loverboy8

Surprised Jake hasn’t jumped into fixing/improving these. Seems it could be profitable

Carl_

Hi, I have a 991.1 C4 PDK problem simular to that of TEEVERDI (#658).
The problem started with a PDK-failure while driving , fault code P1765 (Gear disengagement lock) and P1774 (Gear valve 2 stuck on pressure).
After shutting down the engine, was it impossible do drive.
The electro/hydraulic-valves have been changed with refurbished ones from Shinghoi Autoparts, as also the TCU has been replaced with a new one.
The TCU-coding was done with succes by the Porsche-dealer, however they could not perform calibration successfully.
After retrieving the vehicle I also tried to calibrate the TCU in different ways with my PIWIS2 but also without succes.
I have similar problems just like TEEVERDI : an error when bleeding the PDK, and an incomplete calibration (with parts replacement) ending at 71% (column end position not plausible), with TCU-faultode P1870 as result and a none driving car.
I checked the frequency's and the duty-cycle of the distance sensors, only sensor #1 has around 30% while all the others are around 50% in rest. During calibration the duty-cycle of DS #1 varies between 30% and 70%, while all the others vary between 20% and 80%.I was wondering how TEEVERDI solved his problem, because I am a little bit stalled for the moment notwithstanding the very helpful information in this thread.
Regards, Carl

jjrichar

A couple of vids below that might help.

Firstly, getting another valve body to cal, especially an aftermarket recond, is going to be difficult. The whole cal with part replacement is notoriously finicky.

For the gear selection codes, this seems to indicate that the gear selection valves or solenoids aren't doing their thing correctly. Gear selection is done via three gear selection valves, with these being controlled by two solenoids, MV1 and 2. MV 1 controls valve 2, and MV2 controls valve 1 and 3.

Personally I'd be looking at the old VB, pressure testing the MV solenoids and looking at the gear selection valves and seeing if one is stuck or can't move fully.

Also, the DF only getting to 30/70 on DS1 to me is indicating shift rod 1 (gears 4 and 6) isn't engaging correctly. I'm assuming the sensor isn't aftermarket and still the original OEM. These sensors when working fine will be very consistent with the output, so this seems to indicate that when hydraulic power is applied to shift rod 1 it won't fully push the rod to engage fully. Why? Could be a mechanical failure in the gearset itself, but I would be looking at a hydraulic issue as more likely, as it's not just a single selection of one of the gears, but both aren't working.

For shift rod 1 to move, MV 1 is active, and MV 2 deactive. This puts the gear selection valves in a particular position. Then depending on if either EDS 5 or 6 is active, shift rod 1 will be either retracted or extended. It might be the case that the valves aren't fully home, and so full pressure isn't being applied to the actuator.

The only problem with this theory is that it seem all the other shift rods are selecting OK, as the DFs seem to be fine. If my theory above was correct, I would also expect to see issues with other selections.

The only thing I can think of that might be specific to just shift rod 1 is that one of the lock outs in the gear set has failed and it isn't allowing the rod to move. In this case it would stay in neutral when SR 1 isn't activated, but not be able to fully move it home when engaged.

Like I said, I'd start with the VB and check all is good there and then move to the transmission itself.

Carl_

Thank you very much.
I started with testing the old valves but with some troubles sealing the valve noses , tomorrow I will continu.

Carl_

Meanwhile, i have tested all the valves (old en new) with an air pressure regulator and i found no really errors except for a small loss at the nose but that is probably between the test hose and the valve.
I also opened the valve block (thanks to the perfectly documented videos) and i find 2 valves that where a little sticky by removal, but after removal they where smooth and loose.
Everything is back in the gearbox together with fresh oil using the filling mode with Piwis. Now i have done the calibration already 3 times after erasing all the faults in all the ECU's but now the calibration is interupted by an internal teaching error.
When performing "Teach-In hydraulic" under "Stellglieder prufungen" , i get the P1872 fault code. So i am stuk again for today. I will do some tests later on.

jjrichar

Below is a quote of a post I made (#1323) that explains the best I could of what is going on in the calibration processes.

I can't find the part that you describe where it failed. Where in the calibration did it fail? In particular I'm interested in knowing if part C (shift distance calibration) was performed correctly.




The calibration with part replacement carries out all 5 routines a-e. When I've had failures where the distance sensor response was fine, it fails in routine d, the hydraulic teach in (valve body cal).

The calibration without part replacement only carries out routines c and e, that being the shift distance (DS numbers) and the clutch cal. All the other routines are ignored, so there is no adaptation delete and no cal of the valve body. If the distance sensor response is within the prescribed limits set in the TCU, I've never had this fail.

Either routine will not begin until you have the clutch fluid temp above 60C. PIWIS will list a bunch of prerequisites like park brake on and the foot brake pressed. Same prerequisites for both.

Routines are as follows:

a. Reprogramming. I don’t know exactly what is happening here, but it seems like all the adaptations are wiped. Takes about a minute.

b. Pressure sensor calibration. Engine isn’t running. Key is turned off and on a few times. Progressive instructions given by PIWIS. Does it twice that take 30 secs each with about 10 secs between each one.

c. Shift distance calibration. This is where it tests the distance sensor response. Engine is running. Takes about 7 minutes, but sometimes takes a lot less if the numbers it has stored are similar to previously held. Will change through all gears and preselections to get the distances for any possible combination. It will sit there seemingly doing nothing for long periods. This is normal.

d. Hydraulic teach in. This is the valve body cal. You don’t really know what’s going on in the background, but if it fails it will give you the individual steps of the hyd teach in and tell you which it failed on. These are the 12 steps below. Below each is the English courtesy of google translate. A german speaker might give better translations so we can have a better understanding of what is actually happening.

Note that EDS 1 and 2 are the clutch controlling solenoids. EDS 3 and 4 control overall system pressure, cooling and lubrication. MV 1 & 2, and EDS 5 & 6 control gear shifting, and don’t seem to be checked at all in this procedure.

Note that steps 2 and 3 are for the two clutches, but it doesn’t do a clutch calibration like it will do next in routine e, which is much more comprehensive. It might be just doing a calibration of EDS 1 and 2.

1. überprüfung druckstellar beendet
pressure regulator test completed

2. kupplung 1 adaptiert
clutch 1 adapted

3. kupplung 2 adaptiert
clutch 2 adapted

4. notkonzept schwelle EDS 4 abwartsrampe
emergency concept threshold EDS4 downward ramp

5. einlernen notkonzeptschwelle EDS 4
learn emergency concept threshold EDS4

6. notkonzept schwelle EDS 3 abwartsrampe
emergency concept threshold EDS3 downward ramp

7. einlernen notkonzeptschwelle EDS 3
learn emergency concept threshold EDS3

8. systemdruckkennlinie EDS3 abwartsramp
system pressure characteristic EDS3 downward ramp

9. einlernen systemdruckkennlinie EDS3
Teach system pressure characteristic EDS3

10. umschaltschwelle EDS4 abwartsstufen
switching threshold EDS4 down steps

11. einlernen umschaltschwelle EDS4
Teach switchover threshold EDS4

12. einlernvorgang beedet
teach-in process finished

e. Clutch calibration. It increases engine RPM and gently applies each of the clutches over and over to get adaptations of how the clutches behave. Every time I do this it will stagnate at 24% and 74% complete for a few minutes. This seems to be normal.

jjrichar

I just knocked out a video on how to replace the distance sensor or speed sensor on the 911 (991 or 987) without removal of the transmission. It's not a video as such, but a collection of images with voice over and instructions etc.

Big thanks to @stjoh for the images and guidance on the procedure. He's of the opinion that it's every bit as easy as the 981/987. No more than a days work. Removal of the transmission is simply not required.

Carl_

Still no success after several of different calibrations, both with "with part replacement" and "without part replacement" and also via "stellglieder" where you can perform the individual calibrations. Each time the calibration gets stuck when learning the distance of the shift rods (at 58%) with an internal error as the cause. I continue to find it curious that the distance sensor #1 stays at 4.727mm and does not rotate around 0, as seen on many posts, could it be that there is a shift rod stuck.
A separate hydraulic teach-in is also not possible because of a negative feedback of the TCU, probably by not calibrating the shift distance.
Could the shiftrods be manually moved by controlling MV1, MV2, EDS5 and EDS6 ?

regards,Carl

jjrichar

Was the last series of cals with the original valve body installed or still the replacement?

I'm curious about how you get the sensor 1 output when doing the cal. Is there some mode of PIWIS that allows you to see this?

Don't know about how you would shift the rods manually. MV1 and 2 would be simple to apply the correct voltage as they are a simple on/off solenoid. The EDS are not and require a PWM signal to reduce the power and hence pressure applied. I'd be very worried about applying too much pressure and seriously breaking something. Me personally I'd be removing the rear casing and moving by hand.

stjoh

This would be my recommendation too. This should allow you to determine whether the issue is with the actuation or with the sensing.

cbracerx

Long time PDK thread lurker, first time poster.

I've got a 2011 987.2 that has a replaced PDK with about 20K miles on it. During a track event the dreaded emergency error was shown and the PDK stopped working. A full capacitive discharge of the electrical system was performed with no change in the bad behavior. After reattaching the battery we had these relevant DTC codes stored which I read with an x431 device.

DME (Digital Engine Electronics) 1 problem exist
1.P0700 Transmission Fault Signal implausible
Tiptronic Transmission Control 7 problems exist
1.P18B4 Software Error No fault symptom available
2.P177F 7th Gear Not Disengageable No fault symptom available
3.U0418 CAN Fault, Brake No fault symptom available
4.P17DA Hydraulics Fault During Change To Neutral Position No fault symptom available
5.P1732 Shift Rod 2 Displacement Sensors Signal implausible
6.U0155 CAN Timeout Instrument Cluster No fault symptom available
7.U0146 Communication With Gateway Control Unit No fault symptom available

Taking the "what have I got to lose approach" I cleared the codes and started the car, attempted to use R and D. Car just sits still and is in neutral. When the car is put in park it is no longer in neutral, and does not roll.

Read the codes again - and got no errors!

Did some direct observation of the data feeds from the TCU, which suggested that the sensors for temp and shift rods seem to be working as expected: The TCU knows which drive mode is selected. I cleared the adaptations - that worked. But attempting the hydraulic calibration ended in:

P1872 Data Record For Hydraulics Calibration Is Invalid

TCU certainly isn't toast as we had a nice chat via x431, but doesn't seem to be commanding the solenoids to do anything.

What is the collective's suggested next step?

This error is mentioned here https://rennlist.com/forums/997-foru...l#post17584274 with the suggestion in this post https://rennlist.com/forums/997-foru...l#post17584303 to reprogram the TCU.

stjoh

Based on the codes it looks like there's an issue with the hydraulics. I'd start with the basics:

- Check for leaks, confirm the fill level of clutch fluid is good.
- Measure resistance of the two clutch pressure sensors.

The descriptions from your x431 do not exactly match the Porsche troubleshooting guide, any chance you can get a PIWIS hooked up when you get the codes again?

We've seen a number of issues with sensors (pressure and speed) where they work when cold but fail at temperature so when you do your checks make sure you let the transmission get to working temperature.

To unbrick your TCU and get through the cal I'd try a full re-program er:

https://rennlist.com/forums/997-foru...ission-84.html

Good luck and keep us posted.

jjrichar

That seems very strange that when you had zero codes and selected a gear it did nothing and had no codes afterwards.

When during the cal did it fail? During the shift rod cal or the hyd teach in?

The code for the distance sensor (1732) and the 7th gear disengagement code I would think are related, as shift rod 2 controls the engagement of 5th and 7th gear, and distance sensor 2 is there to detect this.

All the other codes seem be related to a bus communication problem. Maybe a voltage problem, or an intermittent electrical connection issue?

Carl_

I have removed the rear casing, every shaft could be moved, I also checked the magnets on the shaft and they where fine.
Next I put an oscilloscope on each distance sensor to look at the PWM-signal while moving a magnet over it, each PWM varies from 10% to 90%.
The only strange thing is that the PWM-signal of the 1st sensor is at 60% and the others at 95% when no magnet is around (see image)
I have never opened a gearbox before, but at first glance I see nothing wrong. What's going on?
Looks like "cbracerx" has similar calibration problems.

For looking at the distance sensors while calibrating, this should be possible while doing the calibration via "stelglieder prufungen" (or drive links check) > Teach-In schaltstangen with PIWIS2 and then scrolling to the bottom of the page.

jjrichar

Channel 1 at 60% with no magnet is wrong. They should all be at 95% with no magnet. It's a check within the TCM that a magnet hasn't fallen off. If power is applied to the sensor and no magnet is detected it should always give a 95% DF return for each channel.

If you have the distance sensor not installed (just hanging free), but the plug for it is connected to the car, the correct distance seen on PIWIS is about 12.7mm, with channel 3 being the negative of all the others. This is the TCU interpretation of a 95% DF return.

If you disconnect the sensor plug from the car so the TCM is getting no signal, ie 0% DF return, you should see about 13.5mm on PIWIS for all channels, with channel 3 the negative of the others again. This is the response of a 0% DF return.

Note that if you are looking at a particular channel and the 95% DF return is giving 12.7mm, the 0% return will be -13.5mm.

Is this what you see?

If you are then the DS channel 1 is faulty. If you aren't seeing this I suspect a TCM fault.