Right Up To the End

By Richard McCuistian

After more than an hour’s worth of painstaking troubleshooting, we had to stop just short of the goal.

1991 Eagle Summit

137,000 miles

Cranks but won’t start

The little white Eagle came gliding into our driveway on the hook as a no-start.  The customer said most of the time it was dead in the nest, but when Danny went to the service lot to retrieve the little car late one afternoon, he turned the key and found the little engine that could (start) under the hood.  It actually fired up for the drive into the shop, but ran with an engine skip that seemed to be compression related, since it spun like one of the four pistons was squeezing almost no air at all.

But the skip wasn’t what the customer wanted checked.  First, the car needed to be repaired to start consistently, then, if there was any money left, we could go after the engine skip.  Danny drove the little car into his service bay left it there all night.  When he came back the next morning, he found that it had decided show us what it could do.  It spun rapidly, regularly puffing its one weak cylinder, but it wouldn’t fire up at all. Problem duplicated!

Beginning With Basics

            In the old days it was easy to condemn the dastardly solid state ‘black box’ when a problem like this landed in our laps.  But most of us have become more cautious in the years since on-board computers started showing up under our dashboards.  How many of us seasoned professionals have whipped a computer out of the box, popped the PROM into it, and plugged it in, only to find that the problem was a bad ground or a blown fuse?  Then there are the intermittent problems that make us all want to lock up and go on vacation for a few days… Fortunately, this Eagle was dead in the nest at this point.

A common sense investigation under the hood revealed that there was no spark at the plugs and no injector operation, although fuel pressure was acceptable and there weren’t any visible harness or hose problems.  Since the injectors weren’t getting any signal and there was no spark available at the plugs, it was time to do a little reasoning.  While the PCM requires reference pulses from the Crank and Cylinder Identification  (CID) sensors in order to operate the fuel injectors at the proper time and speed, the expensive little distributor (which contains both sensors) had already been replaced by somebody, but to no avail.  While we may not have been ready to condemn the PCM yet, it was definitely time to dig out the scan tool.

Scan Tool Data

            Danny’s tool of first choice for this job was Chrysler’s second generation Diagnostic Readout Box, dubbed “DRBII” by the Mopar folks.  This OTC-built machine has always been handy to have around for troubleshooting Chrysler products, but it won’t handle anything newer than a ’94 model.  We have three DRBII units at our shop.  On the Eagle Summit, it was necessary to use the OTC Mitsubishi adapter as an interpreter between the American-made scan tool and the Japanese-speaking Powertrain Control Module (PCM).  The diagnostic connector on these little cars is conveniently mounted right next to the fuse panel under the left side of the dash. 

With the adapter connected and the scan tool booted up, Danny switched on the key and found that the PCM wouldn’t talk.  The only message he could get from it in the beginning was a ‘NO RESPONSE’ message on the DRB screen.  This is annoying when you’re expecting to find some useful data displayed there, but it provided a valuable indicator that things weren’t what they should have been at the PCM.  This can be the result of broken wires between the PCM and the diagnostic connector, or it could be as simple as a lack of power or ground at the PCM connector.

Voltages

            When troubleshooting a problem where the PCM doesn’t fire the injectors or the ignition coil, there are some choices to be made as to what needs to be checked first. 

Obviously, the PCM needs system voltage fed to it from the electrical system, in this case from the what Mitsubishi calls the MPI relay.  It uses this voltage to fire up its internal microprocessors and provide power to the drivers it uses to operate actuators on the engine.

Among the most important of these actuators are the four injectors.  And it doesn’t take an electrical engineer to figure out that none of this internal processor hardware will operate correctly without a proper ground connection.

In addition to all this, the part of the PCM circuitry that generates reference voltage for the sensors obviously needs system power to operate.  Making a trip to the wiring schematic and ferreting out the power and ground pins is a viable choice at this point, but years of experience have taught us that a simple reference voltage check at the Throttle Position sensor can provide those of us who are ‘in the know’ some really useful information about what may be happening inside the Powertrain Control Module. 

Most TP sensors have three wires, all leading back to the Powertrain Control Module, and these wires are usually fairly easy to access at the sensor.  One wire is a 5 volt reference feed, accompanied by a reference ground, and finally, the TP signal wire itself.  If the reference voltage rests at or near zero, then system voltage may not be available at the PCM connector for some reason.  If the reference voltage is much higher than 5 volts, then the car’s electrical system ground feed to the PCM may be absent or making poor contact somewhere in the circuit.

High Reference Voltage↑

In this case, the reference voltage was floating around from what it should have been (5 volts) to as high as 11 volts.  It didn’t take long to research the shop manual connector pin out…

Here’s the pinout of the connector in question.

            ..and find that the PCM ground terminal was clean and fitting tightly on its respective pins. One fairly easy way to check the integrity of the ground feeding the PCM connector is to simply measure the ground wire at a point near the PCM connection   If more than a tiny trace voltage (0.01 or so) is present on the ground wire, then there must be either high resistance or an open somewhere in the ground circuit. Furthermore, spinning the engine while reading voltage here can produce some eye-opening fluctuations here even if the measured static voltage is acceptable.  We found acceptable ground voltage at the PCM in both cases.

  Poking around under the hood, we seemed to find that disturbing the engine compartment wiring harness near the big neutral safety switch connector below the air cleaner was causing the reference voltage reading to fluctuate, and as the voltage floated up and down, we could hear PCM-controlled relays and solenoids clicking here and there. This might have been a fluke, however, since after the first time we were unable to reproduce the symptoms by manipulating the connector. This little anomaly made for an annoying sidetrack and would continue to cast a cloud of doubt over the rest of our troubleshooting exercises on the Eagle.

When the reference voltage slithered into the 5-volt range and stayed there long enough, the scan tool would begin to communicate with the PCM and the engine would fire up when the key was turned.  Exhaustive testing, inspection, and manipulation of the engine harness and its connectors failed to produce any evidence at all that the problem was in the harness.  The one thing that did seem consistent was that when the ignition was switched off and back on, the reference voltage would change, even if nothing anywhere was moving. For the most part, yanking on harnesses seemed to make no difference at all after that false alarm at the safety neutral switch.

Drawing Conclusions

            Since the PCM had consistent power and ground, yet the reference voltage was floating and PCM communication was sporadic at best, it appeared evident that a replacement PCM would be in order.  One of the aggravating things about this sort of problem is the fact that the only way to verify that the PCM is causing the trouble is to plug another one in.  As a matter of fact, there are numerous shop manual matrixes that give instructions to try a ‘known good part’ in cases like this. It appears that the service engineers have the same problems we do from time to time, but who on earth always has a ‘known good part’ in their inventory? If another car could be found on the lot somewhere, we could plug the suspect PCM into the other car and see if the problem moved to the other car.  The big problem was that there wasn’t another Eagle Summit on the lot, so we were stuck.  Salvage yard modules are always questionable, but a lot cheaper than a new one if they can be located. Furthermore, those salvage yard modules don’t come with much of a warranty, and if the car were to quit again, the PCM would instantly be a point of uncertainty.  For this job, we opted to quote this customer a price on a new one.

Sticker Shock

            It wasn’t surprising to Danny and me that the PCM for this little car was over $900.  The sad part about the whole thing was the age-old problem customers have concerning how much cash to spend on a older car.  The little Summit would still have engine problems even after the electronics were straightened out, and who knows how much money and time had already been spent on the little car?  A customer can sometimes pour a couple of thousand bucks into a car that isn’t worth half that much even after the money is spent and the work is done.  This particular customer decided to take the low road.  While we didn’t suggest it and he didn’t request it, a wrecking yard PCM might have been in order, but he was satisfied to pay for our diagnosis and have the car hauled away.

Right Up To The End

            This was a clean little car, but it had some very expensive problems.  And in situations like this one, nobody wins.  Danny and I worked together on it for more than an hour, and we’ve both been chasing sparks for over twenty years.  Sometimes one has to wonder how much time to spend looking for a problem before condemning the ever-present but very expensive ‘black box.’  In this case, the customer got all the troubleshooting we could give him, but our prognosis either scared him off or was too rich for his wallet.  He had paid for a tough hour of shop time, but he wasn’t quite ready to let us make his problem our problem.  And it would have been our problem if we had plugged in the new PCM to find that the problem was somewhere else!  Unfortunately, we may never know…

This is the location of the MPI Control relay that powers up the PCM.