southbay

That’s good; one more item you can check off.

CAVU

OP,

Have you considered giving Tony Callas a call at https://www.callasrennsport.com/ and making an appointment for him to have a try at it? This challenge is well within his wheelhouse of experience.

I acknowledge that this goes against our collective desire to solve the problem. I once had something similar with a volvo. Dead of winter pacific northwest. could not for the life of me isolate the issue using my trusty Haynes manual. Seven days later, towed it to a local independent who called me before lunch and said the car was ready. K-Jetronic FI. Cold start injector failed.

Zygrene

Good suggestion! I'll give him a call. When I dropped my car off at BR Racing in December I assumed they'd be able to figure it out. Thousands of dollars later and I have a very slick custom Motorsport AOS installation but all the same issues as before.



UPDATE: Something very odd happened yesterday. I had been driving with the solid CEL for a few days and all of a sudden the idle began to hunt around 1200 rpm, 400 higher than normal. The CEL then started to flash and I pulled these codes:
P0300, P0301, P0303, P0305 (the usual cyl 1/3/5 misfire codes)
P2177, P2179 (bank 1 and 2 oxygen sensing adaptation FRAU lower load range) - long term fuel trim at cruising rpm
P2187, P2189 (bank 1 and 2 system too lean at idle) - long term fuel trim at idle rpm

Engine then started to run extremely roughly at all rpms. I cleared the fault codes which smoothed it out just enough for me to limp home. I unplugged the new MAF and ran the car - the idle went back to normal. Plugged in an old MAF and the idle went high again - same thing with the new MAF. Then I took apart the throttle body, checked fitment of intake plenum couplers and intake hose, cleaned the new MAF, and disconnected the battery for 5 min to reset the ECU. The car now runs "normally" after 20 minutes of driving.

Here's the weird part: Durametric log shows engine roughness values are now very close to 0 even on cylinders 1/3/5. The values have never been this low, and the engine is barely misfiring. The problem cylinders 1/3/5 are actually running AS SMOOTH or SMOOTHER than the others now. However, long term fuel trims at idle and cruising rpm (RKAT/FRAU) for both banks increased steadily during a 10 minute drive. RKAT ended at 1.17 (should be 0) and FRAU at 1.03 (should be 1.00). So the engine is definitely running lean and adding fuel. I don't think it's MAF related because I replaced it recently and it's reading a healthy 15 kg/hr at idle.

1) How is it possible that the rough cylinders suddenly became so smooth? This makes me think the misfire is not caused by a mechanical problem, but an electrical one. Perhaps the DME is seeing bad data.
2) What can I deduce from the fact that the engine is running lean? It has brand new fuel injectors, coils, plugs.

Zygrene

What are the chances I need new O2 sensors? Readings from Durametric:

O2 sensor voltage behind c.c., bank 1: 0.01 - 0.90V
O2 sensor voltage behind c.c., bank 2: 0.01 - 0.90V
Oxygen sensor (LSF) voltage ah. of cat. c. 1: -50V
Oxygen sensor (LSF) voltage ah. of cat. c. 2: -50V

Lots of fluctuation for the behind-bank O2 sensors. Which ones are pre-cat and which are post-cat? Behind/After bank makes it confusing...

CAVU

.ah. = ahead of cat = exhaust before it goes into the cat
behind c.c. = exhaust AFTER going through the cat.

Samething is happening bilateral, both sides. Both sides' values are the same sooooo....The ECU might still be running off its baseline map with some adjustments. Not sure how many drive and warmup cycles your ECU needs, nor do I know what those time, temperature details are, before the map starts to settle into a "smarter" more fully populated map. You may just need to drive it a few more times to normal oil temperature.?

Please share with us what you can from your call with Tony.

DDE997

From What I have seen from the experts, the long fuel trim should be 1.0 to 1.05. My 997 runs a long fuel trim is at 1.02 both banks, and my car is healthy. When you disconnect the ecu, you cleared the adaptions, and the ecu will re learn the engine in approx 100 miles.
keep a eye on the o2 sensors.

Zygrene

@CAVU I'm driving down to LA tomorrow morning and dropping my car off with Tony. I'm staying nearby for an entire week, which I'm hoping is enough time for them to at least pinpoint the issue. Thanks again for the suggestion.

CAVU

I am guessing that it will be one significant contributor and 2 minor contributors, all straightforward for him to address.

Zygrene

I hope you're right! This car has been anything but straightforward for the past 5 months.

southbay

Good luck man; I’m hoping for the best. Nothing but sunny weather here all week.

Petza914

Tony at Callas-Rennsport? No one better for this kind of stuff. Good luck and I'm very interested to see what he finds.

CAVU

My pocketbook would be a lot lighter and my car even more fun to drive after a visit with Tony. Never met the man but I enjoy reading and viewing his technical experience and his background is extensive.

Zygrene

After 2+ days of diagnostics at Callas, here is what they found:

DIAGNOSIS
Accessed crank sensor signal wire from twist lock connector in engine compartment and connected oscilloscope to monitor crankshaft sensor signal for quality and irregularities during various speeds/loads. Found individual speed and reference pulses could be clearly seen and would change proportionate to RPM (signal quality good) but momentary changes in amplitude during revolution are likely indicative of a flywheel issue: either too much torsional or axial play affecting crankshaft acceleration.

They also checked output from the alternator to rule it out as a possible cause of the crankshaft signal oscillations. They seem somewhat confident that the flywheel is the issue. I authorized them to remove the transmission and inspect the flywheel and replace it if needed. Big $$$$ and I'm sure I could find a shop that will do it for cheaper, but at this point I just want it resolved ASAP. At least if it does get fixed, the car will have a BUNCH of new parts already installed. Motorsport AOS, deep sump, plugs/coils, injectors, crank/cam sensors, MAF sensor, intake plenum hoses, etc.

I have to say - Callas has done a SUPER thorough job and their daily reports have been the best I've ever seen, complete with pics and videos. Really hoping this is the last push.

Hootsama

Good to hear you're on the right path, finally. Also admire the optimistic outlook. You've been through the ringer, and I'm looking forward to much happier posts from you in the future.

Zygrene

Some photos Callas provided below.

Here's the flywheel:


Irregularities in signal between reference marks are likely signs of failing flywheel.




Known good crank signal from similar vehicle.



They also uploaded this photo of the Motorsport AOS that BR Racing custom installed. Looks like the clamp is not all the way on? Also seeing some scuff marks. Will call Callas tomorrow.