No Crank & Stuck On A Hill - Applying Logic To Trouble Codes
DTC's can lead us in the wrong direction if we don't stop and take the time to evaluate what they are actually telling us.
The other week I received a call from a tech at another location who was looking for some guidance on a machine that was stuck on the side of a hill and would not crank. He explained that he had 2 DTCs but wasn't sure how to attack them; he also had a few other machines that were giving him issues. I decided to take a ride up the next day and have him ride along with me as a diagnostic teaching exercise.
The machine in question for this study is a John Deere 35G Mini Excavator (made by Hitachi, equivalent to the Hitachi Maxxis series for those familiar) with a complaint of a sudden no crank/no start. Upon arriving at the machine, first thing checked was battery voltage (12.4V) as a baseline. The next step was to confirm proper operation of any safety interlocks that needed to be satisfied before the controllers would allow a start command. This machine requires only one interlock to be satisfied before a start command will be sent, this is the pilot control lever. The lever is a simple on/off switch which is shown as off in the raised position so as to prevent accidental engagement of any electro-hydraulic functions during cranking or when the operator is out of the machine.
Proper operation of the pilot control lever verification was performed by entering service mode within the dashboard (achieved by holding menu button down and holding while cycling key to the on position) and monitoring switch input. Pilot control lever interlock was fully functional so the next step was to duplicate the complaint and observe what occurs.
When key is turned to the crank position an engine failure light/buzzer is illuminated within 3 seconds of cranking with an E1100 (Engine Failure Fault). This code just lets us know that the MCZ (main controller, can be thought of as the BCM) has received a failure code from the ECM (Yanmar engine in this instance). The MCZ also sets a 1208 code, this is its interpretation of the Yanmar ECU language and denotes it as an engine speed sensor code.
Scrolling through the dashboard to the codes menu then allows us to access the specific failure code of 1078.04. This code translates roughly to a P0340 (CMP Sensor Fault), indicating failure of engine speed sensor. The interesting thing here is that there was no dimming of the display, no audible starter solenoid noise, engagement of the starter, and no voltage drop at the battery during the 3 seconds of key in start position that was required to set the code. Next step was to research code set criteria and system operation, which is often not available or severely lacking in the off-road industry.
The engine speed sensor code refers to 1 of 2 speed sensors located on this engine. The first is a CMP (cam position) sensor located in the injection pump housing, this sensor is non-serviceable per the parts catalog. The secondary sensor, which is used as a back-up and for engine RPM plausibility checks, is located within the alternator on your typical tachometer output. The code set criteria is show below.
Next step was to pull a schematic to see what type of speed sensor we were dealing with, be it a VR (variable reluctance) or a Hall Effect. The sensor is a VR sensor; on a side note: it has no bias voltage on it. At this point we had spent 5 minutes or so looking at the schematic and code set criteria and I explained to the tech that I wasn't concerned with it as all signs pointed to the code being a symptom of the no crank and not the cause. It was at this point that I was informed that the factory had mentioned they have had multiple ECU failures that resulted in the same no crank with a 1208.04 failure code and so the tech had been recommended to replace the ECU. Obviously this hadn't fixed anything and so I had him re-install the original ECU and we continued on from there.
I explained that I was no longer concerned with the code as it was a symptom of the actual fault (no cranking). I reasoned that the engine must be expecting to see a speed signal during cranking and it was not. How did it know when to expect a speed signal? It must at least be recognizing the key switch being moved to the start/crank position. That got us further, but I had no great interest in pulling apart interior panels to confirm switch output (after all that's what data PIDs are for). Unfortunately no data PIDs for the ignition switch status were available through ServiceAdvisor (Hitachi uses MPDr for their OEM diagnostics and so not all parameters can be monitored with Deere software).
I continued with my train of thought from this point and determined that since there was a possibility to set a fault code for starter relay output status, that the controller had the ability to monitor for some type of voltage drop on the starter relay control circuit. This led me to believe that we could go straight to the starter relay and check the output (thought process was that a starter relay fault should set before a speed sensor fault should). I gained access at the heavy gauge white output wire which runs under the cab direct to the starter solenoid.
I had the tech attempt cranking and found that there was a 12V output from the relay on the wire. So we had the fault narrowed down to one small section of wire, as obviously all the proper inputs and outputs were met in order to command the starter relay on. This meant we had to raise the cab as far as we could in order to gain access to the starter. The fault was immediately evident at this point.
Some type of rodent had chewed through the delicious soy based wire insulation and broken the starter solenoid signal wire. There was also damage to the wiring harness in multiple areas around the solenoid wire. These are tucked down next to the starter and so can't be caused by rubbing or pinching on the bottom side of the cab. The ignition switch was cycled in order to move the fault code to stored status and a jumper wire was installed between the ends of the broken wire. Machine was cranked and fired right off. We ran a temporary patch and got it off the side of the hill and into a location where it could be picked up and repaired properly back at the shop.
When troubleshooting we have to remember that no matter how smart controllers may appear, they are rather stupid (for now, neural networks in newer controllers are changing some of that) compared to the human brain. They can only calculate and interpret within their programming parameters. Proper system design and programming, at least in my mind, would dictate that there be a way to monitor voltage drop on the solenoid signal wire and flag a fault accordingly.
The important lesson is to not miss the forest for the trees, in this case what are the faults and what are the symptoms of the faults. We had a very simple no crank fault, easily diagnosed with all the standard methods familiar to anyone within the industry, be it automotive or HD/off-road. This simple fault was complicated with a trouble code that was a symptom not a cause. The speed sensor fault code was a symptom of the non-actuating starter. The ECU & MCZ controllers both saw all the inputs and outputs they expected and so assumed the starter was in fact engaging and cranking the engine. This led the controllers to believe that there must be a speed sensor fault as no signal was detected and so it shut everything down in order to prevent engine damage. Reasoning through the data, codes, theory of operation, and possible interactions gets us our baseline. Mix in observation of actual operating conditions at time of fault with a good dose of logic and you have a winning recipe for diagnosing faults quickly and efficiently.
This entire diagnosis, from getting to the property to loading up to leave, education included, took 32 minutes (less time than it took me to write this up). We can all learn to diagnose quickly, efficiently, and accurately if we put ourselves to it, continue training, and using the most important tool in your toolbox, your brain.
Thanks for the case study. I am slowly messing around with more off highway equipment, my biggest challenge is locating service information.
It can be challenging to say the least. I've found that I can always get a service manual for most brands, but the on demand internet access to the whole library is often hard to come by. Building up a relationship with your local dealers is one of the best ways.
Helping to push R2R into the off-road market would be the ideal solution though. Their R2R is not the same as automotive by any means, and there seem to be less voices on the HD/Off-Road side pushing for change.
The upside is just about every testing procedure that is applicable to automotive is also applicable to HD/ off-road. I may be doing some hydraulic theory and standard testing method write ups in the future here. Focus on Pascals law and its implications. Stay tuned.
Nicely done Chris!
A valuable lesson. That can trip many up.
I've never seen a cam circuit issue cause a no crank. But with the way technology tends to go, it's only a matter of time(these days a blown out dome bulb will cause a crank no start:D).
Excellent strategy based diagnosis and use of your noggin 👍
Blown dash bulbs (Charge Lamp) have causes no charging situations (Mack Midliner). Defective alternators, though still charging, have caused no blower fan operation (Isuzu NPR). Your supposition is not far-fetched, at all.
I appreciate it Mario. Tech wasn't sure if a cam code could cause a no crank and I wasn't sure until I looked at the schematic and theory of op.
Perhaps if it had been a 3 wire and pulling Vref to ground I would believe it, I'm sure that's coming soon to equipment.
This was an excellent illustration of taking what you know and applying to what you don't know. The only issue that I saw is the use of the word "we" so often. As you get older, you'll find the proper word will become "he/him" or "she/her". As in "I had her ...." 😉
BTW, many people still get tripped up chasing VTD codes on vehicles with defective fuel pumps. Crank the hell out of something that long and "logic" dictates that someone must be trying to steal it.
Much appreciated Guido. We can all get tripped up or way overthink things at times; miss the forest for the trees so to speak.
Get a roll of this stuff and keep it in the shop as a 'supply':