E35B Post-Flash Issue - Am I alone?

[juddski88]

Yesterday I flashed 2 E35Bs, both flashed in a truck that has had successful flashes recently since the other truck is experiencing Canbus issues. Both Ecms flashed successfully, both were just receiving minor table changes. Now neither ecm will allow the truck to crank due to theft deterrent communication issues. Neither ECM is automatically discoverable with EFI Live, but they will take a flash and I can log them with the Tech2. Security Light has stayed off, I tried flashing with VATS off, tried flashing their stock files, tried updating EFILive and my firmware, all to no avail. Both ECMs run versions of the 8594 OS. I also just flashed a couple E35As the previous day with no issues using the same older version and v2 as I started these trucks with. Anyone experience anything like this?

Thanks,
Tim

 

[2004LB7]

Aftermarket radio?

Tried disconnecting the batteries? I had it happen to my truck once and disconnecting the batteries solved it for me and it never came back

 

[juddski88]

Aftermarket radio?

Tried disconnecting the batteries? I had it happen to my truck once and disconnecting the batteries solved it for me and it never came back

Nothing aftermarket in either truck that I've found. Unhooking the batteries in one of the trucks didn't do anything, however, I went to the other one (the one with Canbus issues) and it fired right up first try today...so...yeah...

It seems like either a CANbus issue, or weak immobilizer or something, but in my experience, the security light will stay lit when there is a real issue there. I can't connect to many of the modules with my tech2 in that truck, including the tdm, I can only look at the tdm messages at the ecm. Still trying to pinpoint the cause of the data bus issues on it, the modules that are giving problems don't all share the bus, they are HS and LS. Its a roached turd too, rot everywhere and standing water in the floorboards...

 

[1FastBrick]

Nothing aftermarket in either truck that I've found. Unhooking the batteries in one of the trucks didn't do anything, however, I went to the other one (the one with Canbus issues) and it fired right up first try today...so...yeah...

It seems like either a CANbus issue, or weak immobilizer or something, but in my experience, the security light will stay lit when there is a real issue there. I can't connect to many of the modules with my tech2 in that truck, including the tdm, I can only look at the tdm messages at the ecm. Still trying to pinpoint the cause of the data bus issues on it, the modules that are giving problems don't all share the bus, they are HS and LS. Its a roached turd too, rot everywhere and standing water in the floorboards...

That sucks... Some of the NNBS trucks like 07-08 had harness chaffing issues around the floor board foot wheel area. With standing water, It's also possible there is corrosion at the junction points or the grounds...

I do not envy you on this one...

BTW, do you have any way to bench flash the module?

 

[juddski88]

I've had over a dozen lmms in this past 6 months with broken wires in the body harness through the firewall to the fuse box. Some are beat, some are low mileage and well cared for, but none of those were data bus issues.

On the truck I have with CANbus issues. I started with checking resistance on the bus circuits, checking for shorts, cleaning grounds. Now I'm unplugging modules 1 by 1.

I don't have bench harnesses yet, but the flash process hasn't proven to be the issue for me. My initial hypothesis and reason for starting the thread was that there may be something funky in the files, once flashed with the last version of v8 that is causing a CANbus miscommunication. Kind of similar to the bug that was in the early e35b OS from the factory.

 

[1FastBrick]

It's possible

 

[2004LB7]

If you don't already have it

 

[juddski88]

Quick update: the truck without the CANbus issues that wouldn't start after the flash ended up being a coincidence also. Lucky me..
The UBEC panel had some loose female sockets and the DLIS circuit was open and there was an partial open in the power train relay circuit.

Still haven't resolved the CANbus issues on the other truck, but at least they both run and my concern that the flash had caused an issue can be put to rest.

 

[juddski88]

Another update:

The truck with Canbus issues ended up needing a new TCCM and new VCIM to resolve shorts to ground on the high side circuit of the HSCan.

The other truck, I assumed had no Canbus issues because I was able to communicate with every module fine until an ecm file flash, acrually had more issues.... After the successful flash, the ECM would show no comm. Tried 3 different ECMs. All would wakeup and initiate comms with the rest of the modules before a flash, flash successfully, and then show no comms right after. Found a couple pin contact issues in the fuse panel, switched it out, truck started once, and then went back to not starting, no ecm comms. Well, yesterday I find there is a dead short to ground on both the high side and low side HSCan circuits...lucky me

Edit: dead short to ground is only present after about 5 minutes of key being off, otherwise, while key is on and modules are powered up, no short to ground. Then when key is initially turned off, I get about 5000ohms to ground from both high side and low side HSCAN circuits. I've already gone through all the 12v circuits and grounds on the ecm twice, they are all as they should be. Time to go through the bcm I guess

Edit 2: BCM it is...gotta love these POS NBS electronics....granted, it's 13yrs old, but low miles and pretty clean c&c w/t bucket truck.

 

[1FastBrick]

What a fiasco... Thanks for the update.

 

[buick455]

The 5000 ohms to ground at initial key off is likely voltage. Even if the module is bad, it's still pretty likely the case. The modules don't normally "sleep" right at key off. At least on the heavy duty side of things, you can be keyed off forever, but if for example the door is open (or any module is online), you'll have continuity showing on a mm when checking data links but it's actually voltage. The newer the vehicle the worse it is but it's kinda fun honestly. I'm quite sure you already understand fully, not trying to step on any toes, just my perspective anyways.

 

[juddski88]

The 5000 ohms to ground at initial key off is likely voltage. Even if the module is bad, it's still pretty likely the case. The modules don't normally "sleep" right at key off. At least on the heavy duty side of things, you can be keyed off forever, but if for example the door is open (or any module is online), you'll have continuity showing on a mm when checking data links but it's actually voltage. The newer the vehicle the worse it is but it's kinda fun honestly. I'm quite sure you already understand fully, not trying to step on any toes, just my perspective anyways.

I still have a lot to learn about databus systems. I was comparing my findings to what the diagnostic procedure was telling me I should be seeing. The 5000 ohms didn't bother me much because it was consistent with what I was seeing in a couple of the other truck I had here to compare to. The dead short to ground after everything sleeps is the anomaly. I still don't have an explanation for why it happening, I also still can't explain why the bcm is causing an issue with ecm comms with key on, and not any other module nor itself.

 

[juddski88]

Did some more messing around with this truck today. Installed new bcm and updated it with SPS. No more short to ground on the HS Canbus with key off. Updated the ECM with SPS. Communications with the ECM to and from other modules was immediate, truck started. Read out the new ECM file with EFI Live, modified it appropriately, flashed it back in, and back to my original finding of no communication to the ECM from the other modules. Flash in the stock file I read out, still no communication with other modules. So it is a problem with EFI Live.

 

[juddski88]

So I flashed this truck with HPTuners yesterday, just to prove a point, and it worked fine, no issues, truck started. So I decided to try the 6/17 beta build of Efilive, and flashed in the newest version of my modified file, and the truck started and runs. Two issues I noted: I tried to read out the file using the beta release that I had previously flashed with the public release, before flashing with hptuners, and the read "completed" but finished with an error message saying .cam file was not detected, therefore, didn't get a complete read, but now I had a new data point. Also, there are now EGR codes and EGT sensor codes currently reporting that should not be....so still some bugs to work out with EFILive...

 

[juddski88]

Update:

Dealing with Paul at EFILive on this issue has been frustrating to say the least. They have not been able to provide any answers as to why the modified file is still reporting codes that Editor is showing are "not reported." I have provided multiple trace files of flashes and reads and they cannot give an explanation of why the post-flash comms issue is happening.

 

[DAVe3283]

Can they replicate it locally? Or do they not have a compatible ECU on hand. It seems really reproducible from what I've read here.

Sent from my FlashScan V2 using Tapatalk

 

[juddski88]

Can they replicate it locally? Or do they not have a compatible ECU on hand. It seems really reproducible from what I've read here.

Sent from my FlashScan V2 using Tapatalk

I narrowed it down today to an issue with their address for the APP expected voltage. Once in a great while, I was able to change the values and get the flash to fully take. Success rate was 2/12. Most times that I tried to recover the ecm with a stock file flash afterwards with EFILive, I was not able to. Had to flash with SPS or HPTuners. Cal flash, full flash, didn't matter. Also, I still don't have an explanation as to why p0405 code and other related codes were still active after being shut off in the file...the only way since upgrading to the latest beta release to keep those off was to first flash with hptuners and then read it out and modify the rest with efilive.

The reason why I was trying to change the APP expected voltage for circuit 1 was because this truck has an issue...I found a while ago bad wires in the harness for circuit 1, ran new wires, load tested fine - .1v drop at 12.7v, still had intermittent issues. Scope showed a random very momentary drop to close to 0v on circuit 1 voltage signal. Replaced the pedal with new gm. Didn't fix it. Replace ecm. Didn't fix it. Load tested the 5v ref wires to the APP, no issues. So I was like, ok, I'll just change the threshold in EFI Live since it allows me to do so. Well..not so much apparently.

All I can do now is, again, take the connectors apart and attempt, again, to tighten the terminals.

 

[juddski88]

I received a response back from Paul again, saying they have absolutely no idea why changing that APP parameter value is causing an issue. My best guess is that it is changing the engine cal ID checksum to one that is unrecognizable, because when the corrupted file is installed, no tools I have are able to properly identify the segments.