Interesting stall issue

[NorthernThrux]

So, as I understand the situation, you are slowing to a stop, the revs are dropping to idle, you push in the clutch and the engine dies. No fuss, it just dies. Yet it will restart as though nothing is wrong. So you had two instances in a row, and on the third clutch action it did not repeat. I find it hard to imagine an ignition problem there so it seems to indicate the fuel was shut off. The rarity of this problem occurring certainly makes the diagnosis difficult.

Correct. And it has happened twice (2 times in a row) in 2019 on the way to the Whaletail swap meet and 2 times in 2020 (this past Sunday). Both times after long drives mostly in 5th gear. I've probably come to a stop a thousand times over the two years. So whatever it is, it's very rare and it doesn't point to any consistent problem. When it happened in 2019 I had an original Bosch speed transducer. Now I have the replacement new one with the solid state logic in it. Brand new microswitch as well and the engine oscillates as it should when you test it. Stuck my LM-2 in the exhaust yesterday evening. 11.1 AFR at cold start and 13.2 after idling for 10 minutes. Same as when it was set up on the dyno, so nothing has changed. Car has run flawlessly the rest of the time and idles rock steady at 875-900 rpm when warmed up and with the hand throttle all the way down.

My hunch is that the fuel shut off solenoid which gets activated as you roll to a stop with the foot off the throttle is getting stuck or hesitating and that's what leads to this occasional problem. I've done many other long drives and not had the problem, so it's going to be hard to chase down.

[xpensivewino]

Based solely on your description of the way this occurs, my experience would tend to think vacuum loss. Generally with a fuel issue you will not get instantaneous cut off. I had a vacuum leak on one of my cars and it behaved exactly as you described. Have you noticed any changes with altitude??

[NorthernThrux]

I would tend to agree about fuel, except the one thing that can absolutely and positively cut off the fuel is that overrun MFI solenoid. Still, vacuum is a possibility too except how come it fixes itself? And if I had a general vacuum problem, I'd see the idle go all over the place and it is rock steady. Still, I do have a couple of tatty vacuum lines as I mentioned, and on my next order from Belmetric or whomever, I'll get some vacuum hose as well to replace the lines from the stacks to the vacuum advance. It could be the leak opens up after a long period of suction (like cruising) and then closes up when the engine stops and the vacuum ceases, so when you restart it is fine.

One other clue. Both times it occurred when the temperatures were relatively crisp. Like below 10C. Not that anything in the engine compartment would be that temperature after a long drive.

Not sure where else I would look for a vacuum leak given the entire motor has been apart in between the two years that this ghost has been around.

[BURN-BROS]

Could be an intermittent electrical gremlin. Do a "wiggle" test while the engine is running.

Move the engine harness and the wires at the fuses up front to see if you can get it to generate the complaint. Pay close attention to the 14pin connector in the engine bay as it could get corrosion.

[gled49]

Unscrew micro switch to eliminate speed switch. I’ve had as many cars backfire with the switch working as not. Mine has been disconnected forever and no backfiring with my 2.4 or my 2.7. I’ll bet that’s your problem. G

[steve shea]

I think he suffers from the same thing I do. If the original system worked then it should still work; given proper trouble shooting and maintenance on each aspect of the system. I had a similar problem but I live at 2000m so idle air, contemporary fuel and other factors necessitated some creative tuning. I used Mayo's seat of the pants method and voila, it works as it should.

[NorthernThrux]

It’s going to be hard to troubleshoot something that happens once or twice a year. I did the jiggle test on the ignition switch and everything I could in the engine compartment and the front fuse boxes. Nothing. Not surprising. The contacts are all cleaned and everything is pristine.

Microswitch is new and works fine too. Speed switch (both ones I have, made 47 years apart) works fine. I used the hand throttle to set the rpm to 2000 and then pushed the microswitch and the engine speed cycles as it should. 15 minutes of this and no obvious glitch although on one cycle it may have gone lower in rpm before come back on. Hard to know as I didn’t have the timing light hooked up. That could point to a glitchy solenoid on the mfi. I’ll have to investigate further. In daily driving since I first posted this, no further stalls. So no real complaints. Will disconnect the microswitch next year and see how it goes. Want to see if I can recreate it before the weather turns.

[Boulder Bob]

My money would be on the anti-backfire system , probably the rpm transducer , certainly seems suspicious . Maybe disconnect the MFI pump shut-off solenoid and and see if the problem reoccurs .

Never heard the term "anti-backfire system" - what exactly are you referring to?

[NorthernThrux]

Never heard the term "anti-backfire system" - what exactly are you referring to?

It’s the system that involves the microswitch, the speed relay in the rear electrical console and the cutoff solenoid at the end of the mfi pump.

On the MFI cars. when you lift off the throttle at higher rpm, the throttle plates slam shut, but the pump keeps spraying fuel at a high rate until the pump catches up and provides only idle fuel flow. That leads to a super rich condition with lots of unburnt fuel and often a backfire or two. To counter this, Porsche installed a circuit that does the following. It has an speed relay (technically engine speed relay but Bosch calls it a speed relay) that always puts out 12V whenever the rpm is between 1300 and 1700 rpm. The speed relay gets its rpm signal from the black/purple wire on the dizzie. This 12V signal from the speed relay is applied to a solenoid on the mfi pump that pushes the rack to its fuel cutoff point. Now obviously, you don’t want the solenoid active when you are on the throttle, so Porsche installed a microswitch that feeds the speed switch 12V signal to the mfi solenoid ONLY when the throttle plates close.

Net result? When you let off the throttle above 1700 rpm, the MFI pump still sends fuel (avoiding a massive lurch) until 1700 rpm where the MFI rack is pushed by the activated solenoid and the fuel cuts off. The fuel comes back on below 1300 rpm so you don’t stall. At that point you are almost at idle and you just have idle air-fuel ratios, so no backfire.

The circuit can be cantankerous. Hence people pointing fingers at it. Speed relays fail due to vibration and corrosion of the leads. Mine has been repaired by Bob Ashlock. Bosch also sells a modern replacement. It uses an integrated circuit. I am using it at the moment. My pump is rebuilt, but it is the original solenoid. Microswitches fail. Mine is new. In failure modes, you just get backfires. But if the solenoid sticks, you get a cutout of fuel.

In my case, any intermittent electrical glitch in the car should also occassionally show up when not at idle. The car has mever missed a beat in 1700 miles. So I discount that. It has to be something that only occurs at idle. Hence my focus on the speed switch circuit, but the solenoid in particular.

Ravi

[steve shea]

I addition to everything Ravi posted you can further eliminate occasional backfire by lifting completely allowing the system to do it's thing. CMA addresses this in a roundabout way. Lazy foot, my term, will still feed fuel. You just need good dorsiflexion in your left ankle and hours behind the wheel to get this to be second nature.