Bilateral ignition failure--Anyone seen this before?

Okay, so this a continuation of a previous post, but after going through my whole CIS system without finding any serious discrepancies, I was finally able to replicate the original fault, twice over now, and have come to the realization that it's not a fuel problem at all. It's an *electrical problem* (Yes, I know the old adage that most fuel delivery problems are actually spark-delivery problems. True enough in this case.)

Basically, with a hot engine, when I go to idle or low-load (coasting down hill, for instance) I suddenly lose both front and rear bank ignition. The car will not start (no spark) until it is thoroughly cool again. Then it starts/runs just fine . . . until the motor gets up to full operating temp (based on oil temp).

The failure seems to be preceded by small, but noticeable hang-up in the engine RPM when backing down off the throttle--like you might expect if the ignition timing stayed fixed on the fast side of the curve. (Of note, I've replaced the throttle cable and rebuilt/lubed the throttle bottle linkages recently, so the source of that hang-up is probably not mechanical. It ain't my floormats either.

Immediately after the ignition failure, I get a nice, loud backfire--as you would expect with the engine still drawing fuel as it winds down.

Logically, whatever is happening when my car gets hot must relate to something that the front/rear bank ignition systems have in common--which isn't very much, actually. They share a power source and ground; and they share a signal from the engine speed sensor on the lower-front side of the clutch housing. That's about it, as far as I know.

What I've done so far:

--Examined, cleaned, re-seated all connections between the sensors, the ignition boxes and the coils. (Found that the big multi-connector by the oil cooler was soaked in oil from a hose that ruptured last summer)

--Relocated sensor wires so they are not touching the engine, oil or coolant lines.

I was hoping that those actions would have done the trick, but the fault re-occurred today. I am now at a loss on how to trouble-shoot, short of just throwing new ignition system parts at the car until she decides to cooperate.

Any suggestions?






They even have separate crank position sensors.

 

I *believe* they were in that range when I tested them, but I will double-check. Thanks.

 

I have the Birdman kit. Installed almost seven years ago. I had a glance at the fuse that the coils are powered through (and the tail lights, apparently) and nothing looked off, but short of obvious corrosion or a loose connection, what should I be checking for?

 

Additional detail: If I run the car to temp but not quite enough to trigger the failure, then shut it down, I have a brief window during which I can restart the car. After about a minute or two, heat-soak seems to take effect, and whatever is shorting out my ignition system takes out both coils.

 

Okay, so I re-checked the crank position sensors with my multi-meter. The motor is still slightly warm-to-the-touch from running a few hours ago, but I didn't check to see whether it would start (too late and don't want to wake everyone).

The front bank sensor reads 754 Ohms. The rear bank reads infinite.

Earlier, I tested the engine speed sensor (cold) and got ~650 Ohms. I didn't re-check it just now because that would require getting under the car.

I should mention that the rear sensor lead had previously been resting against the oil return hose until I just relocated both a tad, so heat-transference might have done some damage. Still, both banks have their own sensor, so why would one failing bring both banks down? Do the ignition modules cross-compare or something?

 

That was my first assumption--despite the backfire--and I spent several days going through my CIS system as a consequence. Even bought a CIS test kit for the purpose. The accumulator and check valve are new with 6 months, and seem to be doing their thing. Finally admitting it might be ignition related, I got out my, in-line spark tester. Low-and-behold, I have good spark on cold start, and then nothing on either bank while trying to start the engine hot. Not the result I expected to see, but undeniable.

 

Let me whack myself in the forehead first ...

It turns out, what I assumed was the rear (1-4) bank TDC sensor is actually the engine speed sensor. And *that* is the one which has had a hard life pressed against the oil return hose and showed infinite resistance last night.

I'll give myself a little slack for getting that wrong. I didn't get my protractor out and measure to make sure what I thought were the front/rear sensor were actually 90 degrees apart--and without removing the clutch cover like in the picture you posted, it's not intuitively obvious just looking at it. Thanks!

Now, lets see where I can get one of those... Standard Bosch part--shouldn't be too hard.

 

Follow up:

Okay, so it is a Magnetti Marelli sensor, same part number as the TDC sensors, and it is DEFINITELY my smoking gun.

I pulled it out this morning and did an quick test. I hooked the sensor's leads to my multimeter, set to resistance, then applied my heat gun to the sensor head, while monitoring with my laser temperature meter.

At 61F, resistance was at 686 Ohms -- within specifications
As soon as heat was applied, the resistance started moving upwards
At 80F, resistance was up to ~1000 Ohms -- out of spec
At ~81F, resistance went infinite. No worky-worky.

I had to put the damn thing in the refrigerator to cool it enough to get it work again, so I could repeat the test. Got the same results next time around.

In retrospect, I think this has been coming on for some time, with the temperature threshold of failure gradually decreasing. The final nail in the coffin was a day spent hooning the car around an autocross circuit.

Anyway, I am now much smarter on both the ignition system AND the CIS system (due to the week I spent needlessly trouble-shooting that side) and I have a new sensor on order.

Thanks for the support from this forum. This is, without a doubt, the friendliest, most helpful online car community I've ever been involved in.