[FC Saab]

I know it's a vague question, but I've only owned manual Saabs. Lol.

I picked up a new to me '05 Arc 2.3T Auto on the cheap. Previous owner drove it to breakfast and then the car wouldn't start. Had it towed to reputable Saab shop to have the shift cable replaced. Now the car has a P0722 Output RPM No Signal. The tech says he swapped sensors and checked continuity and everything looked ok. I asked if he checked for short to ground or even scoped the sensor to see what signal it was giving and he looked at me with a blank stare.

So, I bought it and got it home. The battery was dead. Not slow crank dead, I mean DEAD. I haven't thrown an amp clamp on it yet but I'm assuming parasitic draw.

Here's my question. Isn't the output RPM on the trans reading the differential? I charged the battery, checked for codes and reset everything. Car started and idled for a while until I would assume it reached close loop and then it threw the P0722 code. I never shifted from park, just idled. So now it's in trans limp mode and shifting to reverse is brutal and its stuck in 4th in drive. What is the output sensor reading? Is it the speed of the diff while moving? I'm going to hope this is a wiring issue. I'll dig into it this weekend, I'm just wondering how these trans work.

 

[ksw92_mk2]

The output speed sensor is a two-wire hall sensor mounted on top of the trans and under the battery tray. It reads the output shaft speed. The output shaft is coupled to the diff by a gear set.

There are two connectors under the battery tray that connect the trans electronics to the TCU. They are known to get crusty and cause problems. I'd start by pulling the battery tray out and looking at that area. The input and output speed sensors on the 5-speed seem be reliable, more so that the CPS on the engine block, so I think you're on the right track by checking connectors and wiring.

 

[Atokad]

From the WIS:

If you have the capability, firstly I would throw a scope on the output and see if the sensor is outputting right. If the sensor is good, check the wiring from the sensor. If it's bad check the wiring to the sensor. If all of that is good, it's more than likely the sensor itself.

 

[FC Saab]

Does the output shaft spin at idle in park? Until you select a gear, then the pressure changes to engage certain clutches. Just wondering what the sensor should see at park.

 

[unclemiltie]

the sensor will see nothing in park or N, the output is not turning.

the sensors are reasonably robust but the good news is that the input and output sensors are the same. So swap them and see what happens. If you get the same error you have a wiring issue. if you have the error on the input you now the sensor is bad. (the sensors are expensive BTW)

 

[FC Saab]

The shop said they swapped sensors. And they also wanted to charge the previous owner $3k for a new trans. I was very skeptical and actually tried to argue at the shop on behalf of the previous owner.

I knew something wasn't right after I cleared the code with my Tech2 and it came back within a minute and I wasn't even driving.

The shop said they had to do some wiring. Im guessing they screwed it up somehow. 8 dont mind wiring issues, I like the challenge. I just wasnt familiar with the operation of this transmission.

I'll dig in this weekend and report back.

 

[unclemiltie]

OK, someone who is saying the this thing needs a new transmission is someone you should never talk to again. This is ENTIRELY an electrical fault and there are no electronics "in" the transmission

there can only be three causes of this code:

• The sensor itself is bad and not sending any signal. If the shop swapped them and the probelm is still there, this is not the problem
• The wiring between the transmission and the Transmission control computer. The sensor is connected at the sensor via a plug. That wire goes to a connector under the battery (in the WIS this is H16-7, the connector pins are 2 and 10. This then goes to the TCM on pins 23 and 33. What I would do is unplug the wires from the TCM and the sensor, use a wire to jump the pins where the sensor plugs in and then measure the resistance between the two pins at the TCM. If it's "zero" then pull the jumper and see if the circuit is "open". If that's the case then check both leads for a short to ground. Then you know if the wiring is good
• The last thing is the TCM itself. you can find them on ebay, they do need a Tech 2 to "readapt" the transmission although if you put it in the car should run. Make sure you find one that is the same part nubmer as yours. That should set you back $100 or so.

But ignore the people who say mechanical problems, for now, since this is electrical.
(if you had a sensor mismatch code I'd be thinking a mechanical issue but you don't)

 

[FC Saab]

The wiring checked out even though someone had been in there before. I even wiggled the joints and everything checked out. I had connectivity to the TCM and no opens or short to ground.

I'm starting to think that there was damage at the connector and when the shop jostled the wires around to change out the shift cable they moved the wires and then the 12v signal to the output sensor shorted to the trans case. They nosed around, found the wire, but it shorted the TCM. They connected everything but still had no signal because the tcm was internally compromised.

There's a couple options on Ebay but I'll check the classifieds here first. I'll report back what I find.

 

[FC Saab]

I got a couple of units ordered. One from Ebay and one from a fellow Saaber in the great Northeast.

I'm trying to understand this wiring to figure out what may have caused an internal failure to the TCM (IF that is the problem). As far as I can tell, the TCM is sending a 12v signal, but the TCM can't see the response. I'm not sure what the symbol is that is parallel to the 100ohm resistor that is internal to the TCM. Anyone have any ideas?

If the rodent damage created an external ground to the trans case, I can't see how that would have caused any harm other than a signal loss. Regardless, the wiring was repaired and I verified everything back to the TCM. Just looking for possible causes of the TCM failure. It was fine one day and bad the next.

 

[unclemiltie]

If you have an oscilloscope you could scope the signal coming back, it should be a square wave.

the TCM feeds voltage to the sensor that then switches it on and off based on the slotted disc in the transmission to carry the frequency, that gives the speed.

If you had a short on the supply side that could blow an output transistor on the TCM so that it is no longer feeding the voltage.
If you had a short on the sensor side, one would hope that the design could tolerate such but if it's got a pull-up on it or something like that you could kill a transistor on the input circuit since it's designed to have a specific duty cycle and a straight short to ground could fry something.

 

[FC Saab]

Final update

I don't have a scope, but it is definitely on my Christmas list this year.

So I had a revelation on Sunday when sitting in church. Does the polarity of a Hall Effect sensor matter? If you recall, the shop checked for continuity on the wires from the TCU. They also checked for opens and grounds. As did I and everything checked out. The pigtail they attached was a pigtail from an input sensor, the wire color didn't match but the connector is the same.

I contacted a buddy of mine who is an electronic control wizard. He said, depending on the sensor type, especially a 2 wire sensor. Reversing the power will cause no detection events. AH HA!

I dug back into the to wiring diagram for both the input and output sensors being sure to identify the orientation of the working input sensor and sure enough, the wires on the output pigtail were switched. I did some cutting and soldering, taped it up nice and tight, put everything back together, cleared the codes and voila. Problem solved.

The tech at the shop recommended a transmission when all it required was a little more due diligence on his part to double check his shoddy work to begin with.

I'm considering my approach to the establishment. Do I simply write a scathing review or do I call into the the owners weekly car discussion live radio show and get his reaction? I'm considering the latter.

 

[unclemiltie]

Glad to see you got it fixed. I knew it was electrical. didn't think about the wires being swapped, and yes that will certainly make it stop working if you put the bias voltage on the wrong terminal.


FWIW there is an outfit that makes a thing that you plug into your iPhone and it turns it into a simple o-scope. not a high speed one but one that you could use to diagnose stuff like this.

Oh, and I'd get my hands on some shrink wrap tubing and then re-solder the joints and shrink them. tape is going to leak at some point!

 

[FC Saab]

I'll look into the phone scope thing. That would be handy.

I did solder, shrink the wires. I double wrapped with some 3M Super77 to help keep the elements and green crusties out.