Crank Crank Crank Crank… Even When It's Complicated, Don't Overlook The Basics
I'm going to be completely honest here. This case study is on a job that kicked my butt multiple times. Enjoy and learn from it, I know I did.
Machine is: 2011 John Deere 333D Compact Track Loader - 1T0333DMVBD207327
Customer hauled this machine in with a complaint of a crank/ no start. They also supplied a CMP sensor, as a code popped up on their monitor for CMP signal out of range. Story was that the machine had been sitting for approximately 3 months, but had run before being parked. Their mechanic went to fire it up to get ready for a coming snow storm. The machine apparently fired up, threw a bunch of codes, and then shut off and wouldn't restart.
I attempted to start the machine and just got a slow crank due to the heavily discharged battery. After dragging it off the truck and into a bay, it was time to get to work. First order of business was to get a battery charger on it, pull codes and begin to set-up a plan of attack. Why not just replace the battery? With a no run condition there is no easy way to get the boom arm up in the air, which means there is no easy way to wrestle a 60lb battery out of it's hiding hole in the back, and not arc it against the machine somewhere in the process.
While waiting for the battery to charge I started my diagnostic plan (excuse the handwriting). This is an example of my mental process put down on paper.
It was a bit of a gamble as to which direction to go next, but a lot of the codes were communication and/or strange voltage codes. I narrowed my focus to codes that would prevent the engine from running and ignored the other codes for now. I focused on the engine de-rate and shut-off. These will occur if coolant temp and/or oil pressure readings are too high, low, or implausible.
At this point the battery was charged at least enough to give a good attempt at cranking. I cleared out all the codes and attempted cranking. The machine cranked, albeit a bit labored, but hey, the machine ran before right? Plus the battery was weak, so I put it to the back of my mind.
One code re-appeared after cranking and that was the … (CMP Signal Invalid). OK, so once again, boom arms are in the way, no easy way to get to the CMP sensor; got to love re-occurring themes. Instead I raised the cab and tapped in at the ECU.
Time to fire up the scope and get some waveforms. The waveform looked normal, despite the fact that SI (service information) calls for a 45 tooth with a 3 tooth gap for the crank , and a 12 tooth with a single offset notch cam gear for timing #1 TDC. I run into this quite a bit, as an OEM engine can be shipped out with different sensors and flywheel configurations depending on it's end-use application. There isn't exactly a database of known good waveforms floating around (yet).
So now I have verified that the signal is good, and this application will run with only a CKP or CMP sensor (it just cranks a little longer). The computer doesn't see it that way though as evidenced in the data here. Something is getting scrambled somewhere, but it's internal to the ECM. Maybe corrupted data due to excessively low voltage?
I wanted to do a relative compression test quickly before going into further diagnosis as to why the ECM was glitching out on the CMP signal. After all, it had cranked over with a slightly strange cadence. Boom arms are in the way of the engine, plus it's not a cheap or quick job to pull an injector or glow plug to install a pressure transducer as the valve cover encases everything. As in the previous capture, yellow trace is crank and green is cam. This is a 5030 engine (3.0 Liter Inline 5 cylinder). Notice there are some low humps and timing seemed to be skewed. Machine does have 1,861 hard hours of use on it.
Back to the ECU. After letting the battery charge up more, I decided to hook up a brand new battery up with heavy duty jumper cables in order to eliminate the funky effects a shot battery can have on a system. This ended up eliminating the … CMP code, but now I was getting a recurring code for engine oil pressure fault, and it still wouldn't start. I now also had a … ECU Boot Block Error (this means either ECU was reprogrammed or there is an internal error).
I'm leaning more and more towards the ECU, so I check and load test my powers and grounds. Everything checks out just fine.
Next step was to see if I had injector control, this I could easily access because I was right at the ECU. No injector control or power (this system fires 90V out to the injectors and the computer controls the ground, so not exactly something I can easily actuate my self).
I ran a Harness Diagnostic Mode Test (actuates various components in a self-test). No injector buzz. Fried injector driver circuit. ECU it is. I made the call and explained that we would need to put an ECU in (around $2k) and program it. If additional diagnostics were needed from there we would continue on.
A few days go by and the customer approves an ECU replacement. I download old calibration data (injector codes, etc) from the original ECU, swap in the new ECU, and program. Cycle the key and hit the key for the moment of truth. Crank Crank Crank Crank Crank Crank......my heart dropped. What did I miss?
I start double checking my connector seating, battery voltage, fuses, and started to doubt my entire original diagnosis. After an hour (yes, an hour, this is the back to the basics part), I finally decided to check my injector driver output and control. It's present. So ECU is doing what the old one won't. Now on to the rest of the diagnosis.
I checked fuel coming into the cylinder head (this machine uses cam actuated Electronic Unit Pumps), so there is no easy way to check fuel at the injectors or actual injector integrity.
I found debris in the fuel pick-up tube and an inline fuel filter installed backwards. Clean the pick-up tube, remove the filter and run an auxiliary tank of known good, clean fuel. Fuel pressure (with inline gauge) is at 24 PSI (well within spec). Crank Crank Crank Crank Crank Crank.... still no good.
I'm starting to get a bit frustrated at this point. I stuck my hand over the exhaust stack and could count 5 distinct pulses as it cranked over, I pulled the air cleaner out and could feel consistent pulsing there as well, no long gap in between. I figured the valves have to be working at least somewhat.
Then it hit me, the scenario of the original complaint. Machine sits, incredibly cold weather (negative wind chills at times) and a weak battery. Overanxious mechanic or operator and some ether? Could explain the strange cranking cadence, although the relative compression test wasn't terrible. It was time to pull the valve cover.
So did the exhaust valves get stuck in the valve guides and just bend the push rods before freeing up? Perhaps they were stuck, and some ether made a bad situation worse? Maybe there was no ether at all and it was just a freak thing as it supposedly ran? Side note: the rocker arms were all numbered with paint on top, so someone had been in there at some point, though there is no history of it according to the customer. The root cause of the bent push rods will be determined once I start my post mortem on the top end of the engine.
Edit: After tearing down and checking for stuck lifters or injection pumps, I wanted to check my cam timing which appeared to be off as evidenced in the scope capture up top. Cam gear on these engines is a press on with no key. A stuck plunger on one of the injection pumps is enough to spin jam the cam and spin it in the gear. This will cause your CMP & CKP correlation to remain in time, while the timing has actually changed. This was the case and the camshaft had spun in the gear. This means engine has to come out in order to retime.
Moral of this story is that when you are chasing yourself in circles and nothing seems to make sense, stop and take a deep breath to organize your thoughts. Go back to your basics, verify the story and history of the machine, think things over clearly, and consider what you may have possibly overlooked as not concerning at the time. I got so caught up in doubting my initial diagnosis that I was overlooking the obvious.
I should have quickly verified that my original problem with the ECU was resolved with the new one and continued on with additional diagnostics. I'll be completely honest and admit that I have a full 6 hours into this diagnostic and getting to the point of bent push rods. Not my finest job, but I always learn more from the jobs that kick my butt than I do from easy ones.
Now I just have to figure out what tests or procedures I could have done to save myself a lot of tear down in this case. Time for more research.
Thank you to anyone who reads this, and I hope it helps you in the future.
Some additional pictures of where everything lives.
A couple of things seem odd to me. You mentioned that the unit has 1891 hours on it. The report shows a missing 30 hours. The report also shows shutdowns 6 minutes apart. I kind of wonder why. :)
When I start to question myself, I treat it as the first time it came through the door. If my test results are the same, I have to believe that I am correct in whatever I have determined to that point. Could you have done it faster? Possibly. Did someone, perhaps, not give you the whole story? I'm kind of leaning that way. Therefore, you had "help". When someone has helped and doesn't give me the whole story, I don't beat myself up because the crime scene has been contaminated.
Guido [can I call you that now?),
I had forgotten about the hours. 1861 would be the actual correct hours. The ECU hours remain the same in, but if someone reprograms the EMU (Engagement Monitor Unit), which is where the hour meter is displayed (think dashboard), they have to reset the hour meter display. Unfortunately most just go for close enough or "I think it was". This is also a pain when you need to generate an unlock code if the owner has forgotten theirs as the master codes change randomnly hour by hour, sometimes down to the .1 hour.
The 6 minutes apart had me intriqued as well, but not enough to choose a diagnostic route off of. I'm actually reading over my notes and data now. Trying to assemble a mental picture of what actually happened based off known data, previous experience, and deduction. It's a fun excercise and also helps me piece together who contaminated the crime scene, how, and for what reason. See below:
- Gregory : "Is there any other point to which you would wish to draw my attention?"
- Holmes: "To the curious incident of the dog in the night-time."
- Gregory: "The dog did nothing in the night-time."
- Holmes: "That was the curious incident."
I don't beat myself up (well maybe just a wee bit), but no matter what I do I can always find room for improvement. I like that lean manufacturing principle of Kaizen, improve every day, even if it is only 1%.
I've been called Guido for well over 35 years now. When someone calls me "Anthony", I start looking for bulges and/or badges.
I wasn't suggesting that any 1 item should dictate a course of action. Based upon what you wrote and what I can see, I would wonder if someone attempted to jump start it backwards and alos used ether. (I've seen valve covers blown off of the engine.) After all, it stalled 2 times. The driver/s was/were open. If more than 1 driver, that's an unusual occurrence without help. Granted there was a breathing issue but how can you determine that when you're showing 48 RPM on Engine Speed, 82 RPM on the CKP sensor and almost 3K RPM on the CMP? Sure, you should be able to hear if the RPM is really that low but you have a bad battery also. Could the battery cause the false readings? Absolutely though, as you know, that doesn't mean it is the only thing wrong. All of those pushrods being bent? One (multiple) hell of a backfire.
The way that I see it, it would be a good …e study if you out and out screwed up. Since it appears that you had help, not so much. (No offense meant.)
Long story short, sometimes, there is no good take-away (teachable moment for you though there may be [$$$$] for the customer). On those, I tend to kick the situation to the curb. 101 Proof Wild Turkey has been known to help.
Actually, I just learned something. I didn't know they use a rolling code. Thanks.
Nice job. Get yourself some pressure transducers, that would have flushed out your mechanical fault. I usually put one in the intake, exhaust, then I do starter current and then either Injector synch or crank and cam.
I surprisingly do have a couple pressure transducers. I enjoy doing in cylinder so this will be a great extension on that. I've got a 0-100psi and a 0-500 psi. Or are you saying to go with something more like a FirstLook? I hadn't even thought of checking it that way on a diesel. I've seen the intake waveforms on the gassers, but I've been rather negligent in doing any experiments that way.
I'll be giving my 100 psi transducer a shot today and see if I can come up with anything. I have a known bad for sure.
I appreciate the advice.
Chris....Excellent write up. I have to say i was totally impressed that you actually wrote down a diagnostic game plan on paper! I try to keep it simple when I get a no start. Remember you have to have fuel, spark(in your case compression) and good electrical supply. You also need to make sure that its all together at the right time as well. A compression test would have revealed your problem at the get go
Like Michael stated get a psi transducer and set up a scope. I want to thank you for sharing with us. This is what this site is all about.
Thanks Edwin, I appreciate that and I appreciate any and all input. Positive or negative comments are welcome (though I prefer to read negative ones on mondays). I can assure everyone that unless I'm cursing, I'm not offended so fire away; I'm my harshest critic. I have more to learn than I can imagine.
I did play around with the 100 psi transducer I have after Michael's comments, but couldn't get anything to work. I believe I'm going to get a couple firstlook sensors.
Compression isn't an easy test to do on this (as I'm sure you've experienced on many pieces) and I didn't have the correct glow plug adaptor, which I was a bit surprised at.
Looking back, there are multiple ways I would have approached this diag to make it easier, but that's why I do these. So I can learn and others can as well.
Thanks again Edwin, and I look forward to more write ups from you.
Additional Information - During tear down and confirmation that the camshaft had slipped in the gear, I found what I believe to be the root cause of the injector circuits in the ECU getting fried.
There was a rub through/short of the 90V supply to three of the control wires and ground in the injector harness. Trace 1 (yellow) was supposed to be the common 90V supply to the injectors (JD outputs a 5V Bias on this wire FYI). Trace 2 (green) was #3 injector if I recall correctly.
It was an accidental find to be honest.
Chris, that find is like winning the lottery. That's an awesome video clip.
I think that is a safe assumption. Take the good with the bad.
How many amps to those injectors? (rh) If you get bit by that, it could make you sit down and think about the role that pain and confusion plays in your life. (Mitch Schneider, IIRC)
They don't publish a spec but based off the math, each injector coil has ~4 ohms resistance and 90v applied, so 22.5 amps give or take.
Would have made life interesting for sure.
Hi Chris, from what I am seeing on your injector waveforms, this is a solenoid type Injector Correct?