By Jeff Bach
Did you ever find yourself up a tree and out on a limb with no place to go but to drop off and start climbing all over? That's where I wound up, working on this '92 Cadillac Seville.
It came in with a complaint of losing A/C and fuel data center display after the key was on for about 30 seconds. A code check revealed multiple instrument panel controller (IPC) codes: IO33, IO34 and IO37 were showing up as current codes. There is a specific chart in the factory service manual to follow for these particular codes, which seems logical enough.
I thought I'd write about this one because I've found myself in similar situations on some of the older sedan Devilles that had data problems ranging from bad body control modules (BCMs) to aftermarket radios.
This one had a nice flip face CD player with the equalizer and subs in the back. The owner had a bluegrass CD playing when I got in the car. Somehow I got the idea that this system was in the car when he bought it. I found out later that he had bought the car from the son of a friend.
When something takes down a data line, it can be difficult to trace, especially if it's a party line. This vehicle uses a communication link called an 800 line arranged in a series loop (see Diagram 1).
Between the IPC, diagnostic energy reserve module (DERM), assembly line data link (ALDL), electronic brake control module (EBCM), air conditioner programmer (ACP) and the powertrain control module (PCM), any one of these components can take down the line.
The trouble tree climb begins with me disconnecting the PCM and unplugging the ALDL cover, which cuts the loop in half. I then measure the voltage on the end of the 800 circuit that is remaining. I still had 5 volts but no fluctuation. The only component left for the IPC to talk to from here was the air bag module (DERM) and it was next on the chopping block. The next measurement I took was of pin A8 in the plug from the DERM, which now revealed a fluctuating 3.2 volt signal. The next step had me "jumpering" the two DERM circuits together and then checking for a data signal at the ALDL, which read the same fluctuating voltage signal.
From here, it's off to get a new DERM but the new one didn't help. I backed up and started over to see where I jumped onto the wrong branch and wound up in the same place. Next, I decided to create a known problem and see where the chart led me. I connected the 800 circuits together at the DERM connector and left it unplugged. The chart directed me to the same spot with the same codes. I discovered that with all the components on the line disconnected, I could get the fluctuating voltage out of either of the 800 lines coming from the cluster. But any one module I plugged in would make the line go to 5 volts and remain high.
To do this test, I had to connect the 800 lines together at the disconnected modules to transfer the data beyond them. I got a call from the customer to inform me (more of a confession really) that this car had a salvage title and that it had not worked since he owned it. He was sent here by another shop that preferred to remain anonymous. It was at this point that I went for my scope and looked at the data directly. Figure 1 was my first captured image.
It was here that I finally realized what had me so confused. Even though it quacked like a duck - the voltage signal fell into the chart's "correct" range and was fluctuating - it was not a duck. It was a chicken. It moves up and down in the right voltage range, but I don't think any of the boxes on the line can communicate with it. The voltmeter does, though.
Figure 2 was my next capture and confirmed to me that the source of the data was the point of corruption, which was the IPC. I found a bulletin addressing this very issue that spoke of a water leak and a corroded ground. I checked the ground and assured myself that this was not the problem I was dealing with.
This picture begs the question of which circuit is pulling the signal low. You can see the shape similar to that of an inverted primary current ramp on top of the last part of the signal. Checking the power and ground sources of the cluster with the scope revealed no similar drop in voltage, indicating to me the problem was within the cluster itself. I obtained a good used one and tested both serial data lines before connecting the DERM or PCM.
Figure 3 was the first shot I got of real serial data. I never would have thought that the sight of digital data would bring a smile to my face.
I spent a bit of time searching for this elusive data signal, and once I caught it ... I became enamored with it and had to play a little. Figure 4, for instance, is an overlay of several random snapshot samplings.
I noticed that I could take instant screens of the data stream and get strings of characters that looked similar. I wondered what they meant. I wondered how many bytes it takes to request the defroster to come on. I saved a few of the similar snapshots. Notice the bits to the left and right of the cursors in the waveform. They were similar outside the cursors.
In Figure 5, I separated them.
Looking at this data gives me that feeling of being in a foreign country where everyone around me is speaking a language that I'm struggling to grasp. I wonder if there's a pocket translator for serial bits. I didn't get to charge for the total time I spent on this one, but the enjoyment I got was worth twice the difference between what I got paid for it and the time I spent on it. There should be a formula for that somewhere.
I explained to the customer what had transpired that led to the fix and showed him the scope pictures. He was relieved that he had heat now and could defrost the windshield. He was so happy with the outcome that he said he's bringing me a bottle of his homemade wine and he has a Tennessee Walker that will be perfect for my granddaughter when she's big enough (he thinks 5 is old enough!). I'm not sure I ever did get the whole story on this car. But I somehow got the feeling that I won the prize.
Jeff Bach is the owner of CRT Auto Electronics, an ASA-member shop in Batavia, Ohio. For more information on this topic, contact Bach at (515) 732-3965. His e-mail address is email@example.com and his Web site is www.currentprobe.com.