toddlamb

Excellent - glad you’re back on the road trouble free!

Drh2

Thanks for the great information and a glimmer of hope. I read your guide however, I am one of the guys who takes it to the dealership for an oil change. Does anyone know of any shops that will do this type of repair in the DFW area?
I spoke with Becks in Scottsdale and they have a repair but it is still $$$.

GTSpure

Hi (:
I'm sorry I did not quite understand what the problem is with your PDK? Do you have a fault?
Do you need to replace PDK oil or clutch oil?
Are you with 997.2 or 991.1?

toddlamb

Hopefully there’s a close option for you but if not we get a lot of cars shipped to us in Atlanta for PDK repairs. Let me know if you need help.

Drh2

Today a.m. -I am getting the error message light Transmission fault for the PDK. "Poss no R gear Drive on Possible". The car does not go into gear. It did this driving home at a stop light, the light came on I proceeded and the transmission shifted to 6 gear, I drove a half mile, pulled in my drive and it would not go into a forward or reverse gear.

The car / error message light has done this once before, but turning off and on basically reset the car and there have been no issues until today. The car has 50,000 miles.

This afternoon p.m. - After sitting an hour, I cranked up and got the same message, After sitting 3 hours the light went off and I could change gears and drive the car. Strange.
The model is a 991.1 with 50k miles

toddlamb

Well that’s frustrating but does sound like distance sensor (guessing without code info based on symptoms). 50k miles isn’t indicative of failure one way or another. We repaired a 4100 mile PDK recently that was out of warranty.

saabin

Sorry to hear this. I have a buddy with the same car as you (991.1 PDK) with 56K miles here in DFW that got the same message and the car would not move forward. He took it to Autobahn in Ft Worth since he normally gets all his service done there.
The total time from tow-in to pickup was about 3 months. Part availability was a big issue from Porsche. He got some goodwill from Porsche but it still cost him close to 13K since the dealer did not want to do any kind of "repair", just a full replacement.

YMMV, this was about 2 years ago. If this happened to me I would look into shipping the car to a shop that can repair it vs. one that is just going to replace the entire unit. No idea if any shop here in DFW has successfully done this repair before.

PV997

Maybe Todd can chime in here but I've never seen an intermittent distance sensor failure, in my experience they just outright fail. It is electrical part so it's possible it could be acting up when getting hot and working when cooling off but I have not seen that pattern (though I don't have anywhere near the hands on experience with these like Todd or Beckers).

Prior to shipping the car anywhere I'd find someone to run the codes. If you know someone with a high end scanning tool (a good Autel, Durametric, etc.) then they can do it or you could take it to the dealer and just pay for a diagnosis. Insist they provide you the full code and description though, not just tell you the PDK failed and needs replacement. Once you have the codes post them here and we can help pin down the problem.

Kuro Neko

... and to preempt the next question and searching the many posts above, with the cheaper Foxwell being able to confirm the appropriate distance sensor codes too:



With the range P173X for the shift rod displacement sensor?
The last digit matching the rod number?

notfastenough

The very first PDK distance sensor we did was a 991 that was intermittent for just a day or two then went toes up.

@Drh2, I wouldn't be taking any out-of-town trips with it.

toddlamb

We have had a few that were intermittent and clearable, but the failures were not far apart. Sometimes it takes just the right conditions/circumstances to trigger an error. Eventually though they always trend towards permanent.

mdvigil

@PV997

Hello,

Sorry to jack this thread. I attempted to PM but your box is full.

I saw the comprehensive thread you wrote on the PDK and was hoping maybe you could share your wisdom with us to diagnose an issue that recently came up.

A brief history. The car has a 4.0-liter built engine from EvoSpec, Dodson clutch, ByDesign XR turbo kit, and a Wavetrac LSD. We completed the installation yesterday. No leaks of any kind. All PDK fluids we drained and refilled per Porsche's recommendation. We completed the PIWIS PDK calibration per Dodson's instruction. The calibration was successful.

We went for an initial drive yesterday, and everything seemed to work well. PDK would shift, although a bit sluggish (it has been about a year since the car was driven, so my reference is fuzzy)? We decided to get on it to ensure the car would not boost past 10 psi since the boost controller was set to the off position, and we wanted to check wastegate operation. We then heard some sounds, and the car stopped moving. We pulled to the side, and the transmission failure park vehicle safely, and no R faults came up on the dash. The car had no gears afterward. The indicator on the dash would show the car could switch gears but no forward or reverse movements of the wheels. We towed the car back to the shop and put it on a lift. Wheels would move with no odd sounds. PIWIS presented the following codes: P1765 gear disengagement lock (no fault system) and P1870 PDK control unit internal fault (no fault symptom). We cleared the codes but still no wheel movement when put into gear. We plan to pull the bottom PDK pan to examine the valve body today.

Any help or suggestions would be greatly appreciated.

Thanks!

-Mikey

PV997

No problem at all posting this here, that's exactly what this thread is for. Apologies for the delay in responding.

We've seen P1870 a few times and it seems to be some sort of fault within the transmission control unit (TCU - the computer within the car that controls the PDK). Unfortunately people aren't real good about following up once things are resolved so I don't know a conclusive fix. One I do recall was fixed by the dealer wiping the TCU and reprogramming it. Unfortunately P1765 isn't in my diagnostic code reference (it's for a 997.2) so I don't have any additional info on that one.

The readout on the dash comes directly from the shifter via the canbus so that would explain why it's showing a correct gear even when the PDK isn't responding. My hunch would be there's something fouled up in the TCU. Did you try connecting to the PIWIS again and rerunning a cal? I doubt it will work but it may give you some additional info when/if the cal fails.

There seems to be some weird mode a TCU can get in when there are major changes to a PDK that causes a failure that we have never conclusively resolved. The TCU has adaptation tables (different from the calibration tables) that are developed over time and match the PDK response to driver's style. I've long suspected that if there is a sudden change to the PDK (such as a new clutch) then those old adaptation tables can be so out of sync with the new parts that it causes the TCU to fault. If so, that would also explain why the PDK seemed sluggish when you first drove it as the adaptation was all wrong. That's the rationale behind wiping the TCU clean and starting over so it can develop new adaptation tables from scratch matched to the new parts.

Please let us know if you get this resolved and what you did as that's how we learn.

mdvigil

I appreciate the reply.

I read up on the posts with a P1870 code however, as you mentioned, whether a resolution is found or not, forum members don't follow up.

We attempted to re-run the calibration. It would start the recalibration process but then kick us out.

The car has not been driven in about a year. What I hear you saying is that the computer learned my driving style before all the changes in the engine and transmission. The PDK TCU threw a fault because the changes did not make sense to the PDK.

How do we go about wiping the TCU? We have access to a PIWIS. Also, what do you make of the fine metal shavings and the clunk sound? My buddy believes that it may be from the synchros trying to engage/disengage and not aligning, similar to grinding gears in a manual transmission since the PDK is essentially an auto/manual gearbox. From your post, it seems the gear part of the PDK is robust, and the hydraulic and electronic parts are likely culprits. We hope there is no internal damage to the gears.

We'll keep everyone posted as we move forward with troubleshooting. The valve body, pressure sensors, and new filter should arrive sometime next week.

Below are the faults from the PIWIS.

PV997

Check out this thread starting at comment 18 (linked) where the poster reprogrammed his TCU (that he thought was bricked) back to factory conditions using a PIWIS3 clone:

https://rennlist.com/forums/997-foru...l#post18270139

You may want to reach out to him or post on that thread. Others have also done it with a PIWIS2 (I think @amet has IIRC). Who knows what's actually going on during some of these PIWIS tasks as the documentation is terrible and it's designed for for tech following an algorithm rather than actually understanding what they are doing. However, it looks as if the actual software for the module is part of the suite provided with the PIWIS, no connection to Mother Porsche is needed to retrieve it.

Those codes you posted are all over the map and perhaps there is some common mechanical cause but I don't see it. Those "measured value implausible" codes aren't an open or short (they have their own different codes) but mean that the values don't make sense to the TCU, i.e. they are outside the range the TCU expects to see. This would also fit with the hypothesis that the TCU as configured is mismatched to the transmission. As to the clunks that could just be from harsh engagement but the fine particles aren't a good sign. It may be clutch pad wear from the clutch slipping due to the improper application pressure. That would also be consistent with the "implausible pressure" fault codes as those sensors are right at the input to the clutch.

You can see the adaptation values if you run a VAL (Vehicle Analysis Log) with the PIWIS. It's in the PDK section of the VAL report and you'll see pages and pages of alpha-numerical characters. The characters in and off themselves aren't compensable but comparing them after making TCU changes may lend some insight into what is actually changing over time.

The good news is that usually these types of faults don't indicate something physically broken within the PDK, but that the TCU doesn't like what its seeing so it disables the PDK to prevent damage.