should be buzzing. If its not, then there's a problem.

Before you go replacing it, I'd suggest checking over all the grounds first. Make sure the FV is plugged in. Make sure the CPR is plugged in. Check all grounds, especially those around the CSV area.

[gull]

Will do. Where is the Lambda control unit actually located?

[tolusina]

..........One odd thing I noticed, that may be unrelated: Normally, with the engine running, I can hear a humming or buzzing coming from the driver's side of the engine bay, in the general vicinity of the fuel distributor. That sound is absent now.

Go to the frequency valve. Check for power on the orange wire with the car running. It's probably not. If so, the CIS power supply relay is burned or full of water. Bentley shows it in location 15.
http://dana.ucc.nau.edu/~kaw5/electrical.htm#Relays

If the freq valve has power, the Lambda Control likely does too, they are both powered from the same relay. Hmm... separate contacts though, could still be the relay.

[gull]

After waiting all morning for it to stop raining, I got out my DVM and went out to the car. It then proceeded to run perfectly, demonstrating the wicked sense of humor all VWs are known for. The basic diagnosis is correct -- I was able to reproduce all the symptoms by unplugging the frequency valve. Until it acts up again, though, I won't know which component in that system is the culprit.

[tolusina]

....... I was able to reproduce all the symptoms by unplugging the frequency valve. .....

Find and pull the Jetronic power supply relay. Un-crimp the cover, open it and look at the contacts. I think you'll find them regular or extra crispy.

Look closely at the circuit board of the relay, light and magnifier if you need, cold solder joints are real easay and cheap to repair.

It's not your standard relay, dang, you can't just swap with another location. It has separate power paths and contacts, one for the Jetronic ECU, the other for the Frequency Valve. Pin numbers are standard 85, 86, 30 and 87 plus an 87b. Sorry, I have no idea it's current rating.

It is powered off the same pin of the fuel pump relay that powers the fuel pump etc.. Could be an intermittent connection inside the fuse/relay panel, it's certainly not the fuel pump relay since the car does stil run.

Make a jumper, carry it for the next occurrence. You'll want to bridge 30 to both 87 and 87b.

The ground connection for the relay is in the dashboard wiring harness, that's a lot of help to know, huh?

BUT!! The ground connection for the Jetronic ECU is over on the right of the intake manifold at the cold start valve, those sometimes get wonky, be sure to look and wiggle that one.

I'd check that ground at the CSV, if nothing bad is found there, throw a relay at it. Or wait, and bypass. A firm diagnosis is really comforting if you have the time and patience. (we have Cabbys, we have patients)

[gull]

I actually fixed a broken ground wire at the CSV a few days before this started, but it was a ground for one of the idle control valves.

I'll probably put a new relay in, just for peace of mind; it's only $19. Until then, I have a couple of jumpers with male push-on connectors at the ends that I'll throw in the glovebox. One of these actually has a long history; I originally made it to use for reading trouble codes on a '90 Ford Econoline I used to own.

[gull]

Just as a footnote to this: All that work I did while I was trying to figure out the problem wasn't wasted. The car idles and runs much smoother than it did before I started. I can't point to any one item but I suspect the biggest difference is due to replacing the distributor cap and rotor, which were beyond worn out. I've never seen one where the center button was eroded back into the cap before; I'm amazed it ran at all.

[gull]

I have a small request -- could someone who has a CIS-Lamba Cabriolet reach under the dash and carefully(!) feel the Lambda relay after the car has been running for about 30 minutes? Both the new and old relays run very hot to the touch, and I'm wondering if that's normal or if I have excessive current being drawn in this circuit. This relay is the metal-can one directly above the seatbelt warning relay, which is easy to find as it's the only one that has a speaker on it.

[gull]

Hmm, never mind. I bench-tested the old relay and it seems the coil alone produces that much heat, with no current through the contacts at all.

I opened up the old relay and examined it. The problem seems to be a mechanical one, maybe even a manufacturing error. Here's a shot of the relay with the cover off:


Notice how the contact on the left side is misaligned. Here's a close-up, with the coil energized:


Notice how the misaligned contact barely closes, and only a fraction of its surface area is actually being used to carry current.

Normally this is where I'd rant about cheap Brazillian/Chinese/etc. parts, but this relay is the original one and its can is marked 'W. Germany.' So I guess it just goes to show you can get a bad part from anywhere.

[Briano1234]

It is not unusual for a relay contact point to become misaligned, you should try working on a 50's pin ball machine...

Repetative action heat of the relay and current going thru all tend to induce that malfunction. I would just manually re-align it, clean the contact points and put it on a shelf for a rainy day.....