Question1995 Cirrus. In the morning car starts as usual. If I run an errand in afternoon, car in sun, when the car is started the odometer reading has a 0, then all the gears are lit but none selected, then they are all selected. Car drives in limp mode. A few times the odometer read "no bus link" and key test shows code 66. If I do not run errand and leave work with car then in shade, car starts as usual. We talked to dealer service manager and decided to just change the input and output sensors ourselves. Car still not fixed. Took car back to dealer, explaining every sympton. They hooked it up and code indicated the output sensor was bad and so was the transmission sensor. Even though output was brand new. They said they took them out and both were not sending signals. They replaced the output sensor and GUARANTEED it was the solution, plus recommended we also replace the other sensor. Test drive showed car working fine, of course they only drove around lot, never got past 3rd gear. I drove home and it shifted strange. Had no shift problems when I first came to dealer. Figured computer was reset and relearning gear location. Still shifiting horribly, when past 3rd gear it drops in and out of gear. Not hard shifting but more like slipping in and out. Did key code test and still getting 66 code that I had when I first took to dealer. Service Tech says now tells me we have to start from square one, says that if the car was not in limp mode when they tested it it may not have shown those problems, etc. They are the expert, they should have told me that car needed to be "in problem stage" to be tested. Car was in shade, testing in morning not showing limp mode. I am thinking they should have tested the bus wires, the TCM and the BCM for activity. We did replace the PCM two years ago. Dude wants me to now bring it back while in limp mode so they can look at it again, of course charging me again. We agreed to half price but I am thinking the tech did not fully inform mechanic of my entire problem, therefore mechanic could only diagnose what the guy told him to. I have done lots of internet reading and some say output sensor is simple solution but readings go on to indicate this may be more into the connectivity of the bus between the main components. Your help will go a long way as I may be taking car in real soon.
AnswerHi Ellen,
It is hard for me to evaluate the communication difficulties that you recount so I will not do so but rather focus upon the problem and the symptoms.
The 66 code (and "no bus link" in the odometer) is very relevant because it says that the digital data (including that from the speed sensors) that originates in the transmission and its control module is not being received by the other participants in that data sharing network (most notably the PCM and the body computer). That alone would have been the clue as to where this whole problem should have begun to be addressed at the shop rather than with the torque converter or output speed sensors which were rather obviously a consequence of the data not coming from the TCM rather than the cause of the problem. So you might be able to get them to reverse the charges on those sensor items and take them back into stock (oops I slipped into evaluating the history).
The first thing to do would be to verify whether the wires and plugs that connect the TCM to the body computer are connected (have continuity). The TCM plug is 60 wire and the data is transmitted from pins 4 (white/dark green) and 43 (violet/dark green) to pin 14 and pin 6, respectively, of the body computer's 14 pin plug that is black in color. That body computer is located under dash, the TCM is in the engine compartment just inboard of the power distribution center box. The harness of wires and the plugs should be inspected and shaken to divulge any loose/broken connections. Also, the voltage bias (+ and - 2.5 volts) on those two wires should be verified. I believe that the Chrysler diagnostic readout box can be brought to bear on the cause as well. There could be an open or short to ground/battery in bias wires on pin 5 and 44 of TCM plug.
There could be an open or short to ground/battery in the diagnostic bus connector wire.
It is possible that the TCM itself has failed when it gets hot.
The dealer should be able to sort this out in a straightforward fashion if competent and in possession of the proper manual.
The temperature dependence of the problem is relevant. They should do the testing when the engine compartment is heated up (or in the sun). There could be a component in the TCM that is temperature sensitive and fails when heated. Or it could be the plug/socket interfaces that I first pointed to.
So go back with this approach and see how much of the previous misdirected efforts you can get reversed and go forward from there.
Please let me know how this works out.
Roland
