Transmission Update- (kinda long)
05-01-2006, 10:55 PM
#12
AudiWorld Super User
Finally got to the bottom of this on my 4.2...long but hopefully worth it for tranny sufferers
Long story short, bad wire between the transmission output speed sensor (sensor G195) and the TCU. I looked for this problem for the last 30K miles and three years. I have searched and read every post about 5HP24A's, especially on the D2 A8 board where there is much more cumulative experience than on the C5 A6 board or the S4 board (the S4 board has more posts about the Level Ten valve body that acts like a shift kit). But, I don't think anyone has ever fingered the output sensor or the wire, except by indirect reference to an old 2000 C5 A6 4.2 TSB (see below).
To review my 2000 4.2 (5HP24A, tranny code ECF) issues: Starting about 40K, the infamous clunk--as car shifts up, typically out of first, car makes a big clunk that shakes the car as tranny shifts. Feels very similar to getting rear ended at a stop light and passengers notice it big time. Most likely to happen as tranny left first, hestitated momentarily, perhaps with continued slightly increasing throttle pressure like when you are going up a slight incline and it seemed to want to kick back down to first. It goes into safe mode, which is either 4th or 5th gear (people say 4th, but the scans I ran seem to say it was in 5th). Turn ignition off and on and it reset. If it clunked twice in one driving session CEL also came on. An MAF replacement at a little over 40K helped a lot, but issue happened every few weeks still. In past six months it progressed from weekly to quasi daily. City driving made it act up more (I now know why; the sensor is critical once it leaves first). Using TIP would not avoid the issue either. In past few months it started missing shifts, particularly kickdowns. It could go over tach redline and start bouncing off the limiter (particularly 2-3 where the rate of engine/tranny speed change is slower than 1-2), and 3-2 downshifts were harsh, sometimes to the point of it overreving as though it was headed to first before I let off gas. This is when it got critical, since I didn't want to screw up the motor. I changed AT fluid and filter at 40K, and again at 65K. I have always used Audi fluid or Esso numerical equivalent (the Pentosin brand). I go through the whole painful only-on-a-German-car fluid level procedure per Bentley. At 65K, I also took out all the solenoids (pretty easy to do with pan down) for inspection. My theory had been a faulty solenoid or a related weak spring, either the main pressure solenoid or the torque converter solenoid. Nothing was obvious though.
The recurring code I have gotten is 17114. This is a scary code, in that there are a lot of posts tying this to tranny slippage and failure. But I've noticed those posts really never mention whether the code says it is intermittent (as mine always did) , or continuous. Presumably slippage would be continuous.
Based on more reading of posts and my continued hunch it was a solenoid related fluke, I just had valve body rebuilt and I installed it myself. That was worth it in the end too--shifts just like new once I later got to the root of the issue once and for all. A ZF distributor, Ericcson Industries in CT, did it on a 2 day turnaround. The valve body is an indicated possible problem/fix for 17114 in Bentley. Cost was $550, and the tranny stays in place and valve body is removed with common mechanic's tools. The whole subassembly is close to a foot square and 3 inches thick, weighing 10+ pounds overall. Relative to the cost, hassle and down time of rebuilding the tranny I figured it was a bet worth making. I had also never found any telltale broken clutch type pieces in draining fluid and I didn't think converter was slipping, including based on a VAG scan I logged over a 30 minute driving cycle onto the freeway and back. I also replaced the internal tranny wiring harness for about $100--parts are much cheaper through ZF than through Audi. Old one looked kind of cooked but sound, but I broke a connector when removing the valve body so that made it mandatory. I can post pics of the valve body in and out of the tranny if folks really want to see it.
The rebuilt valve body still didn't fix it, though shifting felt better. Dropping the valve body by the way resulted in about a 9 quart fill, so 75%+ of the fluid had come out, vs. 7 - 7 1/2 quarts max before. With the valve body out you can see up into the clutch area a bit, which looked all in order visually. I also added one container of red LubeGard too when I refilled the tranny, per the A8 posts I found on the subject that tied to an obscure Saab TSB for a different ZF unit that also uses a high slip converter and suffers from shudder issues like the ZF units and some others.
I went back to old TSB 01-00-02 about the bad jumper harness on early 4.2's based on some C2 A8 posts I read yet again. Mine is out of the VIN range though--it's even earlier than those listed. Nonetheless I had bought the harness a long time ago, but visually never could figure out how to connect the two ends to anything on or near the tranny under the car--the TSB never makes clear where it attaches exactly, and doesn't even make clear it is the output speed sensor on the side of the tranny rather than the input speed sensor that it on the top of the valve body internal to the tranny. No connector ever showed any water ingress on my car by the way, and the output speed sensor connector is so heat soaked by the right converter as well as high up in the tranny tunnel area I find it hard to believe water would ever stay there unless the car were submerged in a flood or some such.
A week ago I finally tore the whole dash apart to try to get to the inside passenger compartment connector--if it had one, I figured it was deep under the ash tray, which is where I finally figured the cable entered the passenger compartment. Turned out there is none on my car--I did find the wire there after hours of dashboard disassembly. Thus the TSB was technically right the connector wasn't for my car. So I carefully looked over the wire, cleaned the tranny side connector and put everything back together. Bad news was no fix, the odd good news was it now FINALLY did it every time leaving first into second. Since right before the work it did more or less still work but with the sporadic clunk, I finally knew it was almost certainly electrical. I could also see on the VAG-COM tranny output RPM was now always ZERO--that made no sense of course since I was logging it on a road test and the car was moving (albeit in limp mode once it left first).
Ohmmeter testing from TCU to tranny output speed sensor connector confirmed one wire was broken (it did not test as shorted, but rather discontinuous). The other was fine, and there is a third shielding ground that tested okay too. Looking at the cable there is NO POSSIBLE WAY to ever see a break or problem. MY guess is it was probably broken at the BRAIN DEAD rough wire clamp they put at a severe angle attached to the heat shielding just above the right side catalytic converter. It's real clear to me Audi and ZF didn't coordinate at all on this wiring or the sensor and exhaust locations. Plugging new harness onto the sensor showed good continuity, and sensor tested just at the high end of spec for resistance.
With no other practical choice due to limited cable length and impossibility of reaching original cable routing hole from underneath, I drilled a 1/2 inch hole through the transmission tunnel sheet metal right beside the sensor and about 2-3" above the converter, cut off the inside connector end of the replacement jumper, ran the cable through the hole I made and under the carpet to the TCU, splicing it into the complex TCU connector. I Permatexed the hole I made with high temp sealant to make it watertight and protect the cable housing at the hole edge, leaving just enough slack to disconnect it cleanly at the tranny if desired (the jumper has the "factory" connector on that end). No one who doesn't know the car intimately would ever know I modified it. I reused the silver heat insulation around the connector and wire--with it much shorter than before I could get the cable more distant from the converter, and cover both the connector and the cable all the way to the exit hole. Frankly, the factory should have done it something like this; their installation is dumb--huge heat soak from the converters, terrible cable routing, etc.
Proof in the pudding--instantly codes went away. Every shift is right there. Zero hunting, zero missed shifts. With OOO32 it kicks down just as it should. Shifts cleanly right at 6500. Tranny went from ragged and tired to like brand new, just like a light switch before and after the rewiring. Prior fluid and filter changes didn't do it. A recent attempt to change MAF just caused it to overrev more, recoding didn't do it no matter what I tried, recalibrating throttle etc. only lasted a day, etc.
Going back to Bentley, sure enough problems with the rear output sensor also tie into 17114. But who would guess a bad wire buried in a connector cable that visually looks fine, and is super hard to test. My best guess is for years it was internally frayed or cracked, and when I finally really worked on it the first time I separated the break and forced the issue so it happened on every shift.
Finally if I were looking for this problem by way of testing, I would now look at the VAG COM output for the output speed sensor, setting it to take rapid measurement points. (When I logged earlier I was focused on the solenoid action, and again no posts really get into the output speed sensor and only now do I know how critical it is to each TCU shift decision.) I would also log the tranny gear it is in. I would then test drive it trying to force the "clunk" and resulting safe mode. Then I would review the log data looking for an error with the output speed sensor data (probably a drop out or sharp speed reading reduction) at the point the clunk occurred. My practical experience though now is if you have the "clunk" with a a kick out to safe mode, the 17114 code (ESPECIALLY if it says "Intermittent") but no evidence of tranny slippage, zoom right in on this sensor and especially the wiring.
Hope this saves some frustrating tranny rebuilds!
To review my 2000 4.2 (5HP24A, tranny code ECF) issues: Starting about 40K, the infamous clunk--as car shifts up, typically out of first, car makes a big clunk that shakes the car as tranny shifts. Feels very similar to getting rear ended at a stop light and passengers notice it big time. Most likely to happen as tranny left first, hestitated momentarily, perhaps with continued slightly increasing throttle pressure like when you are going up a slight incline and it seemed to want to kick back down to first. It goes into safe mode, which is either 4th or 5th gear (people say 4th, but the scans I ran seem to say it was in 5th). Turn ignition off and on and it reset. If it clunked twice in one driving session CEL also came on. An MAF replacement at a little over 40K helped a lot, but issue happened every few weeks still. In past six months it progressed from weekly to quasi daily. City driving made it act up more (I now know why; the sensor is critical once it leaves first). Using TIP would not avoid the issue either. In past few months it started missing shifts, particularly kickdowns. It could go over tach redline and start bouncing off the limiter (particularly 2-3 where the rate of engine/tranny speed change is slower than 1-2), and 3-2 downshifts were harsh, sometimes to the point of it overreving as though it was headed to first before I let off gas. This is when it got critical, since I didn't want to screw up the motor. I changed AT fluid and filter at 40K, and again at 65K. I have always used Audi fluid or Esso numerical equivalent (the Pentosin brand). I go through the whole painful only-on-a-German-car fluid level procedure per Bentley. At 65K, I also took out all the solenoids (pretty easy to do with pan down) for inspection. My theory had been a faulty solenoid or a related weak spring, either the main pressure solenoid or the torque converter solenoid. Nothing was obvious though.
The recurring code I have gotten is 17114. This is a scary code, in that there are a lot of posts tying this to tranny slippage and failure. But I've noticed those posts really never mention whether the code says it is intermittent (as mine always did) , or continuous. Presumably slippage would be continuous.
Based on more reading of posts and my continued hunch it was a solenoid related fluke, I just had valve body rebuilt and I installed it myself. That was worth it in the end too--shifts just like new once I later got to the root of the issue once and for all. A ZF distributor, Ericcson Industries in CT, did it on a 2 day turnaround. The valve body is an indicated possible problem/fix for 17114 in Bentley. Cost was $550, and the tranny stays in place and valve body is removed with common mechanic's tools. The whole subassembly is close to a foot square and 3 inches thick, weighing 10+ pounds overall. Relative to the cost, hassle and down time of rebuilding the tranny I figured it was a bet worth making. I had also never found any telltale broken clutch type pieces in draining fluid and I didn't think converter was slipping, including based on a VAG scan I logged over a 30 minute driving cycle onto the freeway and back. I also replaced the internal tranny wiring harness for about $100--parts are much cheaper through ZF than through Audi. Old one looked kind of cooked but sound, but I broke a connector when removing the valve body so that made it mandatory. I can post pics of the valve body in and out of the tranny if folks really want to see it.
The rebuilt valve body still didn't fix it, though shifting felt better. Dropping the valve body by the way resulted in about a 9 quart fill, so 75%+ of the fluid had come out, vs. 7 - 7 1/2 quarts max before. With the valve body out you can see up into the clutch area a bit, which looked all in order visually. I also added one container of red LubeGard too when I refilled the tranny, per the A8 posts I found on the subject that tied to an obscure Saab TSB for a different ZF unit that also uses a high slip converter and suffers from shudder issues like the ZF units and some others.
I went back to old TSB 01-00-02 about the bad jumper harness on early 4.2's based on some C2 A8 posts I read yet again. Mine is out of the VIN range though--it's even earlier than those listed. Nonetheless I had bought the harness a long time ago, but visually never could figure out how to connect the two ends to anything on or near the tranny under the car--the TSB never makes clear where it attaches exactly, and doesn't even make clear it is the output speed sensor on the side of the tranny rather than the input speed sensor that it on the top of the valve body internal to the tranny. No connector ever showed any water ingress on my car by the way, and the output speed sensor connector is so heat soaked by the right converter as well as high up in the tranny tunnel area I find it hard to believe water would ever stay there unless the car were submerged in a flood or some such.
A week ago I finally tore the whole dash apart to try to get to the inside passenger compartment connector--if it had one, I figured it was deep under the ash tray, which is where I finally figured the cable entered the passenger compartment. Turned out there is none on my car--I did find the wire there after hours of dashboard disassembly. Thus the TSB was technically right the connector wasn't for my car. So I carefully looked over the wire, cleaned the tranny side connector and put everything back together. Bad news was no fix, the odd good news was it now FINALLY did it every time leaving first into second. Since right before the work it did more or less still work but with the sporadic clunk, I finally knew it was almost certainly electrical. I could also see on the VAG-COM tranny output RPM was now always ZERO--that made no sense of course since I was logging it on a road test and the car was moving (albeit in limp mode once it left first).
Ohmmeter testing from TCU to tranny output speed sensor connector confirmed one wire was broken (it did not test as shorted, but rather discontinuous). The other was fine, and there is a third shielding ground that tested okay too. Looking at the cable there is NO POSSIBLE WAY to ever see a break or problem. MY guess is it was probably broken at the BRAIN DEAD rough wire clamp they put at a severe angle attached to the heat shielding just above the right side catalytic converter. It's real clear to me Audi and ZF didn't coordinate at all on this wiring or the sensor and exhaust locations. Plugging new harness onto the sensor showed good continuity, and sensor tested just at the high end of spec for resistance.
With no other practical choice due to limited cable length and impossibility of reaching original cable routing hole from underneath, I drilled a 1/2 inch hole through the transmission tunnel sheet metal right beside the sensor and about 2-3" above the converter, cut off the inside connector end of the replacement jumper, ran the cable through the hole I made and under the carpet to the TCU, splicing it into the complex TCU connector. I Permatexed the hole I made with high temp sealant to make it watertight and protect the cable housing at the hole edge, leaving just enough slack to disconnect it cleanly at the tranny if desired (the jumper has the "factory" connector on that end). No one who doesn't know the car intimately would ever know I modified it. I reused the silver heat insulation around the connector and wire--with it much shorter than before I could get the cable more distant from the converter, and cover both the connector and the cable all the way to the exit hole. Frankly, the factory should have done it something like this; their installation is dumb--huge heat soak from the converters, terrible cable routing, etc.
Proof in the pudding--instantly codes went away. Every shift is right there. Zero hunting, zero missed shifts. With OOO32 it kicks down just as it should. Shifts cleanly right at 6500. Tranny went from ragged and tired to like brand new, just like a light switch before and after the rewiring. Prior fluid and filter changes didn't do it. A recent attempt to change MAF just caused it to overrev more, recoding didn't do it no matter what I tried, recalibrating throttle etc. only lasted a day, etc.
Going back to Bentley, sure enough problems with the rear output sensor also tie into 17114. But who would guess a bad wire buried in a connector cable that visually looks fine, and is super hard to test. My best guess is for years it was internally frayed or cracked, and when I finally really worked on it the first time I separated the break and forced the issue so it happened on every shift.
Finally if I were looking for this problem by way of testing, I would now look at the VAG COM output for the output speed sensor, setting it to take rapid measurement points. (When I logged earlier I was focused on the solenoid action, and again no posts really get into the output speed sensor and only now do I know how critical it is to each TCU shift decision.) I would also log the tranny gear it is in. I would then test drive it trying to force the "clunk" and resulting safe mode. Then I would review the log data looking for an error with the output speed sensor data (probably a drop out or sharp speed reading reduction) at the point the clunk occurred. My practical experience though now is if you have the "clunk" with a a kick out to safe mode, the 17114 code (ESPECIALLY if it says "Intermittent") but no evidence of tranny slippage, zoom right in on this sensor and especially the wiring.
Hope this saves some frustrating tranny rebuilds!
05-02-2006, 03:48 AM
#13
AudiWorld Super User
Join Date: Apr 2001
Location: Northern NJ
Posts: 9,211
Likes: 0
Received 0 Likes
on
0 Posts
wow! Congrats on resolving your issue, and that is a heck of a post. Must have been
excellent to feel everything running properly for the first tiem after you spliced your cables together.
05-02-2006, 07:25 AM
#14
Audiworld Junior Member
Thread Starter
Join Date: May 2004
Posts: 49
Likes: 0
Received 0 Likes
on
0 Posts
holy cow! great post!
That sounds exactly like the problem I am having, my VAG did indicate "intermittent". Since I do not posess the level of skill you do, I am going to take this post with me to the dealer tomorrow.
thank you so much for the time you spent detailing this issue, I am sure that I will not be the only one to benefit from this!
Spinny
thank you so much for the time you spent detailing this issue, I am sure that I will not be the only one to benefit from this!
Spinny
05-02-2006, 07:29 AM
#15
Audiworld Junior Member
Thread Starter
Join Date: May 2004
Posts: 49
Likes: 0
Received 0 Likes
on
0 Posts
Re: It is hard to fathom that a local Denver dealer can't diagnose a defective transmission.
Took it to Prestige. I also have a 2.7, and used to have a 911. they have always done a good job, so this is frustrating. I have had bad luck with the only indy shop, they were just as expensive, in fact I think they were a few bucks more than Prestige.
05-02-2006, 08:47 AM
#16
AudiWorld Super User
Join Date: Aug 2002
Posts: 2,505
Likes: 0
Received 0 Likes
on
0 Posts
Lying and stupid - Haldex service should be covered. My '01 225 interval for the
Haldex is every 20K, so they had to do it twice during the free maintenance period. Guess they upped it to 30K to only have to do it once.
05-02-2006, 04:43 PM
#19
AudiWorld Super User
See A8 cross link where I have now posted pictures...
and more explanations with them.<ul><li><a href="https://forums.audiworld.com/a8/msgs/117873.phtml">https://forums.audiworld.com/a8/msgs/117873.phtml</a</li></ul>
Thread
Thread Starter
Forum
Replies
Last Post
Tgr_Clw
A4 (B6 Platform) Discussion
20
02-26-2007 08:44 AM
rayzor
A6 / S6 (C5 Platform) Discussion
3
04-29-2004 12:48 PM
