Guide to Repairing a PDK Transmission
#181
Three Wheelin'
Thread Starter
Fair warning that this is a big job. It involves removing the transmission, splitting the gear case open, replacing the sensor, resealing the case, then reinstalling the transmission. You will not find a dealer who will do this because it's not authorized by Porsche. It's unlikely you will find an independent garage to tackle it since it's not documented anywhere and many indys are afraid of the PDK. So this is probably a DIY unless you are on really good terms with a shop that is willing to give it a try.
To be blunt, it's gut check time. If you are an experienced DIY, who is willing to do the research needed, and is motivated to save tens of thousands of dollars then it absolutely can be done. If you are one of these guys that pays for $400 dealer oil changes then it probably isn't for you. You'd be better off eating the cost of a replacement transmission or dump the car. BTW, $30k is way too high, it should be more like half than since it would be replaced by a remanufactured transmission from Porsche.
Good luck and please let us know if you come across more information or find a willing shop.
The following users liked this post:
Kpadas (09-17-2020)
#186
Thanks for the response. It is difficult to get a picture. The bolt is on the drivers side of the transmission Parrell with the shift cable and has a round head with a indentation for an allen wrench to remove it. There is a small hole that continues from the allen head indentation through the length of the bolt. Porsche refers to this bolt as a bleeder bolt., Mine was weeping trans oil from this small hole. Every Porsche mechanic I talked to had no idea what the purpose of the bolt was, so I sealed it and stopped the leak. Now I am having second thoughts as to whether the hole in it was some sort of breather? Porsche's answer is to replace a perfectly good working transmission if this leaks, not something that I want to do.
#187
Three Wheelin'
Thread Starter
I took a look at a few photos and didn't see the bolt you mention. Take a look at Ebay and search for PDK transmissions. They have photos of junkyard transmissions out of the car, let me know if there's one that shows the bolt.
#188
Log states temp is in units (F).
Quoted from your #81 post on this thread in regards to Overheating Protection.
“...A calculation model, which calculates the clutch temperature from the engine torque and the slip at the clutch, is also used...”
I think all the added codes are just fail-safe protection.
Have not been back behind the wheel in several days. Might be able to get some new log data next weekend. Problem is that it may be another 6 months before the codes re-appear.
Quoted from your #81 post on this thread in regards to Overheating Protection.
“...A calculation model, which calculates the clutch temperature from the engine torque and the slip at the clutch, is also used...”
I think all the added codes are just fail-safe protection.
Have not been back behind the wheel in several days. Might be able to get some new log data next weekend. Problem is that it may be another 6 months before the codes re-appear.
#189
HI Rage360, were you able to get a solution for this problem. I have a 2011 911 Carrera 4 and just had the same codes (P17F0, P17F1, P17F2 AND P0730). The "Transmission Emergency Run" warning came up twice, but went off within 10 to 15 seconds without me doing anything. These happened once when I was on the highway at about 85 MPH and once when I was standing still in an DMV inspection station. No other symptoms, all gears worked properly, no loss of power and no other issues. I drove it to the Porsche dealer the next day (60 miles away) and no other warnings. I am in Europe and the Porsche dealer has said they have never seen this before and are just going by what their manuals say to do, replace the PDK at U$20K. Hope you have something that can help me.
#190
HI Rage360, were you able to get a solution for this problem. I have a 2011 911 Carrera 4 and just had the same codes (P17F0, P17F1, P17F2 AND P0730). The "Transmission Emergency Run" warning came up twice, but went off within 10 to 15 seconds without me doing anything. These happened once when I was on the highway at about 85 MPH and once when I was standing still in an DMV inspection station. No other symptoms, all gears worked properly, no loss of power and no other issues. I drove it to the Porsche dealer the next day (60 miles away) and no other warnings. I am in Europe and the Porsche dealer has said they have never seen this before and are just going by what their manuals say to do, replace the PDK at U$20K. Hope you have something that can help me.
But at least its not just Porsche discriminating against us "stupid americans" and making us buy $20k transmissions for simple small problems.
#191
Three Wheelin'
Thread Starter
Log states temp is in units (F).
Quoted from your #81 post on this thread in regards to Overheating Protection.
“...A calculation model, which calculates the clutch temperature from the engine torque and the slip at the clutch, is also used...”
I think all the added codes are just fail-safe protection.
Have not been back behind the wheel in several days. Might be able to get some new log data next weekend. Problem is that it may be another 6 months before the codes re-appear.
Quoted from your #81 post on this thread in regards to Overheating Protection.
“...A calculation model, which calculates the clutch temperature from the engine torque and the slip at the clutch, is also used...”
I think all the added codes are just fail-safe protection.
Have not been back behind the wheel in several days. Might be able to get some new log data next weekend. Problem is that it may be another 6 months before the codes re-appear.
After thinking about this I keep coming back to the pressure values being so high (14 bar). Those ZF solenoids only have a spec'd control range of 0 to 4.7 bar so I don't know how that's possible.
Edit: Do you have pressure logs when the PDK was acting normally? It would be interesting to see the pressures then. I logged mine a while back just for fun, I'll see if I can find them and what they read.
Edit 2: Just double-checked the documentation and the P17Fx faults specifically reference the sump fluid temp (e.g. P17F0: oil sump temperature > 150 deg C). I suppose it could be triggered by the calculated clutch temp also and the documentation isn't specific.
Last edited by PV997; 06-29-2020 at 03:04 PM.
#192
Three Wheelin'
Thread Starter
HI Rage360, were you able to get a solution for this problem. I have a 2011 911 Carrera 4 and just had the same codes (P17F0, P17F1, P17F2 AND P0730). The "Transmission Emergency Run" warning came up twice, but went off within 10 to 15 seconds without me doing anything. These happened once when I was on the highway at about 85 MPH and once when I was standing still in an DMV inspection station. No other symptoms, all gears worked properly, no loss of power and no other issues. I drove it to the Porsche dealer the next day (60 miles away) and no other warnings. I am in Europe and the Porsche dealer has said they have never seen this before and are just going by what their manuals say to do, replace the PDK at U$20K. Hope you have something that can help me.
The fourth fault (P0730) has about four variations, did they give you any text from the error? Depending on the variant it has different causes.
Don't believe a word the dealership says regarding PDK diagnosis. Their flow diagrams are incredibly simplistic and only have a few solutions, most of which end in "replace PDK transmission". Even when there are things the dealer is authorized to do (such as replace the valve body) they still often recommend PDK replacement. I'm not sure of it's incompetence, cowardice, or greed, most likely a combination of all three.
The following users liked this post:
C4SJOHN (11-15-2020)
#193
Hi
My 997.2 pdk shut down when I was waiting for a traffic light, without any sign of malfunction before.
I am not the only one, a few I know in my area.
P center informed me the same, P1731, and they are not capable of and not able to repair my pdk, except replace a new one.
When a more complicated Rolls Royce aircraft engine could be disassembled, repaired and fly, a PDK cannot be repaired become funny.
My 997.2 pdk shut down when I was waiting for a traffic light, without any sign of malfunction before.
I am not the only one, a few I know in my area.
P center informed me the same, P1731, and they are not capable of and not able to repair my pdk, except replace a new one.
When a more complicated Rolls Royce aircraft engine could be disassembled, repaired and fly, a PDK cannot be repaired become funny.
The following users liked this post:
maschinetheist (10-23-2021)
#194
Three Wheelin'
Thread Starter
Log states temp is in units (F).
Quoted from your #81 post on this thread in regards to Overheating Protection.
“...A calculation model, which calculates the clutch temperature from the engine torque and the slip at the clutch, is also used...”
I think all the added codes are just fail-safe protection.
Have not been back behind the wheel in several days. Might be able to get some new log data next weekend. Problem is that it may be another 6 months before the codes re-appear.
Quoted from your #81 post on this thread in regards to Overheating Protection.
“...A calculation model, which calculates the clutch temperature from the engine torque and the slip at the clutch, is also used...”
I think all the added codes are just fail-safe protection.
Have not been back behind the wheel in several days. Might be able to get some new log data next weekend. Problem is that it may be another 6 months before the codes re-appear.
There is a big pressure spike up to about 14 bar that corresponds to right after shifting from first to second for some reason (edit: see edit 4 below as I figured this out). The pressure never goes above above 5 bar when rowing through the rest of the gears. Speed sensor data is very smooth and doesn't have anything like the erratic behavior seen in your plots.
Edit: The cause could be a flaky speed sensor on mainshaft 1 that is causing the TCU to think clutch 1 is slipping when it really isn't. In response the TCU is cranking up clutch 1 pressure to try and stop the slipping. The more I think about this the more plausible it seems. Hopefully it isn't this as it requires the transmission to be dropped and the case cracked open. Last off clutch 1 could actually be slipping due to wear but that seems unlikely as the clutch is very robust and you would see it all the time.
Edit 2: Just looked at your logs again and I'm thinking more and more it's a flaky shaft 1 speed sensor. Not sure what the x-axis is on your tool but I'm assuming it's seconds (it's seconds on the PIWIS clone). If so, the speed is jumping 1000 to 1500 RPM in fractions of a second. I don't think this is possible but if it was happening you'd certainly notice it when driving. The TCU is seeing this and futilely increasing clutch pressure pressure to fix it, that's why the pressure goes up to 14 bar when it normally doesn't go above 5 bar. After a few tries the TCU realizes it isn't helping and triggers emergency run.
Edit 3: This theory is consistent with the temp warning if it's being triggered by calculated temp. The actual temp was fine as your monitoring of the sump temp showed. However the TCU would think the clutch temp was very high from the phantom slipping indicated by the flaky speed sensor. BTW, there is also a clutch cooling circuit controlled by it's own solenoid (EDS3). The 14 bar pressure could also be from the TCU cranking up fluid flow to cool the clutch.
Edit 4: I think I have this figured out. When I captured the data below I also captured the currents to the different solenoids. When that big spike up to 14 bar was seen in the lower right hand plot it was when both the clutch was engaged and the clutch cooling circuit very briefly opened to maximum flow. That's probably why the pressure goes so high as the clutch is being fed high pressure fluid by two solenoids at once (i.e. the clutch engagement solenoid and the cooling solenoid). The pressure is normally five bars (for normal clutch engagement) but can go much higher if it's being pushed to maximum engagement and supplemented by the cooling solenoid (EDS3).
In Mr. Hobbyiest's plots the pressure goes to 14 bar both times when the speed sensor data is erratic. Clearly the TCU thinks the clutch is burning up and applying maximum clutch engagement and coolant flow to try and get things under control. When that doesn't work it shuts things down. So a flaky shaft 1 speed sensor pretty much explains everything, the overtemp warning, the speed ratio mismatch, only gears on clutch 2 working, the erratic shaft 1 speed sensor data, and the high clutch 1 pressure. Just to note that if the clutch was truly slipping it would do the same thing so that needs to be ruled out. It should be obvious to the driver if it's slipping. Also since the sump temp was fine it's a pretty good indicator that the slipping isn't real.
There's a small possibility it could be a bad connection at the transmission connector or the TCU (less likely) so it's probably worthwhile to remove the connector and clean the contacts prior to dropping and splitting the transmission.
Last edited by PV997; 07-01-2020 at 08:34 PM.
The following 4 users liked this post by PV997:
#195
I found some PDK sensor data I captured on my car (2010 TT, no PDK issues). See the plots below, the top shows speed sensors and the bottom shows clutch fluid pressures. Left column is clutch 1, right column is clutch 2. During this sequence I started from a stop and manually short-shifted from 1st through 7th gear.
There is a big pressure spike up to about 14 bar that corresponds to right after shifting from first to second for some reason. The pressure never goes above above 5 bar when rowing through the rest of the gears. Speed sensor data is very smooth and doesn't have anything like the erratic behavior seen in your plots.
Edit: The cause could be a flaky speed sensor on mainshaft 1 that is causing the TCU to think clutch 1 is slipping when it really isn't. In response the TCU is cranking up clutch 1 pressure to try and stop the slipping. The more I think about this the more plausible it seems. Hopefully it isn't this as it requires the transmission to be dropped and the case cracked open. Last off clutch 1 could actually be slipping due to wear but that seems unlikely as the clutch is very robust and you would see it all the time.
Edit 2: Just looked at your logs again and I'm thinking more and more it's a flaky shaft 1 speed sensor. Not sure what the x-axis is on your tool but I'm assuming it's seconds (it's seconds on the PIWIS clone). If so, the speed is jumping 1000 to 1500 RPM in fractions of a second. I don't think this is possible but if it was happening you'd certainly notice it when driving. The TCU is seeing this and futilely increasing clutch pressure pressure to fix it, that's why the pressure goes up to 14 bar when it normally doesn't go above 5 bar. After a few tries the TCU realizes it isn't helping and triggers emergency run.
Edit 3: This theory is consistent with the temp warning if it's being triggered by calculated temp. The actual temp was fine as your monitoring of the sump temp showed. However the TCU would think the clutch temp was very high from the phantom slipping indicated by the flaky speed sensor. BTW, there is also a clutch cooling circuit controlled by it's own solenoid (EDS3). The 14 bar pressure could also be from the TCU cranking up fluid flow to cool the clutch.
There is a big pressure spike up to about 14 bar that corresponds to right after shifting from first to second for some reason. The pressure never goes above above 5 bar when rowing through the rest of the gears. Speed sensor data is very smooth and doesn't have anything like the erratic behavior seen in your plots.
Edit: The cause could be a flaky speed sensor on mainshaft 1 that is causing the TCU to think clutch 1 is slipping when it really isn't. In response the TCU is cranking up clutch 1 pressure to try and stop the slipping. The more I think about this the more plausible it seems. Hopefully it isn't this as it requires the transmission to be dropped and the case cracked open. Last off clutch 1 could actually be slipping due to wear but that seems unlikely as the clutch is very robust and you would see it all the time.
Edit 2: Just looked at your logs again and I'm thinking more and more it's a flaky shaft 1 speed sensor. Not sure what the x-axis is on your tool but I'm assuming it's seconds (it's seconds on the PIWIS clone). If so, the speed is jumping 1000 to 1500 RPM in fractions of a second. I don't think this is possible but if it was happening you'd certainly notice it when driving. The TCU is seeing this and futilely increasing clutch pressure pressure to fix it, that's why the pressure goes up to 14 bar when it normally doesn't go above 5 bar. After a few tries the TCU realizes it isn't helping and triggers emergency run.
Edit 3: This theory is consistent with the temp warning if it's being triggered by calculated temp. The actual temp was fine as your monitoring of the sump temp showed. However the TCU would think the clutch temp was very high from the phantom slipping indicated by the flaky speed sensor. BTW, there is also a clutch cooling circuit controlled by it's own solenoid (EDS3). The 14 bar pressure could also be from the TCU cranking up fluid flow to cool the clutch.