Help me trick the 98-01 TCM

[Agreen]

I just swapped in an AX-15 in my 98 and I'm looking for a way to fool the TCM to keep the CEL off. Yes, I know that manual ECUs exist, but I want to crack this egg as there isn't a lot of information about it.

Here's what I have done:

Initially I had codes p0700, p0753, and p0743. That's right after I cranked it up and read what came on. 0700 is the general TCM fault, 0753 & 0743 are shift solenoids and TC lockup solenoid codes.

I got the 0743 and 0753 to go away by wiring in some 15 ohm resistors to make the TCM think the shift solenoids are still there. All was great. I cleared the codes and went for another drive. I stopped for a coke and cranked it up again and got p0700 and 0705 this time. I probably had 0705 when I cleared it last time, I just didn't check. So I got to brainstorming about what could cause the 0705 code. It's the range sensor code, so obviously it's not seeing the range sensor. I thought initially that since I have the clutch switch wired to the P/N portion of the range sensor wiring that it saw both reverse and park at the same time and it saw a conflicting input (since I leave it in reverse because it's parked on a hill). So I cleared the code, put it in neutral, cycled the key and the CEL came right back. P0700 and p0705. A quick look at the wiring diagram showed that P/N doesn't even make an input to the TCM at all! So there went that theory.

So now I'm confused, since you have to start the (automatic) vehicle in either park or neutral, and it doesn't input either of those to the TCM, so why would it be giving me a code? I only turned the key on with it "simulated" in park and it gave me the range sensor code; didn't even start the engine. So what is it looking for?

To clarify, the two connectors that go from the main harness to the auto transmission are cut. One is solely for the range sensor, the other is the transmission control connector. The reverse switch wires are connected to the reverse signal on the range sensor connector, the clutch switch is connected to the P/N wires on the range sensor connector, and the 3 solenoids are re-wired as resistors on the transmission control connector. The rest of the wires are cut, covered in heat shrink, and tied back in the wire loom. P0743 and P0753 never came back, so it seems like I'm just fighting the range sensor code now, although I figured it would be mad about the input and output speed sensors.

Any help, or did I just confuse everyone?

 

[lawsoncl]

I'm confused on why the tcm is still complaining as well. Like you said, the tcm only has inputs for 1-2,3,D,and R, so what's it bitching about? I don't think the NSS grounds those inputs when not selected, but maybe you need to deliberately ground them?

Without the speeds sensors hooked up, it should think the engine isn't running and your sitting still. You could throw a resistor across the speed sensor inputs just to make sure their not floating and sensing speed, but I'd expect a different error if that were the case. The value wouldn't be critical, and I'd think anything from 100-1k would work.

 

[Agreen]

I got to thinking about it today and I think I want to simulate 12v to the "D" input when the clutch is released.

It seems as if it's not so much a problem with the TCM, but the ECU is now having a WTF moment. The ECU does get the VSS input. So if the ECU is seeing 1) not in any gear and 2) vehicle speed, it's probably getting some weird vibes and on the next key cycle it says "ok something is up here". I figured out that if I turn they key on and clear codes, then start it and drive it, the CEL won't come on. At least, until the next key cycle. Then it throws 700 and 705 again. So over the entire last key on cycle it never saw a forward gear selected and speed increased, it might (and this is just a theory here) give the CEL because of it.

So I'm trying to think of a simple, practical way of sending 12v to pin 22 of the TCM. What I'm thinking now is an LM741 circuit to invert the output of the clutch switch. So when the clutch is pushed in it sends 0v, and when it's released it sends 12v. I haven't really hammered out all the details yet.

If that doesn't work, I'll measure the ISS/OSS resistances and simulate them as well. If that doesn't work... then I'll just be stumped and try finding a manual ECU.

 

[YJeXtreme]

Would love to know more on your fix when completed. My 98 was swapped, no TCM or trans wiring left. Even with a 5spd ECM i get the same codes (pulled one from factory 5spd early 99 model)

I’d love to get the auto trans codes gone so I can fix the CEL bulb that’s missing or blown


Sent from my iPhone using Tapatalk

 

[Agreen]

I don't think you have a 5 speed ECU then. What's the part number on it?

 

[YJeXtreme]

I don't think you have a 5 speed ECU then. What's the part number on it?

It was a chrysler updated ECU with the dealer sticker added onto the xj I pulled it from. I’m recovering from ankle surgery so unable to get into my shop.

Here is a pic of the decal I grabbed when pulling it from salvage yard.

Sent from my iPhone using Tapatalk

 

[lawsoncl]

I got to thinking about it today and I think I want to simulate 12v to the "D" input when the clutch is released.

It seems as if it's not so much a problem with the TCM, but the ECU is now having a WTF moment. The ECU does get the VSS input. So if the ECU is seeing 1) not in any gear and 2) vehicle speed, it's probably getting some weird vibes and on the next key cycle it says "ok something is up here". I figured out that if I turn they key on and clear codes, then start it and drive it, the CEL won't come on. At least, until the next key cycle. Then it throws 700 and 705 again. So over the entire last key on cycle it never saw a forward gear selected and speed increased, it might (and this is just a theory here) give the CEL because of it.

So I'm trying to think of a simple, practical way of sending 12v to pin 22 of the TCM. What I'm thinking now is an LM741 circuit to invert the output of the clutch switch. So when the clutch is pushed in it sends 0v, and when it's released it sends 12v. I haven't really hammered out all the details yet.

If that doesn't work, I'll measure the ISS/OSS resistances and simulate them as well. If that doesn't work... then I'll just be stumped and try finding a manual ECU.

P07xx codes come from the TCM. Is it possible the TCM is getting info over the canbus from the ECM? Maybe, but I'm skeptical. I suppose it's possible you need to un-plug the TCM to clear any codes, since it does have ignition-off power. Maybe it just hold onto that P705 until it sees something that looks correct?

 

[lawsoncl]

Simple relay would be easier and more reliable than a 741, btw.

 

[Agreen]

Info update:

I have noticed a couple of things that are good info.
1) I was initially wrong about the NEXT key cycle thing. It's within the same key cycle that the code has to be cleared for it to stay gone. So if I turn the key to the on position, clear the code, then start it, that's the same key cycle. If I turn the key on, clear codes, turn it off, then back to KOEO, it still gives me the code. This is important because my initial thought about the ECU not seeing a selected gear with speed input... was wrong.
2) I went ahead and wired the remaining wires of the range sensor to the gray connector (where they were before). That leaves 4 wires. 2 (red and blue) for the reverse switch and 2 (black and black/wht) for the clutch switch. The remaining 3 wires got connected with Wago connectors temporarily (if you don't know what a Wago is, check them out) and the CEL still comes on after I cleared it. So that tells me there's something else I'm missing here.


Unrelated to the range sensor, I learned something else about the auto vs manual ECU. If I engine brake, the auto ECU seems to self learn a new IAC position. That means as I come to a stop while engine braking, the engine idles extremely low when I push in the clutch. I did read about others having this issue. One guy in particular thought he was having an IAC issue, then replaced his ECU with a manual one and all problems were solved.

I'm going to keep playing with it until I can find a 98 manual 4.0 ecu. I've already tried sending out quote requests on car-part and haven't heard back yet.

 

[Agreen]

Scratch that. I wrote this after only trying to reset the codes in my driveway. I took a drive and the CEL went away. So right now it's in park on the range selector, the park wires are wired to the clutch switch, and the reverse wires are wired to the reverse switch.

It still does the low idle thing if I engine brake before stopping though. Maybe I'll fool it into thinking it's in drive when the clutch is released. I wonder why engine braking was fine when it was automatic though...

 

[boncrshr]

I'm dealing with a stubborn P0705 and I'm leaning towards getting a new TCM. This vendor is VIN specific, maybe they could help you with the proper TCM?
https://store.allcomputerresources.c...htcmtcutr.html