Vehicle problems that change without notice

Back in the writeup area, a QC ticket was written, and since the transmission guy was so busy and I was related to the owner. I pulled the car in onto a lift, raised it and found that one of the transmission cooler lines had cracked at its roll crimp – there was absolutely no way the trans guy had caused this, and the timing of this failure was one of those Murphy’s Law reverse miracles that tend to make us look like incompetent boobs. I fixed that leak quickly with a brass fitting and a double flare and refilled the transmission with fluid to get the customer back on the road. Every good mechanic has done quality work on a vehicle only to be dealt the dirty hand of a repeat repair that wasn’t his fault.

Granted, sometimes the subsequent failure is our fault.  One of my guys removed and reinstalled the transmission in an S10 Blazer and the owner came back complaining that the hatch release button on his dash was inoperative.  It turned out as we analyzed the wiring schematic and did some hands-on rechecks that the Transmission Range sensor connector wasn’t seated good on that one. The ground for the hatch release solenoid travels through that switch to prevent inadvertent hatch release while driving down the highway.

Smooth and straightforward repair jobs are the ones most of us like the best. Those are the routine work orders where there are no surprises on either side – simple “Condition, Cause, Correction” flow.

Such a case was the 2007 Mitsubishi Raider that had been at another shop for an engine skip under load. The owner said the other shop had replaced the ignition switch at a cost of $400, but his engine skip remained the same. In my way of thinking, cheap and easy is the best first move, so we yanked the spark plugs to find the originals still in their holes.  All it took to fix that truck was 30 minutes and a simple set of spark plugs. Then there are the tough but obvious ones, like the Hyundai Santa Fe with an inoperative air conditioner due to a dead evaporator sensor. That’s a cheap part and about eight hours of tedious work, but it fixed a vehicle that had been at the dealer four times for the same concern and had never been fixed.

But then there are those jobs I have dubbed “circumstantial land mines” that blow up in our faces. We perform a simple repair and the job goes south to the point that it feels like a T-Rex has walked into the shop with his beady little eyes on our bottom line. We want to find out what went wrong, and fast. Sometimes situations blind-side us that even make us think we’re incompetent. This job turned out to be one of those.

To Drive For a Year
The foreign exchange student who bought this car at the local GM dealer got it for a pretty decent price to drive for the year she’ll be here. The A/C was inoperative and the “LOW COOLANT” message was constantly displayed on the message center, but other than a couple of burned out stop lamp bulbs, this 1998 Buick was in pretty good shape.  We gave the car a good once-over right at the end of that work week, checked the coolant concentration and the availability of a coolant level sensor and then attacked the A/C for a prelim.

We drew the air out of the A/C system because it had gone completely atmospheric since the last time anybody had injected any cold gas. Because she had been told when she bought the car that the A/C was leaky, we were surprised that it held 30 inches vacuum as good as it did. Next, because were almost out of time that week and she needed the car for the weekend, we injected some UV dye along with the refrigerant charge, felt it blowing cold and let her take the car with instructions to bring it back the next week for a follow-up.

When the Buick returned the following Monday, the “LOW COOLANT” message had been joined by a “TRUNK AJAR” message and the A/C’s refrigerant charge had mostly gone away. We replaced the coolant level sensor and took care of the Low Coolant message, but the Trunk Ajar warning remained.

As for the lost refrigerant, our black light lit the bottom of the A/C compressor up like it was radioactive (no surprise on one of these), and so we recovered the 0.3 pound of refrigerant that hadn’t yet leaked out and followed up with a 150 pound dry nitrogen pressure test. There were bubbles between the compressor housing shells, so we replaced the compressor, the orifice (which was clean) and the accumulator.  We drew the air out and charged the system, ready to feel nice cold wind at the register, but the compressor wouldn’t even engage. While we were investigating that, one of my guys noticed that the temperature needle was dangerously near the red line and the cooling fans weren’t working either. The loss of compressor operation, the dead cooling fans, and the trunk ajar message had all come about over the weekend, but none of these symptoms had existed during our previous service.  Developments of this nature always feel like a smack in the face, especially when the customer was expecting the car back that same day without a hitch, and inoperative cooling fans is an engine-destroying hitch!  One simple problem had been replaced by two larger problems that required stronger troubleshooting techniques. 

I put a couple of guys to work on that trunk ajar message while we began to gather data on the A/C no-engagement and the unresponsive cooling fans, but all they found was a normally operating trunk ajar switch, along with good voltages and circuits, so we closed the trunk and put that trunk ajar concern on the back burner.

Holding the Car
I had to call the owner and explain that we couldn’t release the car, primarily because of our concerns that she might damage the engine sitting in traffic with fans that wouldn’t spin. Investigating the fans for opens using my test light method, I typically remove the relay, find the terminal feeding each fan, connect a test light between B+ and each terminal and it should light up. Then I slowly turn the fan in question and watch for the light to go out. If it ever does, the fan must be replaced  Both of these fans passed that test. I did notice that at some point in the vehicle’s history one of the three fan relays had been replaced.

With a two-way talking scan tool, we told the PCM to turn the fans on, and it did – on both speeds. And with the ECT sensor disconnected, the PCM would also activate the fans, albeit at low speed. What we now knew was that everything was in place (drivers, circuits, good relays, good fans) so that the PCM had the ability to operate the fans but for some reason it was choosing not to. Using a 50k potentiometer connected to the

ECT’s yellow wire, we started the engine and, watching the scan tool, we dialed in a temperature of 267 degrees, but the PCM still wouldn’t activate the fans, even though it was seeing engine-destroying temperatures on the ECT PID. What manner of madness was this? 

As for the A/C, the scan tool showed that the Control Head was sending a request and that the BCM was receiving that request, but that’s where the request stopped. The BCM wouldn’t message the PCM to turn on the A/C – that’s what the BCM was telling us via the datastream, and the PCM’s datastream verified the lack of a request.

This isn’t a voltage deal you can measure or even interpret with a scope. It’s a Class 2 message transfer, which is kind of hard to track with anything but a scan tool. We obviously were dealing with a software issue of some kind. We disconnected the battery terminals and touched them together in an attempt to reboot everything, but tono avail.  I even went so far as to reflash the PCM to its 

latest calibration, but that did nothing at all to correct this concern. Truth to tell, I didn’t expect it to, but a guy can hope, right?

Identifix was our next stop, and posts on this problem reflected issues with the BCM, PCM, wiring, etc. There was no silver bullet in that set of posts. I checked with the GM dealer to find that a replacement BCM runs about $400 – a non-starter as far as I was concerned, so I called a salvage yard and found that they had one on hand for $45. I swung by and picked it up, but when it was installed the next day nothing had changed, either on that LCD scan tool screen or on the vehicle itself. The fans and the A/C were both dead and the same scan tool data prevailed.

A talk with the Identifix hotline guys took me through a battery of tests, one of which

was the exercise of comparing the ECT voltage using a meter and a sensor chart the hotline guy had in front of him with the temperature reported by the scan tool – those numbers didn’t quite line up, but I knew the whole time I was jumping through these hoops that we were barking up the wrong tree.  These two problems had to be related, but how?

Ignorance is NOT Bliss When I have a really tough one like this, I like to work on it when there is nobody else around, and this was one of those. What I discovered during those quiet moments alone was that the “BCM” the salvage yard had handed me was, in reality the HVAC Programmer. The BCM has pink connectors, is about the same size and is mounted right next to it right up there behind the glove box. When I went back by the salvage yard, I handed them that HVAC programmer and they handed me a pink-connectored BCM from the same car. When I plugged that one in to the Buick in my service bay, everything came online – the A/C worked, the cooling fans began to operate normally and the trunk ajar warning went away to be replaced by “Change Oil Soon.”  

What I re-learned here I knew already – you can’t count on the “Description and Operation” section of whatever manual you’re using to explain the complicated relationship between the various modules. Sometimes the best and most reliable information comes from tough experiences, clear reasoning and retracing our steps when we find ourselves at a dead end.

At this point, let me digress to say that people with a lot of GM trench experience have a better feel for concerns like this on a GM car because they understand the relationships between the modules better than those of us who work on anything and everything every day – and while this car is sixteen years old and a GM dealership guy could probably have found the concern in short order.

After the oil change was done and the bill was paid, that Buick driver got her wheels back with everything working very smoothly, but then two weeks later she was back with harsh transmission shifts and a stalling concern.  She also wanted her brakes checked. The transmission had harsh 2-3 and 3-4 shifts, it would stall sometimes during parking lot maneuvers, and the MIL was on with a MAF code we hadn’t seen previously on this car. The datastream showed MAF readings about 1,000 hertz lower than accurate, and after inserting a $200 MAF sensor from the parts store, the transmission cleaned up its act, and there was no more stalling.

As for the brakes, she got a full set of pads and all the rotors measured and machined, and the pulsation was gone. Maybe this ride will last her a year without needing any more work. We all hope so.