M1078A1 - Crank no start

[fuzzytoaster]

I picked up an M1078A1 that's been uncooperative to start. The POC at pickup said they couldn't get the truck to start but I noticed the battery disconnect switch was in the off position and was hoping that was the issue..but it's deeper than that.

Attempting to slave start the LMTV off my wrecker gives it plenty of power to crank but it's not firing off. No warning lights or dash codes, it all acts normal. I pulled the fuse cover off and inspected the fuses which look good. All LED lights show green when power is on and ready to crank. This is the newer "green board" and not the blocky older style board.

I've bled both fuel filters and pressurized the tank to ensure it's getting fuel. There is a fuel pressure governor/restrictor on the rear of the engine for the return fuel. I've read those can be a problem by not allowing fuel pressure to build up when starting the engine. I took the return line off the fuel tank and pressurized it again getting nothing from the line. When cranking a steady stream comes out so it tells me pressure is being built and retained.

I wanted to eliminate the possibility of the dead batteries causing a problem when cranking so I pulled all 4 batteries and managed to save 2 to be put back in. The 12v and 24v systems were hooked back up correctly (verified by fellow members) and she cranks like a champ.

A quick squirt of either/starting fluid for an internal completion test was done. She does run..or wants to.

I bought a truck code reader to plug in the j1939 socket and see if the computer would talk to be but it ends up with a communication error when scanning. I can talk to individual parts of the truck for any current or stored codes but nothing is logged. (I thought this stuff was supposed to make this easier?)

I pulled the P1 connector (right side) on the ECU/ECM (engine computer module) to check for power to the unit. Pins 52 and 53 are always hot reading 24v. Pin 70 reads hot when power is on as well as cranking, reads at 24v.

It's clear to me the injectors aren't getting the signal to fire but I wanted to verify it anyway. I pulled the valve cover and used a non-contact voltage meter to sense the current to fire. None was seen.

At this point I'm suspecting a faulty ECU/ECM unit. Thoughts? Anyone have a good one on the shelf I can use to test my theory?

 

[snowtrac nome]

I suspect no power to the ecm, cat ecm's are pretty robust, aside from too much voltage or reverse polarity it should be good. The no communication error most likely is because the ecm isn't powering up. We did have one in Iraq with that same problem and s&s replaced an engine under warranty the never divulged what was wrong though.

 

[Suprman]

Back to that troublesome green power board. There are 2 big relays under it also.

 

[coachgeo]

Is there a shut off/trip on the Reverse Polarity Box that might need resetting?

Is there any conditions to where the ECU demands a no start such as:
. low oil pressure?? ( aka bad oil pressure sensor, wiring or engine oil pump problem )
. neutral safety switch issue? (Switch, wiring or transmission ECU )

 

[Suprman]

He has power thru polarity box. Neutral safety would prevent crank from dash.

 

[fuzzytoaster]

I suspect no power to the ecm, cat ecm's are pretty robust, aside from too much voltage or reverse polarity it should be good. The no communication error most likely is because the ecm isn't powering up. We did have one in Iraq with that same problem and s&s replaced an engine under warranty the never divulged what was wrong though.

Watching some cat videos and looking up schematics shows wires 52, 53, and 70 are power which are all getting 24v up to the ECM. Is there something else I should be checking pin wise?

Back to that troublesome green power board. There are 2 big relays under it also.

I'll see if I can scope it out tomorrow before I hit the road. How do I access those relays? I did already pull the ECM and will have it tested (maybe flashed) this week so I can put it back in next weekend. It's either bad and the issue or good and we can eliminate it from the equation.

 

[Coffey1]

There was a guy on here a little while ago, his truck would not start after he changed oil.
He unplugged oil pressure sensor on motor and plugged it back and truck started.

 

[Suprman]

Bad power board usually causes intermittent electrical issues. I usually do a visual diagnosis, you can usually see the bad solder joints or move relay sockets in the board and see broken solder joints. It's mounted to a metal frame the whole thing comes up with a few screws on the front and back lip. Not all are bad. I think they military figured it out eventually. The newer boards that were intended for the uparmor trucks don't have that issue.

 

[fuzzytoaster]

Alright, the judge has spoken. Holt Cat took my ECM to their test machine and it wouldn't communicate so it's dead. It gave me the same error when trying to connect through the J1939 plug - "Failed to Communicate". They aren't aware of any alternatives except replacement at $1921.xx for a new blank unit. Then of course $250 to flash it to stock and another 3 hours @ $145/hr to calibrate it to make sure it's in spec. I certainly won't be taking that route on cost alone. Luckily they didn't charge me for the failed diagnostics, they were much more occupied with how I've come to own such a vehicle.

I'll be looking into a used or refurbished unit pending cost (any leads let me know). The service manager said the physical ECMs are all the same from that era but if starting with a blank unit and flashing it to stock settings dictated by the serial number on the engine it will require tuning to run right. This is obviously expensive and requires the truck to be physically at the shop so that's not an option. He did say they'd be able to copy a history or profile on a working ECM from an identical vehicle as a starting point and this would not require further calibration. I did mention increasing the horsepower to 330-335hp which is well within capabilities of the engine. He said with a good profile he could create a "power profile" that could be duplicated for other trucks and for a backup.

 

[TNriverjet]

Fuzzy, there is a relay that controls the "Start Inhibit" switch. It may be worth checking that out to verify it's not something that simple. Check out relays K19 and K24.



Also, I wonder if that excitation relay or circuit could be faulty?

 

[Suprman]

His is A1. Totally different beast.

 

[Suprman]

I bet Justin Whering would know someone with the right program you could send a used ecu to be flashed. I think he's in TX also.

 

[fuzzytoaster]

Fuzzy, there is a relay that controls the "Start Inhibit" switch. It may be worth checking that out to verify it's not something that simple. Check out relays K19 and K24.

View attachment 710080

Also, I wonder if that excitation relay or circuit could be faulty?

That's a different board than my truck, mine's the newer "green board" says Suprman. My board also has a green LED on the board of some key relays and the start inhibit was one. I also tried swapping a few to make sure it wasn't a faulty as I've seen before. My concern here is that the same issue that I had in the truck was replicated outside the truck (communication error) so unless I can find another good ECM to test with we won't know.

 

[lindsey97]

Machinerytrader.com usually has parts listed, you might check there, or maybe with local forestry service. Sometime cash talks at those places.

 

[Suprman]

If cat said it's bad then it's probably bad. I would move on to finding a replacement computer and not worry about troubleshooting further at this point.

 

[snowtrac nome]

regardless of where you get an ecm it's going to need to be reflashed. Only some one with cat e-t can do that, my suggestion would be to find someone in your area they can get an image off of, than it can be transferred to your new or refurbished ecm. The cat shop isn't cheap but the cummins or Detroit shops aren't either, the software subscription is spendy to keep active and it has to pay for its self some way. I'm not a cat dealer I do have software for about 8 different manufactures on my lap top, my customers are lucky that I only charge what the dealers charge in anchorage, as if they come out here there is a thousand dollars in airfare and per diem on top of the standard charges.

 

[fuzzytoaster]

Updating this thread for a final solution.

The original ecm was dead and no indication as to why. I had spoken with HOLT CAT many times on the issue and it came down to replacing the unit. The issue was sourcing an exact identical ecm #239-3879-00 which was discontinued in 2003 along with production of the engine. Using the engine serial HOLT CAT found an alternative part number that was more common and was listed for use with the 3126. #239-5303-00 if I recall right.

I went through a rodeo trying to get a working unit (new or used). One went MIA in the mail, the other the sender never shipped, and HOLT CAT wanted $1900 for a refurbished blank unit so another $250-$500 to program..absurd. One seller finally came through for about $550 with a working unit. HOLT CAT tested the unit, it was good. Attempted to program the unit and hit a wall. My engine had two stickers for HP rating and caused a lot of confusion (330 vs 275). They called corporate (Catipillar) who too was confused but shut them down saying their truck division could not program an ecm for my engine as my serial was designated industrial/off-road despite being in a truck. Previously power/industrial services division wouldn't work with me due to it being in a truck so...catch 22.

A bunch of higher Catipillar people talked and finally strong armed HOLT CAT's industrial division to service my unit. They had to call other dealers for the right flash file due to the little demand on this model. During the final stage of flashing I was told the ecm would not work as that unit was for on road trucks as it was designed to factor in emission calibration. So I was out of pocket for an ecm they said would work but didn't.

Speaking with the division supervisor at this point started to get the gears rolling in a more direct manner. He told me even if the ecm would have flashed and worked correctly it would not be timed correctly so it would be in a safe mode for idling only. He informed me the four options to fix this were:

1. Copy an existing ecm for this engine (no volunteers from the community).
2. Split the engine and transmission, install a flywheel, and calibrate it.
3. Use the current ecm and change injectors ($6000) and drop the hp down to 205.
4. Possibly sub-flash the personality module of the ecm giving it the timing code directly.

Like **** I'm going to pay to lose HP or split the truck. I took up the task again to find an ECM identical to mine. Apparently these 3126 ecms are not often rebuildable and those who could take them on a limited basis (about $1500 with a good core).

I did finally get lucky after 6 months of searching and found an identical new old stock unit from a company who liquidates dealerships.. $229.00. Done deal! But of course it's not that easy.

Though the unit was NOS it was activated as in a new ecm has a 24 hour "test mode" built into it that is activated upon first power to the unit. When the time is up the primary flash info is hard locked to the ecm and can't be changed. My unit was activated with 20 hours left on the clock and programmed for a 3401.

With some more communication to Catipillar the unit was cleared and ready to flash. The tech flashed the unit while in test mode and it immediately and permanently flashed the unit. Unbeknownst to the tech and myself the older flash files do not have a "test mode" and will immediately flash the ecm. luckily the flash was correct for my engine and we did a sub-flash to get the timing requirements in place.

The truck runs perfect now and no issues. Here's the points I made for simple viewing.

1. If you suspect a bad ecm, take it to a shop to test
2. Expect to pay high to get what you want now or be vigilant to get a deal on an ecm (they're not common)
3. Take a picture of your valve cover with all engine specs
4. Dont assume the dealer knows what they're doing with these trucks, ask questions
5. Try to get ahold of a good working ecm with data identical to your engine or else you may be SOL if they dont have the flash file.

Last note is the actual flashing of the unit. The onboard batteries fail on these ecms and can't be replaced (tried it). CAT's new method is to hard flash the ecms (hints the test mode) which permanently imprints the code to the ecm. Early model engines may still suffer from bad batteries leading to memory loss and my no start situation. Hard flashing is their solution to this. Trying to flash an ecm that has been hard flashed to a different engine will not work. It must match your engine or be blank.

 

[NDT]

Head still spinning after reading that. So basically all the early 2000's trucks are "ticking time bombs" as to when the hold up battery dies and takes the ecm with it? Wow, what a mess.

 

[coachgeo]

Head still spinning after reading that. So basically all the early 2000 trucks are "ticking time bombs" as to when the hold up battery dies and takes the ecm with it? Wow, what a mess.

while battery may not be replaceable... wonder if it could be supplemented???.... as in.. if energy will flow thru bad battery even though it will no longer "store it".... could one pigggy back another battery to it? Have not seen the board..... so wonder if.. bad battery could be un-soldered and new on soldered in? Granted that won't help if battery has already died and your ecm's program was lost during the lack of power time.

 

[fuzzytoaster]

Head still spinning after reading that. So basically all the early 2000 trucks are "ticking time bombs" as to when the hold up battery dies and takes the ecm with it? Wow, what a mess.

That's what I've read and had been told. That's also why Cat now hard flashed ecms. The battery in the ecm is supposed to stay charged by the truck but if they go below 9v they will "die". I've read some forums where internal batteries have been leaking acid and failing within the first 10 years.