Of Corvette's and Kenworths

"I've tried everything on this Corvette and I can't get the AC compressor to turn on. It had a bad programmer, so I put another one in. Then it had a bad pressure switch that caused the ECM to set a code 66. I cleared the code but the compressor still won't turn on. The HVAC programmer was $780 and I think it is defective, but I don't want to get another and find out that something else is the culprit."
Max is a good technician. He graduated with a technical degree in both auto and diesel technology and is willing to learn. Some of the older technicians, just like instructors, don't like change. Some--not all--are unwilling to get out of their comfort zones and in this business, that's a no-no. An instructor owes it to the students to keep current. Students are reflections of their instructors.
The HVAC programmer is a rebuilt part. It was an R-12 system I had a suspicion what it was. I pressed the fan up & fan down buttons simultaneously until-00 appeared on the HVAC display. I was now in diagnostic mode. Once in diagnostic mode, you can use the fan up & down buttons to scroll through the PID list. To view the value at the PID location, you press the AUTO fan button. I went to PID position 00: System Faults. I found )), which indicated no codes. Then I went to PID position 09: A/C System Mode. It was 01, which meant MAX A/C. The command was there. I moved to PID 11: Program Number, which represents a heating or cooling command. The programmer was commanding 00, a command for FULL COLD. I moved to PID 34: TEMP DOOR POSITION REQUESTED. It was at -155, which means when you see a minus sign at the value, you add 100. So, -155 was 255, which was FULL COLD.
Everything looked good. I remembered the hidden programmer code:48. When I taught HVAC controls in another lifetime, I called Code 48 the "Claude Rains" Code. the young technician didn't know who Claude Rains was.
"He played the invisible man in the movie back in 1933," I said. "It was a classic movie."
"Oh," the young technician said. It was way before his time.
I plugged the Tech 1 scan tool into the data link connector. First, I went to PID 00 and used the clear codes command on the HVAC head. Then I used the Tech 1 to clear ECM codes. Then I turned the ignition key off, took the Tech 1 off the data link connector, started the Corvette, and the compressor energized.
"How did you do that?" The wide-eyed technician asked.
"With experience," I said. "The code 48 was intentionally masked from the programmer's data list, but if it set, it sent a signal over the UART bus to the PCM to disable the compressor. Code 48 was identified as a LONG TERM FREON LOSS code."
"But I disconnected the battery. Wouldn't that have cleared the hidden code?" The technician asked.
"Sometimes, but not always," I said.
That was an easy one, but only because I learned the hard way on that one, all those years ago. We learn good lessons from our mistakes. We have to--so history doesn't repeat itself.
Onward to the Kenworth in my next post.
 

Tools of the Trade

Like every good technician, I have an electrical tool kit consisting of wire and terminal repair tools and soldering equipment. I also keep a stock of other materials, such as:

Stabilant 22:
This material is electrically active and stays resident within a contact pair, enhancing conductivity without leakage between nearby contacts. It's initially non-conductive, amorphous that becomes conductive when exposed to an electric field.

Rodent-Deterrent Tape:
This tape is treated with capsaicin to protect those wires with soy-based coating.

Liquid Metal Defogger Grid Repair:
Great for a variety of repairs.

Bare Conductive Electric Paint:
Like any other water-based paint except that it is conductive. I also have a conductive paint pen.

Nyogel 760G:
A silica-thickened, synthetic hydrocarbon grease to protect electrical contacts.

Silver Conductive Glue Paste:
Cures at room temperature and air dries.

Nickel-Copper Cobalt Tape:
Great RF shielding--70 dB @ 1 MHz and 100 dB at 1 GHz.

The 2016 Impala that I worked on had intermittently activated the SERVICE STABILITRAC warning on the dash and the sound of "electrical arcing" under the hood. My homemade, electronic electroscope detected
a spark from the inside hub to the outside ring on the harmonic balancer.

 

Measuring the resistance from the center of the hub to the outside ring of the harmonic balancer, I found about 763 K Ohms. I cleaned the harmonic balancer off with brake cleaner. Next, I used my liquid metal defogger grid material, applying it from the center hub to the outer ring on the balancer across the rubber to control the resistance. I told Samuel it is a temporary repair. The intermittent P0300 and STABILITRAK message was due to terminal fretting at the EBCM connector, terminals #23 and #24. My 40x illuminated loupe comes in handy for close inspection. A loose ECM ground caused the misfire at times--the P0300.

 

The problems with the Kenworth were resolved. The Service Lamp was flashing and there were transmission issues on the UltraShift transmission controlled by the TECU. The culprit was a corroded power supply to the TECU. The shop never suspected this because the wiring had been replaced before. But the root cause? I spotted the dirty powder on and near the wiring with my UV light (just another use for it). That was caused by nitric acid formed in the presence of moisture in the air. So, where did the nitric acid come from? It was caused by the ionization of nitrogen in the air, and that was caused by an intense electric field.

This diesel had some add-in inverters, causing nearby corona discharge, due to the close proximity of the wiring. Corona eventually causes punctures and carbon tracking. Whenever I see wiring in close proximity of other wiring, especially from an upfitter, I use my UV light to look for the residue.

A technician has to have good detective skills and be persistent. Had I replaced the power wire to the TECU, it would have failed because I would have treated the symptom, not the disease.

 

 

Electronic Heartbeat?

“It’s a 2016 International and it was parked out back for the weekend when a storm blew in,” Ketchum Jessup said as we stood by the truck. “I was havin’ me a cup of coffee, watching the storm, when a bolt of lightning zapped my truck. Struck the mirror on the driver’s side and blew the glass out,” he said, gnawing on a plug of tobacco. “After that, the truck wouldn’t even crank over.”

I looked at the documentation of other repairs. Every module on the J1939 data link had been replaced: the ABS ECU, Auxiliary Gauge Switch Pack ECU, the Body Controller, the ECM, the Electronic Gauge Cluster and the Pyrometer Ammeter Module. On the second visit to the shop, the Body controller, the terminating resistors and the J1708 data link had been replaced.

“It ran for a little while and then the Engine Stop warning turned on,” Ketchum said. “So I had a mobile repair truck come out.”

I saw that the Body Controller was replaced again, Ignition 1 relay and the Auxiliary Power relay. The BC supplies ground to the auxiliary power relay when the theft deterrent is active. Then the key is ON, the auxiliary relay is energized and the relay opens. When the relay opens it allows power to be fed to the Ignition Relay 1, the Crank Inhibit relay and the ECM. If the auxiliary power relay is energized, the vehicle won’t run.

“Now the gauges tick-tock as if they were on a clock or heartbeat or something when the key shuts off. Sometimes it starts, sometimes it doesn’t,” he said.

I checked for recalls, campaigns, safety-related reports and the like, but nothing came up during my research. I checked connections at the battery as a standard procedure. I checked all the fuses.

“Those are all basic checks,” Ketchum said.

“Exactly,” I responded. I used my inspection camera mad carefully checked all ECU connections, including the door pods. I visually inspected all ECU grounds. I was going over all worked performed by the other shops. You have to. The one thing that you fail to check will be the thing that gets you.

When I decided to check the terminating resistors, that’s when my ohmmeter readings were all over the place, as if I had current flow after everything powered down. I disconnected the battery and checked the terminating resistors. Individually, they measured 120Ω, which was good. But when I checked the J1939 backbone, my ohmmeter indicated OL.

 I reconnected the battery and started the truck. It ran fine for a few minutes and then the Engine Stop light appeared. I turned the key off and the gauges would jump periodically. I disconnected the battery and measured the resistance of the J1939 backbone again. Again it returned OL. Was the data bus open?

I manually ranged my digital meter to the 4KΩ scale. I measured the bus again and the resistance slowly climbed until it returned OL on the display. I reversed the meter leads at the diagnostic link connector, and the reading on the ohmmeter red a negative resistance, counting down to 0.00Ω. It reversed polarity and counted up again, until reaching OL. Now I knew what was happening.

My meter’s 9V battery was charging a capacitor somewhere on the circuit. When I reversed the lead connections, the capacitor discharged and then charged up again. Once the voltage reached the same value on each side of the dielectric, the capacitor was fully charged. The meter displayed OL. I decided to unplug each ECU, one at a time. I was also cross-checking each number. The fourth ECU I disconnected caused the odd reading stop.

Modern heavy-duty trucks can be described as rolling computer networks with seemingly endless possibilities for data collection and transmission. Many different devices found on trucks have the ability of collecting and storing this data when an event occurs. These devices are collectively referred to as Heavy Vehicle Event Data Recorders (HVEDRs). In most cases, however, an HVEDR is not a device but a software application added to an existing device that has an original purpose other than collecting and storing incident data.

Data collected and stored by HVEDRs can be useful when analyzing a truck crash. The most useful data comes from trip/event data recorders on engine Electronic Control Modules (ECMs) and wireless fleet management systems (a.k.a. tracking/ communication systems, mobile resource management systems, and telematics systems). These systems can record vehicle speed, brake usage, vehicle-use histograms, vehicle position history, active and historic diagnostic trouble codes, and more. An endless number of devices used on trucks that have a HVEDR function and no two devices record the same data. Therefore, an HVEDR is like a Christmas present where you really don't know what you're going to get until you open it. You may think you know what is in the box and sometimes you are right but often it is a surprise.

This truck went to a bodybuilder at some point. The ECU was added that acted as a dedicated HVEDR. It must have contained a capacitor that charged and discharged. I opened the ECU. I found a choke that was charging a capacitor due to a bad conductive trace so I removed the cap, repaired the trace, installed a new cap, and checked it again. The capacitor was now functioning as designed.

Ketchum held the cap in his hand. That little thing called all this grief?”

“To you and me both,” I said.

I'm pointing to the cap.

 

Alien Makes and Models

“This one is serious, RJ,” the owner of a 2014 Kenworth said. “The EGR was non-functional, no DEF injection, reduced engine torque along with the MIL and STOP ENGINE light. Where do you start?”

The truck was a 2014 Kenworth, W900 Cummins ISX 15 CM 2350 truck. This one had a fault code 175 SPN 3464 FMI 3. The code indicated that the signal coming out of the throttle actuator is high. Considering that the throttle actuator was replaced three times, I strongly suspected it wasn’t the fault. I checked the terminals at the actuator and the ECM with my 40x magnifier, 25mm.—great for spotting terminal fretting.

 

We started by verifying the concern, followed by a thorough visual inspection. Research revealed no bulletins, although the Kenworth had a work history. The ECM was replaced and programmed. The throttle actuator was replaced three times, and a terminating resistor was replaced. I decided to decode the CAN bus packets, although I strongly suspected where the problem was—at the terminator that was replaced. I wanted to see how the data packets looked.

“What are we looking at? T.D. asked.”

“It is a Stuff Error, T.D. There is an issue with the terminating resistor that they replaced.”

This bit stuffing scheme is used to guarantee enough edges in the bit stream to maintain synchronization within a frame. If a listening device detects 6 bits of the same value, then it must have been the case that synchronization was lost and what was received was NOT what was sent! This is precisely what can happen with improper termination. Some bits in the frame are transmitted and received correctly, but the entire frame (and the integrity of all the bits in that frame) is not maintained and received at the listening CAN interface. 

Whoever worked on it before disconnected a 120Ω resistor, removed the other one and added two 120Ω resistors in series at one end. I didn’t know why. We repaired that issue.

Someone replaced the throttle actuator three times for a reason. My oscilloscope revealed AC noise to the neighborhood of 700 mV and that is a huge neighborhood.  The ECM was sending a regulated voltage signal out—dirty with AC noise of about 650 mV AC P-P. It was time to check the power feeds going into the ECM. Bingo!  The ECM power feed had about 500 mV AC ripple (the battery acts as a dampener, smoothing out AC current so a bad battery can cause strange symptoms in a vehicle). I had that condition on an Impala once. I checked the alternator with my scope: 1.2 volts AC!

Finally, the ECM ground had excessive voltage drop. After all the repairs, the Kenworth ran great. T.D. was happy to get it back on the road.

“Sorry you didn’t see any green aliens when you were here,” T.D. laughed.

“Did you ever see a UFO, T.D.?”

“No, but others around here have. There has been reports of saucers, cigar-shaped UFO, triangle-shaped, diamond-shaped and round ones,” He said. “We have a fair share of them.”

“You know what has always bothered me about the different shapes? I’ve always wondered if aliens have a choice of buying different models,” I said.

“What do you mean, RJ?”

“Well, we can buy an SUV, or a compact, or a sports car—and we can buy a Ford, Chrysler or GM; or imports. Judging by all the different UFOs, I just assumed that the aliens have choices of what to fly to our planet, T.D.”

He just shook his head.

Of Terminators and Characteristic Impedance

“Terrible what’s going on in this country,” Dan said, watching as I worked, tracing the data bus wires. “We’re proud of our heritage and now folks are taking it away from us,” he said as I worked. His truck had a DTC 148309, kept going into regeneration and he had it in to three different shops that replaced the complete aftertreatment system, including the doser valve three times.

“What do you think about all this stuff going on, Bob?”

“Actually, Dan, your truck is currently consuming whatever brain power I have at the moment,” I said.

He nodded. “I asked the technician working on it the last time to explain the J1939 data bus, but he wasn’t very clear.”

“The J1939 multiplexed chassis uses twisted-pair wiring, twisted to prevent the data backbone from acting as a radio antenna. The wires twist a through a full cycle once per centimeter on the backbone—2-1/2 times per inch,” I said.

My hand was getting cut up from the tight places I was going into.

“Those terminating resistors—they’re for signal noise, right?” Dan asked. “And the technician said they need to be 120 ohms, right?”

I checked V-CAN, E-CAN, D-CAN and A-CAN terminating resistance. The resistors each came in at 120-ohms.

“The truck J1939 data bus has a characteristic impedance of 120-ohms, but don’t confuse this with terminating resistors used at the end of the data backbone, Dan.”

“What is characteristic impedance, Bob?”

“It is just a factor of how a conductor appears to a high-frequency signal, regardless of length. You can’t measure it with a meter.”

Andrea, Dan’s wife, brought us some homemade lemonade. I loved her deep-southern accent. “Y’all thirsty?”

I thanked her and she noticed my bloody hand. “Y’all get into a cat fight?”

“No, just Dan’s truck,” I said.

Andrea offered to take care of the scrapes but I refused, saying that I was planning on getting a few more. She shook her head and went back towards the house.

“The terminating resistors also provide a low-resistive pathway for current flow between CAN H and CAN L. The capacitance can discharge and cancel rapidly,” I explained. “Of course, the length of time for a capacitor to drain is proportional to the resistance of the conductor it is discharging through. If capacitance can’t rapidly discharge when a component is trying to transmit at a low level, then the voltage differential between CAN H and CAN L stays high and that can shut down the data bus.”

“The truck has been stalling at times,” Dan said. “I had to disconnect the battery a couple of times to get it started.”

I scraped my hand up again, getting at the hidden wires while Dan held the light. “How do they come up with this impedance?” He asked.

“They do it by maintaining the dielectric constant, or wire insulation, and keeping the actual physical space between the two conductive wires constant by having a consistent twist pattern. Twist pattern is very important,” I said.

Then I found it. “Here is the problem.” I said. Someone repaired the CAN bus.”

“I could have told you that,” Dan said, “The dealer technician did it. He found some rubbed-through wire.”

“Yeah, but substituting ordinary wire for J1939 cables is a no-no because the dielectric properties of the insulation should be constant. You can corrupt communications on the bus by altering the impedance.”

“Corrupt? How, Bob?”

“Because the amount of time needed for every signal pulse pumped down the bus to neutralize increases; now the reflected signals scramble the data bus,” I said. I repaired the wiring, fired up the rig, and checked the data bus and all was well.

“I think we have it licked, Dan,” I said. Everything looked good.

We had great conversation later on over lemonade on the porch while Andrea insisted of doctoring up my hand. We talked about the protests and expressed feelings, but the world is tainted enough without letting it spill into my blog or Facebook.

Dan and Andrea are great people. I wish the world had more like them.  

Bearings and thrust washers destroyed by electricity?

A great friend of mine, Phil, reminded me of those golden years of the HydraMatic 350 transmission, used from about 1969 through 1984. It was typically paired with small block V6 and V8 engines. Although it was destined to become one of the greatest automatic transmissions ever built, we were having issues with the 350’s burning up back in the day.

I recall the first one I ever disassembled. Almost every thrust washer and Torrington bearing were wiped out. The old-timers out there, like my friend Phil, know exactly where I’m going with this. On the 350’s, it was due to a bad ground. Those pesky electrons were traveling through the shift cable, which was also failing, and transmission parts through the driveshaft and the rear axle.

The repair was installing two braided ground cables--one from the cowl (firewall) to the wing nut at the carburetor and the other was from the bottom of the cowl to a bolt on the bell housing securing the transmission to the engine.

As I explained to John and Greg here in Shepherdstown, some thrust bearing premature failures have been attributed by poor engine grounding. When the starter motor is engaged, electrons need to go to ground and if the engine isn’t properly grounded, current can run through the crankshaft and directly into the thrust bearing’s steel backing. In time, the thrust bearing faces can rapidly eat away, behaving as if the thrust surface on the crankshaft isn’t finished properly.

It’s simple and necessary to check for too much voltage in the drivetrain. You simply connect the negative lead of your DMM to the negative post of the battery, and the positive lead to the transmission. You should see no more than 100mv on your meter while the starter is cranking for about 5 seconds. Although the most current flow usually occurs while the starter is cranking, current in the drivetrain can happen while accessories are operating. That’s why you should perform this voltage drop test with the ignition on and as many accessories operating as possible. Again, the maximum is 100mv. If the voltage is too high, check or replace the negative battery cable, or add ground straps from the engine to the frame; or from the transmission to the frame. Some systems may reach 300mv briefly without causing a problem. For added assurance, improve the ground with a larger battery cable or additional ground straps.

“Is it possible that there is too much latent current flowing through the drivetrain, but without measurable voltage?” Greg asked.

“Sure,” I said, remembering one I dealt with about a year ago. “If you have a ground issue in the chassis, but the transmission and engine grounds are fine, the vehicle could pass the test, because the driveshaft and suspension components become the ground path. Saturn Vues going through wheel bearings can be the result of this.”

“So, how would you check that?” John asked.

“I pull the driveshaft and then measure the voltage drop. Now the frame and drive train must pass the 100mv test. Then you run a ground strap from the engine or transmission to the frame.”

“Remember the rear brake caliper slide pins freezing on W body cars years ago?” My friend Phil reminded me. “The engineers finally determined it was due to the static electricity created be the friction of the pads against the rotors looking for a ground through those mounting pins.”

John and Greg are great technicians who take pride in their work and I was glad to have an opportunity to help them out.

Haunted Car?

Can a car be haunted? On a muggy night in South Africa in 2004, a Renault Megane turned itself on and began jumping backwards, all on its own. This car is notable because there were several witnesses and news reports. Nine people, including two police officers, heard the Megane's engine start before it "jumped" backwards twice uphill, according to officials. The car apparently roared to life on its own, despite having no keys in the ignition and the parking brake engaged. I thought the 2006 Lincoln LS with a 3.9 was haunted. It came from Shepherdstown, a nearby town so haunted that the local police called in “Ghost Hunters” from the TV series. The LS Sport came from a couple of different shops for intermittent misfires, stalling, MIL on and transmission shifting issues. The work history indicated the plugs were replaced, TAC module was replaced, The PCM was replaced multiple times, the fuel pump replaced and coil assemblies were replaced. The throttle body was cleaned. Road test #1 resulted in an engine stall—just like the ignition was shut off. After the vehicle restarted, a P0607 flagged. The PCM’s internal CPU has encountered an error. This could be a calibration update, or aftermarket performance parts installed, or a damaged PCM. Clem came back with another PCM (number 3) and wants me to install it. I said no way. I did a visual inspection first, especially focusing on the PCM considering it was replaced before. Guess what? I found a damaged CAN (+) wire had several broken strands. I did a wire repair, confident that this was a major issue. I continued testing, though. I checked the ground circuits first. I prefer testing the ground circuits first because the remaining circuits will use this ground to verify they are working correctly.  This includes battery cable connections, which are subject to corrosion. No trouble found. Once the grounds are checked, I then move to the voltage circuits starting with B+ voltage because the Ignition, voltage reference, and control circuits require a functioning B+ circuit. B+ is direct battery voltage to a module or component. Those circuits were good. I checked the ignition circuits which includes ignition circuits that pass through a fuse, module, component, or connection.  When testing ignition circuits, I referenced the wiring schematic as to the ignition switch mode or position needed to achieve a valid reading. The ignition circuit test is referenced to the ground or low reference circuit and requires a functioning B+ circuit for a viable test. I checked the reference voltage circuit next, which is referenced to the ground or low reference circuit and requires a functioning ignition circuit. Satisfied that all PCM circuits were now acceptable, I moved on. Road test #2 flagged a P1270 (Engine RPM or Speed limit reached) and P2112 (TAC Control System). I checked data PIDs IACTRIM and IACKAM2. The sum of IACTRIM and IACKAM2 is the total IAC. Total IAC was 0.54 lb/min. I removed the throttle body and cleaned the edges of the throttle plate and throttle bore. It was bad. I reassembled it and took the vehicle on road test #3. This one resulted in an engine stall—just like the ignition was shut off. After the vehicle restarted, a P0607 flagged. The PCM’s internal CPU has encountered an error. This could be a calibration update, or aftermarket performance parts installed, or a damaged PCM. Clem wanted me to install that PCM. I said not yet. I already checked for the common bad coil issue. I always check the Coil-On-Plug ignition systems for defective coils by back-probing the main coil B+ with my PicoScope on glitch capture and using an injector as cylinder reference. I put the coils through the tests, including idle testing, cold, hot, snap throttle and power braking. With throttle cables gone, plus the 6-way BPPS, it is tougher to check these functions under the hood. I always look for a positive or negative voltage spikes on the coil ignition feeds which will indicate the failing coil(s). This car passed this test with absolutely NTF!

“This car is haunted,” Clem said.

“I have to walk away from it for a while,” I said.

“You can’t fix it?”

“No—I haven’t fixed it. I don’t throw in the towel but I do need a break.”

I thought about it while watching the Shenandoah River go by. A shop said they cleaned the throttle body and I could tell they didn’t. Did they even change the plugs? That was on my list of things to do when I would tackle it again.

“You must have one heck of a load, Bob, with all those cars and trucks to diagnose,” Clem said.

Then it hit me. I smacked my head as a token gesture.

“One heck of a load—you are so right, Clem!”

“Huh?” Clem said.

I got out my extra long test leads and went for a test drive. And I caught it--an ignition voltage spike! Right after the spike I lost the injectors. I also wanted to check that TAC module issue. With my PicoScope hooked up to ignition primary, and both throttle position signals (one on glitch capture), I captured a voltage spike on the TPS signal. The voltage spike is what caused the TAC codes.

The coil failed under a heavy load, hitting the PCM with the spike, causing it to reset.

More history of the LS Sport surfaced. The fuel pump was replaced because the original owner jump-started the car with a heavy-duty diesel truck charger on boost. Why did he do that? Because he was tired of the old battery going dead, procrastinating on buying a new battery, and after the original battery would take a charge, he decided to boost it. After that, he noticed the PCM would not drive the fuel pump so he had the car towed to a shop and had the technician replace the fuel pump.

The COP units were replaced, but a new one was bad to begin with, failing only under very high load. The ghosts were finally “exorcised” from the Lincoln.

I was so happy to get this one diagnosed.

“How can I thank you?” Clem was elated and hugged me and kissed me on the cheek. Now, don’t think that was weird. Clem is short for Clementine.

“You can really thank me by paying my bill, Clem,” I said, handing it to her.

“Now—that’s scarier than any ghost,” she said, after seeing it.

“I can put the ghosts back in the Lincoln, then.”

Clem paid.