IDFB

Quick update from the shop:

Turns out Shop was using an Autel all along; they don't have a PIWIS. Shop says that an Autel tech was remoting in to perform the PDK calibration, but couldn't initiate it either because Porsche has blocked access to that function (PDK calibrations) or because of an internal PDK fault. Shop is advising replacing the entire PDK as they believe there's something wrong with the clutch or the internals as they say it should more or less function even without a calibration (it won't shift into gear at all right now since the new valve body, whereas the PDK shifted perfectly fine when not fully warmed up prior to the valve body replacement).

PV997: When you saw the HYTI error before (after replacing a distance sensor), how was the PDK function when the error was active? Was it severely restricted?

I just can't wrap my head around how it could be anything other than the fact that the DME/PDK needs to be calibrated after installing a new valve body. What internal fault could possibly be throwing a P1872 (data record for HYTI synchro invalid) while not tripping any other codes - keeping in mind that the car was shifting fine when not fully warmed up before the new valve body?

I did speak to the shop about providing them with a PIWIS or PIWIS clone, but they're not convinced it will work because 1) they don't want to risk damaging the DME/TCU with faulty data using clones and 2) they say that the data for the calibration has to be obtained from Porsche through an internet connection, which Porsche won't release. Did something change recently in the PIWIS world? I haven't seen anything in the PDK repair thread to indicate that PDK calibrations weren't possible anymore.

Appreciate your thoughts.

stjoh

The clone PIWIS can still do calibrations, I just did it. No connection to Porsche is required.

PV997

Kind of shocked to read the shop didn't have an actual PIWIS. I don't know anything about the Autels but Porsche probably didn't provide them with proprietary information when developing the tool, so Autel likely reverse engineered it. We really don't know how or what it's doing regarding calibration, or if it's even being done correctly. Perhaps it's doing it just fine but it's just another unknown thrown into an already murky process.

This is why I say over and over this is PDK repair stuff best suited for knowledgable DIY'ers (and having an analytical mind sure helps). We don't know everything (and Porsche won't tell us) and you really don't want to pay a shop $150/hour for their on-the-job training, particularly if they don't understand the diagnostic art.

Over in the PDK repair sticky thread @teevirdi is dealing with this exact same error (HYTI) after switching out a valve body. Long story short we don't know what is causing at as multiple people have replaced valve bodies or solenoids without issue. Here's what I told him:

IMO it can only be in two areas; 1) a mechanical issue that is preventing the hydraulics from completing the routine (air in system, pressure leak, etc.) or 2) incompatible configuration issues with the TCU causing a fault. Beyond that I can't think of any other cause.

So I wish I had a better answer but unfortunately this only seems to happening in a subset of cases and we have not tracked down the cause.

PV997

I hope this was just miscommunicated as none of it makes any sense. As stjoh states no connection to Porsche is needed for the calibration at least with a PIWIS. Who knows what the Autel is doing but I highly doubt it's being "blocked by Porsche", that sounds like complete and total BS to me.

The likely reason you cannot shift into any gear is that the TCU noted a major fault (the HYTI error) and locked the gears out. The clutch did not suddenly break, what nonsense. So that fault needs to be cleared through successfully running the calibration before the TCU will allow gear engagement again.

Edit: My 997 fault code reference doesn't include P1872 for some reason, but it does have P1871 (a very similar teaching fault). The "fault effects" section which describes how the TCU reacts to the fault says "clutches open". In other words, the TCU will not let the clutches engage until the fault is cleared, that is why it now won't go into gear. It almost certainly does this to protect the clutches so they don't get trashed due to a bad calibration (a good thing).

What your shop is telling you sounds like they are just shooting in the dark IMO.

PV997

Following up with a few questions:

Are they able to clear the P1872 fault with the Autel? If yes, does it immediately reappear? If no, that's important also.

If they are able to clear the fault and it doesn't come right back, does the car go then into gear (i.e. the TCU has released the clutch disable)?

Assuming they were able to clear it, have they tried to recalibrate the PDK after the fault was cleared? If yes, does the error come right back?

If they can't clear the fault or if it keeps coming right back, have they tried disconnecting the battery to clear all faults?


I came across something in another thread that implied the P1872 fault is persistent, it can't simply be cleared with a PIWIS. This would have the effect of making the PDK inoperable (clutches are locked out from engaging). The dealership doing the work contacted Porsche and was told to disconnect the battery for a while to erase all the faults and start over. It worked for them and it's probably worth a try.

BTW, if you can give us answers to the questions above it will help us have a better understanding of the phenomena.

IDFB

I believe the shop mentioned how the P1872 won't clear and that they couldn't get past it during the calibration. I'll ask if they have tried a hard reset yet - I did mention to them how someone had success clearing that HYTI error by disconnecting the battery. My concern is whether it's even wise to let anything other than a PIWIS or PIWIS clone calibrate the PDK.

Fined

There are a number of shops repairing PDK gearboxes here in Germany. Repairing, rebuilding, strengthening components. And not terribly expensive either. It may be worthwhile to discuss the symptoms of the issue with a shop that works on these gearboxes.

IDFB

Would you happen to know of any contacts at those shops that would be open to discussing?

PV997

You could go the PIWIS clone route yourself, here's the model I have:

https://obd2.market/product/porsche-piwis-2-tester/

It's the $360 version, with fees and shipping it's about $430.

Here's the definitive thread on how to get it running from jjrichar over at Planet 9:

https://www.planet-9.com/threads/piw...achine.238021/

I think you have to get the fault cleared before it will let you do anything else. I really don't think the fault is with the PDK itself but some sort of corruption in the TCU, it's a software error.

Edit: Just noticed the link to OBD2.market says it's out of stock. Many others also had it when I was shopping around so it should still be out there.

IDFB

I'm thinking of going the PIWIS clone route too. The concern though is seeing teevirdi still experiencing the HYTI error even after leaving the battery disconnected. The shop had tried that as well (disconnecting the battery), but said the error is persistent. Apparently the shop doesn't use Autel or PIWIS, but some diagnostic tool they wouldn't disclose the name of. The only other solution is transporting it to the nearest Porsche dealer, which is almost 300 miles away.

What are the chances of bricking the computer if I were to try the PIWIS clone myself? Will this 10 amp charger/battery maintainer be enough to run calibrations?

IDFB

Quick update: the tech from the parts dealer thinks that the HYTI error is from the shop incorrectly performing the fluid fill/installation (40-60 minute process of setting fluid levels and setting parameters when installing a new valve body).

Here's a link for a diagnostic printout from them: P1872

If this were the case, would the next step be a drain and fill of the clutch fluid again?

PV997

Yes, look up at my comment #12 and that was one of the things I suggested, also Teevirdi mentioned he had some kind of strange error during the fill process. I suggested to him that if there was air in hydraulics it could (perhaps) cause something like this as the cal was trying to move things but they weren't responding correctly (due to the air compressing). I don't know if he ever retried filling it.

Your linked fault suggests it's a software problem but it could be a software fault in response to a hardware error (e.g. the hardware isn't working properly so the software adaptation values generated from the malfunction hardware aren't valid). This could just manifest as a software fault as the PDK doesn't know why it happened, but that the calibration doesn't make sense. Did the shop have any errors during the fill? I think teevirdi mentioned a something about a "purge" error that he recalled when I asked him about the fill.

I don't even think you need to drain it but just run through the fill process again and make sure the whole system is purged and topped off without any errors. There's no harm in draining it except the fluid is expensive so use a very clean container if you want to reuse it.

IDFB

The shop isn't divulging much information and is being rather evasive/defensive in general - especially when asked to provide specifics about how they performed the clutch fluid fill. They won't say what software they're using (just that it's not PIWIS) or whether they encountered any specific errors during the fill procedure. The only comment they are willing to make is that they can't get past P1872 and that's it - they're unwilling to go any further.

If there is an air bubble in the system, it sounds like the PDK Fill Mode (on the PIWIS) will cycle through the purge process to remove air without needing to drain the clutch chamber - is my understanding correct?

Thanks.

PV997

Just be clear, I'm not saying the other devices won't work for the fill procedure but that they are unknown. And when you are dealing with an already temperamental and murky process it's throwing in another uncontrolled variable that further confuses things. Maybe they work, maybe they don't (and the fact the shop won't tell you what they are using is really off). But we know the PIWIS works as many of us have done both the fill and the cal procedure using them.

The PIWIS does several things during the fill process. First off, you fill the PDK until the fluid just dribbles out the side fill hole. You then start the car and start the PIWIS fill routine. It ensures the fluid is at a certain temp and it turns off the auxiliary oil cooler (because of this it will abort after 5 or 10 minutes if you dawdle to protect the transmission from overheating). After that it circulates the fluid (which is why the engine has to be on as it drives the pump), and it opens each of the solenoids and pushes the fluid through the various paths to flush out any air. You'll hear the transmission clicking and doing various engagements during this. There are a few times it tells you to manually cycle through the gears IIRC. After it's done with all that it prompts you to again fill the PDK until fluid just dribbles out the fill hole. If there was a bunch of air in the system the fluid level will have dropped from where it was at the start of the process.

It's not hard but you have a limited time to do all this or it will abort out part way through. The PIWIS has on screen instructions that will prompt you through the whole process.

IDFB

Update:

Car is at the dealer. Calibration was completed so P1872 is gone but P17B1 returned. Clutch 1 is inoperable so the car limps using the even gears. Dealer wants to replace the transmission. Looks like it wasn't the solenoid after all.