Ark's City Goblin #187 (2008 SS/TC donor)

[Ark :D]

I'm with Brian on this one. As soon as I seen the POWER STEERING message I had flashback to when I locked up my ECM with HP Tuners. It acted the exact same way! No crank and no start. No communication. You can troubleshoot starters and relays, etc all you want but that ECM or BCM is not going to allow that engine to run.

I'm not disagreeing with you, I feel like I have an ECM problem here. But my question is, how? The engine ran fine with the MAF sensor unplugged, before I ripped the donor apart, and I've not done anything to the ECM since then.

 

[Ark :D]

If your scanner s not reading anything on the data lines, something is wrong there and that's what you need to forum on. I know I sound like a broken record, but makes sure you have good grounds, especially on the grounds to the ECM.

Side note, I've seen articles that with only 4-5 wire to the ECM and HP Tuners can communicate with it and read/write. Positive, ground, TX, RX, and something else. Point being, there is not a lot required to be correct to be able to read data.

Loud and clear. I will double-check the grounds around the ECM tomorrow.

 

[Briann1177]

The power steering module has a 120 ohm termination resistor across the data lines. I dont think you can loop together that way and have them work. It might work, but i wouldn't trust it.

I also seriously doubt you fried your ECM.

 

[Briann1177]

The brown/white MIL wire is not something you really mess with. It's one of the wires that gets extended when you cut out the body-to-dash connector. Make sure you have your 16-pin engine-to-body connector hooked up. Even if you forgot about that, I didn't see any circuits in that connector that would cause the issues you're having.

I discovered early on that it's pretty easy to not seat your ECM connectors down all the way. When I latched them down the first couple of times, one half of the connector would go down and the other half went up. I'm sure you've already been down that path 10x already.

 

[Rttoys]

if you read any trouble chart GM gives you for diagnosing, it will start with “step 1, with key on, is the CEL on”

99.99% of the time, the answer is “yes”. This is a simple function check to make sure the ecm/bcm is working.

your Answer is “no”. So we need to start from there

your gas and other gages don’t move either, when you key up, along with fuel pump or throttle body. I’m sure there is one wire not connected.........but which one???

this was my worse nightmare, when building my harness.

 

[Ark :D]

Did the 10-minute disconnect of the battery tonight, as well as double-checked my body-to-engine connection (it was solid) and ECM connectors (both seem 100% connected solid).

I agree Rttoys, and electrical spelunking through diagrams is solidly NOT a strong point for me.

Hopefully double-checking my ECM ground(s) tomorrow bears some fruit.

 

[ctuinstra]

Loud and clear. I will double-check the grounds around the ECM tomorrow.

Sorry, not trying to come across that strong.

I don't think there is anything wrong with your ECM either, just the communication and most likely in the wiring because that is what has changed since the donor ran. It's all too easy to get the green or the tan wires mixed up with others. Hang in there, you'll get it running!

 

[Ark :D]

I didn't take it that way at all! You're fine. I just appreciate all the help from this forum. I know it'll get there.

 

[bradr]

Attached you will find a good single page databus diagram I stole from someone else. I would use a meter to test both data bus wires end-to-end (diagnostic connector to ECM). If no continuity, then move the probe from the ECM to the next closest item in the chain - in your case, it should be the BCM. If no continuity, then move the prove from the BCM to the PSCM.

Simplified layout of what your wiring should now be:

DLC<--->PSCM<--->BCM<--->ECM

You might logically think the "Power Steering" message is an issue with communication between the PSCM and the BCM, but it is more likely an issue between the BCM and ECM - when the PSCM does not get the messaging it needs from the ECM, it throws the "Power Steering" message.

 

[Briann1177]

I don't think you can measure resistance end-to-end since you'd be measuring through all of the involved circuit boards. From my understanding the high speed GMLAN is a serial network.

 

[bradr]

I don't think you can measure resistance end-to-end since you'd be measuring through all of the involved circuit boards. From my understanding the high speed GMLAN is a serial network.

You are correct, the Cobalt uses differential CANbus and a proprietary single ended CANbus, both of which are forms of serial networks. When speaking about the CANbus in question, there is no circuitry isolating one segment of wire from another in the Cobalt - GM just chose to pass the CANbus "in and out" of each module to add complication with little if any up-side. You can most certainly measure the resistance end to end as it is just a "Wire". There are vehicles that do isolate network segments into broadcast domains using gateway modules. -the Cobalt is not one.

Background:

Differential CANbus is simply a parallel network architecture that uses a bus topology. The normal layout is each module taps into the bus with two wires. GM decided to actually pass the wiring through the various modules. If you use a meter, you will see there is a dead short between the in and out of each CANbus lead on all the modules. This implementation has little advantage and many disadvantages.

You can choose to re-arrange the network layout or even abandon the "in and out" method used by GM all together and the bus will still work. Ideally you still want to maintain a network where you have a terminating resistor at the beginning and end of the network. In the case of the Cobalt, the terminating resistors are in the PSCM and the ECM. The diagnostic connector does not have any active electronics so it does not count, and the only other module the Goblin uses is the BCM and in some cases, the RPD. So these two modules need to ideally stay in the middle of the network.

Bold below indicates a module with a terminating resistor.
DLC<--->PSCM<--->(RPD)<--->BCM<--->ECM

 

[Ark :D]

I have full continuity from end to end. So that's not it, unfortunately!

I also checked the three large ground pins in the ECM connector, all three test good to ground.

Today I am planning to check every connection to the engine block (just to be sure; I don't think this is causing my issue).

 

[bradr]

I have full continuity from end to end. So that's not it, unfortunately!

I also checked the three large ground pins in the ECM connector, all three test good to ground.

Today I am planning to check every connection to the engine block (just to be sure; I don't think this is causing my issue).

If you have continuity end to end, I'm guessing your issue is either one set of the data leads might have flipped polarity somewhere, the ECM is not powering up or some module failure. You may have already, but I would ensure you have the following at X2 of the ECM:

• Ground at pins 1, 2 and 4 (Black/White)
• Power with the ignition on at pins 3, 5 and 6 (Pink/Black - Powered by the ECM/PCM 15a Fuse in the fuse block)
• Power all the time at pin 56 (Red/White - Powered by the ECM/TCM 10a Fuse at the BCM)
• Power when the ignition is on at pin 54 (Pink - Powered by the ECM/Trans 15a Fuse in the fuse block)
• Power when the ignition is on at pin 41 (Brown - Powered directly by the BCM, X3 B5)

EDITED 6/20 to fix error

 

[Briann1177]

Thanks for the info, Brad. I'm guilty of not following this too, but keep in mind that just checking for continuity in a wire might not be good enough. A more thorough check is to ensure there are no shorts between the wires, no shorts to ground on each wire, and no shorts to battery voltage on each wire.

Also make sure you do a good physical check of your connections. Look for things like bent pins, recessed pins, missing pins, etc. I had two instances in the big multi-plugs in the fuse box and BCM where the springy part of the female contact that engages the male pin was bent and it caused me a lot of problems.

 

[Ark :D]

Ok, now I think we are cooking with gas. I do not have power, whether key is on or off, on pins 3, 5, and 6 of the ECM plug.

However, I do not have a fuse specifically labeled "RCM/PCM" in the fuse block. I have a 20A "PCM/ECM", and a 15A "ECM/TRANS". Both of these fuses (actually every fuse, not just these two) are good.

 

[Briann1177]

You need to pull the fuses and make sure you're getting voltage to them.

 

[bradr]

However, I do not have a fuse specifically labeled "RCM/PCM" in the fuse block.

That was a typo, should have been "ECM/PCM"

The wires you dont have power on are the main power wires for the PCM - fix this and I think you will be good.

 

[Ark :D]

If you have continuity end to end, I'm guessing your issue is either one set of the data leads might have flipped polarity somewhere, the ECM is not powering up or some module failure. You may have already, but I would ensure you have the following at X2 of the ECM:

• Ground at pins 1, 2 and 4 (Black/White)
• Power all the time at pins 3, 5 and 6 (Pink/Black - Powered by the RCM/PCM 15a Fuse in the fuse block)
• Power all the time at pin 56 (Red/White - Powered by the ECM/TCM 10a Fuse at the BCM)
• Power when the ignition is on at pin 54 (Pink - Powered by the ECM/Trans 15a Fuse in the fuse block)
• Power when the ignition is on at pin 41 (Brown - Powered directly by the BCM, X3 B5)

Ground at pins 1, 2, and 4 CHECK
Power at all times at pins 3, 5, and 6 FAIL
Power at all times at pin 56 FAIL
Power with ignition on at pin 54 CHECK
Power with ignition on at pin 41 CHECK

 

[Ark :D]

That was a typo, should have been "ECM/PCM"

The wires you dont have power on are the main power wires for the PCM - fix this and I think you will be good.

I will yank the harness and get to work, then.

 

[Briann1177]

Make sure you post a first start video.