Boggy Territory

By Richard McCuistian

Aging vehicles with quirky electronics are the spice of a diagnostician’s life.

1995 Eagle Talon

112,234 miles

2.0L Engine

41TE Transaxle

“Check unit won’t start.”

Old Electronics

            Right after computers first began to appear on vehicles, my AC Delco parts supplier gave me the opportunity to attend a Computer Command Control class in Houston, Texas at the GM training center.  It was 1981, and the questions my classmates asked then-GM instructor Ellen Smith went something like this:

            “What’s going to happen when one of these cars gets a hundred thousand miles on it?  Won’t it be a disaster trying to keep all this computer stuff straightened out?”  Ellen replied (quite truthfully) that a properly maintained computer-assisted engine would run as well at 100,000 miles as it did when it was new.  She cited several specific examples of cars she knew about that proved her point, and we’ve all seen thousands of them since. The questions the mechanics fired at Ellen continued:

            “Will it start if the computer craps out?”  Ellen disconnected the computer on the Buick trainer vehicle and turned the key.  The car started and ran, albeit with the mixture control metering rod in the full rich position and without proper ignition timing control, but the car could be driven that way.  Granted, today’s cars can’t.

            But Ellen believed in computer controlled fuel and ignition, even when the independent shop techs like me weren’t so sure it was a good idea to let “HAL” the computer handle that stuff, especially when he killed off all but one of the human cast and wouldn’t even open the pod door for the sole remaining astronaut on 2001: A Space Odyssey.

            In the here and now, we know that today’s fuel economy and emissions standards would be virtually impossible to meet without sophisticated electronics calling the shots, and they’ve spread their tentacles into virtually every part of the vehicle.  Some new Hyundai models even have electronically controlled engine mounts. 

But what about those high mileage cars like the Lexus SC300 I wrote about a few months back that came in blowing gasoline steam out the tailpipe because of a computer problem nobody wanted to tackle? We managed to fix the fuel problem with a reman PCM ($400+), but that same tired old Lexus SC300 came in a couple months later needing a $450 power steering hose. 

And what about the thousands of unbillable labor hours veteran techs like me have put into diagnosing older electronically controlled fuel systems for on cars that weren’t worth the time and didn’t deserve the effort because the customer wasn’t willing to pay anyway?  Where should the line be drawn?  And how much money should a willing customer be expected to spend on a car that has drifted past the point of obsolescence? 

Expensive Antiques

In 1988, right after we got the Renault/AMC franchise at the dealer where I was working, the first vehicle I checked in my new capacity as an AMC tech was a dead 1981 Renault 18i. We didn’t even have any books on those old cars among the literature that came from the dealership we bought, and I was pretty spooked.

Well, I used a little common sense and got the spark popping by replacing the ignition module (a mid-seventies style GM HEI unit that was mounted externally rather than in the distributor) and then found that the old Bosch fuel injection computer (mounted under the seat) was fried because somebody had baptized it some time earlier by leaving the windows down in a driving rainstorm.  The cost of a replacement engine controller was an eye-popping $1100.  In my eyes, the car wouldn’t have been worth fifty bucks driving down the road, but to my astonishment, the owner okayed the repair without a moment’s hesitation.

Those situations terrify me, because I don’t know what else I’m going to find wrong after replacing a part that cost 20 times what the car is worth, or how many times the junker will boomerang on me. As it was, the 18i owner paid the bill, drove away, and I never saw him again.

Then there are vehicles like the late eighties/early nineties Eagle Summits, Talons, and Plymouth Lasers, those Mitsubishi-made units that were loaded with so many high dollar fuel control system parts that the sum total cost of all the fuel system parts alone was more than the sticker price on the car.  Does there come a time when we should balk at taking a customer’s hard earned green stamps to resuscitate a tired old machine that might not make it 500 miles before needing more repairs?  Distributors for Asian-designed units generally run between $300 and $500.

Hooked by a Talon

This purple Talon wasn’t quite the junker that the18i had been, even though it was twelve years old, give or take a few months, but it was dead as a doornail. The owner didn’t seem to be in too much of a hurry for the car; she called me when we were swamped and working our way through final exams. Her Talon had died and had been sitting for awhile, and it would have to sit awhile longer before we could have a look at it.  As far as I was able to tell when she had it towed to us, nobody had touched it.

This ’95 Talon was one of those peculiar transitory Mitsubishi OBD II compliant units that takes a special adapter cable if you plan to talk to the enhanced room in the PCM.  The Genesis has a special “Y” cable that connects to the OBD II DCL as well as the Mitsubishi data link, which is mounted about six inches from the OBD II connector.  The OBD II generic room (which sometimes has codes that don’t show up in the OBD II Enhanced room) can be accessed via the now-universal 16 pin hookup, but the Talon wouldn’t talk either way.  To a tech like me that likes to grab data quickly and spring into action, this is annoying beyond measure.

In a no-comm situation like this, it’s imperative to determine whether the PCM has power and ground at the appropriate pins. This PCM had those feeds, but there was no ignition and no fuel pulse forthcoming. Could it be a bad crank position sensor?  Maybe, but that alone wasn’t likely to cause a no-communication problem.

This was a set of circumstances where a scan tool isn’t much help, but it was obvious that the PCM had died.  There was no reference voltage anywhere either and with the idea that a shorted three-wire or Hall Effect sensor might be neutralizing the reference voltage along with everything else the PCM was trying to do, I found that disconnecting the pertinent sensors did no good. 

Generally, my experience has been that the whisper of current provided by reference voltage circuits does no permanent harm to the PCM, it simply causes a no-start.  For example, it isn’t all that unusual to have the DPFE sensor short the 5 volt reference line to ground internally on a Ford.  It kills the car along with scan tool communication, but doesn’t hurt the PCM.  I worked on a police Crown Vickie that had the reference voltage wire feeding the fuel tank pressure sensor scratching on a bracket near the fuel tank and it would stall during right turns.

Well, even with the suspect sensors disconnected (crank and cam were still in the loop), we still had no spark and no fuel pulse. Removing the PCM from the Talon, I plugged it into a 1995 Dodge Neon I use for a trainer car because the pinout was almost exactly the same, the difference being that the Talon has a separate computer for the transmission and the Neon PCM controls the torque converter.

With the Talon PCM installed (I didn’t try to start the Neon), the scan tool wouldn’t talk to the Neon either.  That was another indicator that the Talon PCM was cooked.  I had a bad feeling about trying my Neon PCM on the dead car.  A replacement PCM was only $200 from the dealer (the few and far between salvage yard units I could locate were just as expensive), but the PCM would be on backorder for about three weeks before I could get one.

Once I had to replace the PCM on a 1998 Dodge pickup because a body shop had done some welding on the body of the vehicle, and I made a quick check of that one by connecting a Jeep engine controller that I had in my box of junk.  Once again, the pinout was slightly different, but the communication element was all I was checking.  The Jeep controller would talk; the Dodge controller was dead.  Viola!

Fresh Fried PCM

Installing the replacement PCM, my students managed to start the Talon, but before I could get to the car, it had died again.  In a perfect world, we would have connected the scan tool as soon as the PCM was replaced, but a couple of my guys jumped the gun and allowed this angry little Eagle to eat the new PCM for lunch.  This was turning nasty.

Have you ever been at this fork in the road?  What now, especially since no salvage yard in our part of the state had an engine controller for a 1995 Talon?  Would an exhaustive check of every pin for power, ground, and/or shorts to other wires be a viable test?  Maybe, but what if the problem was intermittent?  After all, the car did start and run for a few minutes before it cooked the replacement PCM.

Well, it just so happens that I had a very expensive piece of equipment that was designed to find problems like this, and I was about to put it to work.

First, I called the dealer and re-ordered PCM number two.

Second, I put in a call to Snap-On Equipment sales, where I rented the necessary cable ($50 a day) to connect my now-outdated Snap-On/Sun Dynamic Data Collector (originally produced by Edge Diagnostics) to the Talon.

This pricey little box (called a “pod”) cost my department many thousands of dollars and is the size of a small suitcase. It came with an interface card that had to be installed in a Windows PC with an ISA slot.

No modern PC comes with an ISA slot.  That’s how old this DDC unit is. 

The DDC pod doesn’t pull DTCs or talk to the serial link at all; it simply connects between the vehicle’s PCM and harness just like a breakout box. That’s why I had to rent the $1000 cable for two days.  The cables are vehicle specific, and since they haven’t done any updates on this equipment since you can’t get one for just any and every vehicle.  The DDC checks each individual pin and sends the resulting information via a special cord to the PC-mounted interface card, which operates through the PC’s motherboard and video card to display the data on the monitor in a colorful fashion using special software. 

With the Talon selected, the software designers had been diligent enough to give the program full knowledge of what each and every pin on the 1995 Talon PCM should read in ohms and volts.  With that data at its disposal, the DDC performed what the software program calls a “Sweep Test” on all the circuits connected to the PCM while I stood back and folded my arms.  When it was done checking resistances, the DDC told me to turn the key on and upon checking the volts at each PCM pin, the DDC software flagged all the reference voltage lines in red; in effect, there was no reference voltage to amount to anything.  Pin 44 feeds 9 volts out to the injectors, and that pin was dead as well.

This piece of information was obviously an effect rather than a cause.  The source of this disastrous failure remained obscure, masked by the lack of voltage at all the sensors.   

I took a chance and plugged my 1995 Dodge Neon controller in so as to (hopefully) regain reference voltage, then I re-triggered the sweep test.  The results looked a lot different this time around.  I had reference voltage all right, but the TP sensor input line was dead.  Disconnecting the TP sensor, and doing another test, I found that the TP sensor line read over 4 volts with the sensor disconnected.  Had the voltage remained low, I would have figured on a wiring problem. As it was, the TP was the only faulty input.  Why it might destroy the whole circuit within the PCM remains to be figured out, but there it was.

Bingo, We Scored!

 The TP sensor was $85, and with the new PCM installed, everything was peachy keen. We drove the car around quite a bit to make sure it was okay.  The owner paid for one PCM (Daimler Chrysler graciously warranted the first one) and a TP sensor.  The encounter was a victory, but finding the problem wouldn’t have been so tidy without the DDC.  As it was, I spend $100 of somebody else’s money just to rent the cable. 

I hadn’t seen a TP sensor fry a PCM that way, but this time around, that appears to have been the case.

We managed to get this sporty little Talon going for less than three hundred dollars – the territory was as boggy as any I’d seen, but we managed to pull this one out with minimal damage.

                                                                                                                                                                                                     R.W.M.