New PCM, is it dead?

[antus]

The com port chosen does not sound like the cause. Com port is a yes or no. I don't understand what you mean started on one, then swapped though. The device is only appearing on one at a time, and if you have the wrong one it wont see the device, if it sees the device then the port is right. If you cant get it to init properly and you need to init on one then swap the cable then something is not right and you need to sort that out before you attempt something dangerous like a clone write.

While we are talking about ports I will say bluetooth on windows does not have a good reputation and for something that needs stable communications like a write you are always better of with wired. If you were using bluetooth and windows bluetooth dropped out mid flash, that could have been the cause. As you mention cable length, I guess it was not bluetooth, though.

[ProfessWRX]

The behavior is strange. When I open the app it won’t initialize on com4. I believe com5 initializes every time. But I can get com4 to initialize after com5 does. So I switch usb ports to allow a better place to sit the PC. If com5 is better and initializes every time like I think it does I will stop using com4.

[antus]

Yeah. Its hard to know or explain what windows is doing but let me guess. Its reserving the port number for the physical USB port on your laptop. So when its plugging in to the first side, its on com5. I am going to guess thats the second place you plugged it in. Then you move it to the first place you plugged it in on the other side. Windows lets go of com5 and installs the same driver on com4, and it appears there. Why you have to do it in this order and why com4 doesnt work at startup and even wierder why it works after plugging it in as com5 first is suspicious. I dont know what would cause that, but it sounds like there is another factor here. It may be worth buying a slightly longer cable.

[Gampy]

Personally, I think you should quit using the OBDX VCI until it is fixed, so you do not keep soft bricking PCM's.

Maybe purchase a good VCI (not saying the OBDX is bad) that seems to work real well, like this (Best quality For GM MDI Scan tool with WIFI V2023.07), it has been a good VCI for me and others that I know have purchased it. It most definitely has no hoops you have to jump through to get it to work or keep it working that's for sure. It just works!

Even though mine came with factory manufacturing flaw, they did offer to repair it, I just had to send it to them, I fixed it myself, it was just a cold joint on a chip set. it was also several revisions ago 2020.09.

And I would also build a bench harness, so you can flash out of vehicle for now until we get more use and testing done.
Following this schematic:

P01-P59 Bench Harness Schematic.png

Just my opinions ...

-Enjoy

[antus]

I dont think the xpro is the problem here. When you get a new pcm, dont jump straight to the clone option, do a few test writes first and post the debug log up. That'll tell is if there is anything wrong with the communications in the car/computer/interface before risking the pcm. Then I agree with gampy, do your first write on the bench as a test calibration write only and not a whole clone. The P04s are easier to brick than the later designed ones. Instead read the whole pcm, load the file in universal patcher and change one thing, save as to a new file, then do a calibration test write of that on the bench. Only if thats ok, then do a real calibration write on the bench. If that works, then you can start testing clones on the bench. Save the debug logs as you go incase something happens so we can go back to see where your issue first started. If everything is working on the bench, and you are prepared to take the risk then move to the car. But I agree with MudDuck here viewtopic.php?f=4&t=8008&start=170#p124597 the voltage you are seeing in the car is too low for a reliable write. You may be poor electrical connections, or a low battery. You really want to see at least 12.4v at the pcm before trying to write.

[Gampy]

Remember the P04 is different, it has no calibration segment like the others, it's calibrations are entangled in the Operating Systems. Typically the very top OS segment.

In order to write a P04 Calibration change you have to do a full write ... I.E. Clone.

It is also not easily recoverable like the others, it will hard crash (lose OBDII comms), at least as far as I know!

If someone knows different please school me!

Simple VATS change ...

P04 VATS Only.jpg

-Enjoy

[antus]

Good point, thankyou. We will need to re-work the UI to take this in to account for the early ones

[Gampy]

What do you mean by early ones??

It's the same for all P04's that I know of ... That is why there are so many damn Os's, every calibration change is an Os change!

-Enjoy

[kur4o]

Calibration part is at the end of file, usually last segment, It is possible to write only this segment since checksum is in it.

It is too bad calibration doesn`t align perfect with chip segment part, and very small part overlaps.

There is some cases where OS part is the same for different calibrations[different OS number assigned]. In that case you can update OS only with calibration write.

I figured some way to identify this using some end of OS pointer abd crc-32 checks.

v6_interchange.JPG

[Gampy]

End of file is top Sector of chip!

-Enjoy