FlyingShawn

Just to save everyone some scrolling, here's the post he's referring to:
Hmmm, I must be missing something here; I'm having trouble following your train of thought on this one @dcdrechsel... The theory behind testing the CPS, at least as I understood it, was something like "minor damage to the flywheel might cause intermittent glitches/distortions in the signal from the CPS (meaning the damaged tooth is sometimes seen by the sensor and sometimes not), causing the DME to intermittently miss its cue that it's time to signal the fuel injector." No fuel injector signal=>misfire and no "rebound voltage" on the back side of the injector signal (generating the "injector circuit" code because the DME notices the voltage to test the circuit is absent).

If the test shows "the teeth are there and the CPS is working" as you say, and if we were able to time-correlate missing injector pulses with "normal" CPS signals, how could harmonics or other flywheel issues interfere with the DME sending/failing to send injector signals? How would those issues be communicated to the DME if not through the CPS?

dcdrechsel

I am by no means an expert but If the flywheel is oscillating - the dme predictable pattern is disturbed which could cause strange faults .I would ask Jake for his thoughts …………...Further I believe the IMS was upgraded before your purchase which makes that work suspect .

Porschetech3

Your Indy seems to be very competent and thorough, and is using an oscilloscope which is invaluable in these situations. I don't know if he has Piwis, but it gives a lot more information and actual values than Durametric or other equipment. Most dealers also have a "breakout box" that goes between the DME and wiring harness and allows quick and easy testing of resistance,voltage and scope tests for anything coming or going from the DME. Dealers also have access to the "Porsche Hotline" of engineers in the Atlanta Center that love challenges like this, and have a data base that has seen everything. All problems like this end up at a dealership and most have contacted the "hotline" and solutions are entered into the data base. As a last ditch solution there are " Regional Service Field Engineers" that can get involved and instruct the Technicians on testing and troubleshooting, and sometimes will make a special visit to the dealership in person (or if there is no rush, they come around every so often anyway to check on things). Before I retired Jimmy Hays was our Regional guy, very sharp, I don't know if he covered PA or if he is even still with Porsche, but all of them are good.

The "fresh set of eyes" comment might be what is needed. Many times these kinds of problems turn out to be something really simple, that is causing strange symptoms/codes, and a fresh eye or fresh approach can spot it quickly/easily. Like I said earlier, at this point I would pull the starter to get a good look at the trigger teeth of the flywheel, and fashion an indicator wire and rotate the flywheel by hand to check for irregularities/damage of the flywheel trigger teeth. I know the scope pattern looked good, but at this point nothing should be assumed.

Ahsai

If you reset the DME, does the error code pop up and the engine misfires immediately or the engine runs fine for sometime but then turned into this failure mode?

FlyingShawn

It's good to know they have those resources available. When I originally ran your starter idea by my indy he agreed it'd be a lot easier than pulling out the whole flywheel but then added that pulling the starter wasn't all that quick/easy either (which is when he came up with the idea of scoping the CPS). Aside from your idea of checking the alternator rectifier diodes, my current leading two options are having the indy pull the starter as you said or just giving up now and taking it to Porsche along with the now VERY long list of things we've tried and learned so we don't waste time/money recovering the same ground. Either way it'll be expensive. :/

FlyingShawn

Right away. Or rather: identical behavior to an un-reset DME, which varies depending on which one is installed (ECU A needs to warm up a bit and then misfires on #5, ECUs B-D start misfiring on #1 before you even get it out of the parking lot).

Ahsai

One more question. Has the mechanic ever scoped both the ignition coil signal and injector signal (of the misfiring cylinder) at the same time? I want to know when an injector pulse is missing whether the ignition coil pulse is also missing. If so, it seems the DME is trying to shut down that specific cylinder. If not, it is even more weird.

klr10

Have you tried measuring the supply voltage to the DME ECU? Also the 12V supply for all the injectors and coils?

FlyingShawn

That would have been a good think to scope, but considering how many of the misfires are single-misses it seems like it'd be odd to try to shut down a cylinder for a single rotation and start it again on the next one. ECUs B-D do eventually shut down the #1 cylinder altogether, but only after a number of intermittent misfires. I don't think ECU A ever went into a full-shutdown mode on #5 in our testing.

I don't have it specifically in my notes, but I'm pretty sure we've measured those for at least the ECU and the injectors (not as confident about the coils).

FlyingShawn

So... I caved in today and decided to take the car over to the local Porsche dealer to have them do diagnostics on it. Aside from the couple of out-of-the-box suggestions that have been offered on here over the last few days, we're just totally out of ideas at this point aside from yet another ECU. Before I took it over there I wrote up a list of everything we've done and learned so far to get them up to speed without having to cover the same ground or chase some of the same dead-ends like the injector circuits (it ended up being 4 pages long!).

I'm still pretty-confident that the problem is external to the DME (some sort of ripple or bad input that's confusing/glitching otherwise-fine computers), so my single biggest concern is that they're going to come back and say "it needs a new ECU" before diving deeply-enough into the problem to be confident of that answer (a $3000 not-returnable/refundable gamble if we try it and the problem persists). While talking with the service manager I told him about how Specialized ECU had given ECU A a clean bill of health and he said "they do good work: I've sent stuff to them before," so hopefully that conversation will help persuade them to look deeper during their testing.

They won't get to the car until later next week, so stay tuned and I'll let you know as soon as I hear anything!

De Jeeper

This is such a great technical thread. It sucks the fix has not been found but for us tech heads its about the ride. Hope the trip to the dealer works out.

Ahsai

My thinking is this. Perhaps not so much about DME trying to shut it down but rather the DME cannot find the right cam position to trigger it.

IF the coil signal is missed together with the injector signal, at least the DME is consistent, which points to NOT a DME problem. It's something that's causing the DME to not fire the cylinder. Since the offending cylinders are 1 and 5 and they are next to each other in the firing order, I wonder if there's some irregularity with the cam reluctor(s) such that the DME has to compensate/learn such thus cannot fire all cylinders but can only fire 5 out of the 6. Some DME gave up on cyl #1 and some on cyl #5, due to the way they learn the irregularity. This theory is totally out there but your case is totally out there anyway.

Another possibility is is it possible the ping sensors play a role?

808Bill

Porschetech3

Any update on this? This thread went so deep it would a shame to not have a conclusion on it !!

FlyingShawn

Thanks for asking! I'm living in limbo: no news one way or the other.

It's now been a full month since I dropped it off at the dealership and was told it'd be a week before they got to it. When I spoke with them about two weeks ago they told me that they'd gotten completely slammed with work right after I dropped the car off, so they'd barely spent any time on it at that point (the person I spoke with thought she might have heard the tech had "an idea," but I never heard anything back about that so I'm assuming it didn't pan out). I haven't heard a peep from them since then, so given what they said I think it's reasonable to suspect they still haven't spent much time on it. It's not fun to be so out of the loop, but I've been trying to not bug them because I'd rather they go slow and get it right than rush an answer to get me off their back and get it wrong. I'm planning to check in with them Tuesday or Wednesday if I haven't heard anything.