New ignition switch is $70 and a factory order item (several day wait). The local junkyard wants $35 and two days. So, I removed the ignition switch, disassembled it and could not find anything corroded, broken, melted, or otherwise damaged that would prevent proper function. And, unless I have missed something, ignition power has been present where and when it is supposed to be wherever I have measured it. I re-installed it for the time being at least until I have checked out the effect of removing modules on the CCD circuit one at a time to see the effect on the bus voltage.

 

Bus voltage should be 2.5v on each wire. That's why you wanted to see if you get a response from a scan tool because it's probably an internal module issue. If you can talk to the trans controller with a scan tool and the PCM can't talk to it, the PCM is probably junk as long as the wires in between are good.

 

I measured the CCD voltages at the DLC again this morning and got +2.48 and +2.45. These are a little low, but there is still a difference of about 0.02 or 0.03 volts between the two. I compared these values to those of a friends functioning 99 caravan. On his car, I read +2.47 and +2.50. This leads me to believe that maybe my CCD bus is okay. I'm gonna go back to the engine compartment and make sure all of the wiring from TCM to PCM passes the wiggle test with some load on. Same with any wires from the ignition switch. I am gonna use a test light with a brake light bulb instead of just measuring voltage or continuity. If all the wiring checks out okay, I think you are right, Tizzie, that I might be looking at a new PCM.

 

Sounds like your heading in that direction, good luck

 

I found a wire that would not carry enough current to light my test lamp. It runs from the TCM connector pin 56 to the transmission control relay. With everything hooked up, I measured 12v on this wire with the ignition on. Now I will need to reinstall the battery to test it with my lamp. I hope this could be the cause of the problem.

 

It's fixed! The battery feed to the TCM (pin 56 RD/DB) was the problem. I ran a temporary 18 ga replacement wire spliced in with wire nuts across the radiator cross brace to test it. On start up, after the bulb test, the boxed lights around the gear selector reduced to the one around the current gear and stayed that way. A test drive showed I had a working speedometer and a trip computer. I checked on the interstate and found that I have a 5% speedometer error - (at a steady indicated 60 mph it travels one mile in 57 seconds instead of sixty). I am guessing the junkyard trans must have slightly higher gearing in the final drive than the original trans. I don't think this is fixable, as the tag on the TCM says it cannot be re-programmed without causing it to fail, but it might be worth a call to the guys who sold it to me. I would have been happier not to have spent the $900 on the "new" trans and TCM, but at least I now have a trans with half the miles on it, new seals, new fluid and filter, a new rear main seal, and best of all, everything seems to be working normally.

I'll now need to go back, reroute the new wire, solder it in place, reinstall all the connector covers and reassemble the dashboard. Compared to what I have been doing, that should be fun.

Thanks to jenson611, Tizzy1, master tech, Dan Hetzel and Bill Haas for the valuable help and advice.

 

Gear ratio and tire size for the speedo are programmed into the module and are changable. The calibration of the speedo needle is also possible. Just another use for a scan tool

 

I know this is a few years past the original post but I wanted to get this out there for my experience that I just overcome and where I solved the problem above.


My 2000 Caravan had the same symptoms as BILLBUTTERMORE.


Searching the net I found the usual suggestions without reading codes was to change the input & output speed sensors, I did that with no change. The next suggestion was to change the solenoid body, I did that with no change to the errors. I had tried my code reader but I got a code 905 CCD not able to communicate with the TCM, I thought it must be the TCM so I changed that with no change. I then took it to a shop thinking their reader might work better than mine but theirs gave a P1698 which is the same as 905 CCD not able to communicate with the TCM. So I thought there must have been an issue with the CCD lines. I rung them out and found there was good continuity and there was no shorts to ground. I then found a test for the transmission at this link
http://transmissiondiagnosticinfo.com/41te42le42rletests.html


In doing this test I found that pin 56 at the TCM only had 0.6 volts rather than 12V. I tested the source at the fuse panel and found there to be 12V so there must have been a break in the line to the TCM. I cut the wire near the TCM and run a new fused wire from an always hot wire in the fuse panel to pin 56 on the TCM - Tada - it all worked !!! My code reader communicates, the engine light when out, the speedo works and it wasn't stuck in limp mode.


I also wanted to mention that even though the solenoid body didn't solve the main problem it did solve the shifting issue I had where the transmission was slow to shift down when rolling to a stop and then you hit the gas again, it would do a weird lunge and then shift down and go, it was slower to shift down in the cold. This is now GONE !!! yes !!!

 

I wanted to add that I was seconding the discovering of the missing 12V on pin 56 in hopes that it might help others the check there as well even though it can be a pain if you don't have test tools to check this out. If I would have read BILLBUTTERMORE's post initially I would have checked there a few days sooner and saved me a bunch of other grief.