Bricking a PCM while on the road?

[winchy2]

Hi guys,

Does anyone know what can possibly brick a 1995 LT1 PCM not while flashing, but while driving?

I modified my PCM a few years back with the well known hack of socketed flash ram. Although I only did the T side because, back in the day, that was the way it was recommended. Not the E side. It had been working well all this years. I had reflashed it a few times inside and outside the socket without any issues.

The other day, I took the car out for a spirited sprint and when I let off the gas, I got the dreaded check engine light, the car stumbled bad and eventually shut off. I could start it again but it immediately died. Had to be towed home. Once in the garage, I hooked up my handheld and there was no communication with the computer. Odd!
I brought my laptop out and through several of my tuning apps, also got no communication. I ended up pulling the ram out and reflash it with the base T bin hoping to reestablish coms. Nothing! Nada! Zip!. Totally dead. The only thing I have not tried yet is to perform the other half (E side) of the surgery, bench flash base image and see if I can talk to the darn thing but it it wasn't needed before why would it be needed now?

The was no signs of issues or problems prior this happening. While driving there was no indication of power loss, sparks flying or anything out of the ordinary that can point me to where to even start looking.

Anyone has experience anything like this? Any ideas what could have happened or where to start checking? Even if I bring the computer back from the dead, I fear it may happen again if I don't find out.

Thank you!

[dave w]

I've only had 94-95 LT1 PCM's brick during flash.

Perhaps 94-95 LT1 software (Flashhack) from EEHack website will help? https://fbodytech.com/flashhack/

For the few instances that the LT1 PCM bricked during flash, I sent the bricked PCM to TunerCats ( http://www.tunercat.com/ ) to have the flash chip replaced.

dave w

[winchy2]

Thank you Dave. I have never bricked one during flash. I guess I have been lucky. Flashack was one of the tools I tried but I was not able to get a link.
I don't understand how the PCM could die while driving. I thought these boxes were pretty hearty.




I've only had 94-95 LT1 PCM's brick during flash.

Perhaps 94-95 LT1 software (Flashhack) from EEHack website will help? https://fbodytech.com/flashhack/

For the few instances that the LT1 PCM bricked during flash, I sent the bricked PCM to TunerCats ( http://www.tunercat.com/ ) to have the flash chip replaced.

dave w

[kur4o]

Check for proper seating between chip and socket. It is not unusual that it will have bad contact or bad seating.

Other than that the voltage regulator inside the pcm can goes bad.

With flashhack you can try reading e-side only, to narrow down some of the unknowns.

You need to make sure the chip is flashed with scrambled raw bin other than regular bin read from pcm via aldl line.

[winchy2]

Thank you kur4o.
I do have the plastic spacer in between the board and the chip. It was very tight to pull so I am guessing it has not move since last pulled.
I can do a quick diagnostic on the regulator see if something is amiss.
Flashhack did not connect at all. Neither all my other tools. I can pull the chip again and try to scramble the bin, but I am using the raw T-bin that I have always used when I can't connect. Only then I can load my own .bin. --- Weird.

[NomakeWan]

Honestly I think kur4o is on the right track. Sounds like it's an age-related failure of the socketed part. In general, these PCMs are insanely robust and don't just 'fail'. They can be bricked by a bad flash, but now that Flashhack exists the chance of that approaches zero. So nowadays, a stock unsocketed PCM is essentially a non-issue.

So that being the case, I too would look towards the custom work surrounding the socket and see if something has broken. Maybe reflow the socket, see if it's just cold solder joints.

[steveo]

i bricked one from a busted ground wire

ive seen them die from corrosion

ive had bad solder joints or bad chip height when socketed

seen the ribbon between the two boards crack

thats about it

[winchy2]

Thanks NomakeWan. I am inclined to agree. It may have been its time. Although, as I understand it, if there is no communication to the computer the only way to bring it back to life is by pulling the T side chip and bench flash it with the programmer.

After checking and possibly reseating the PLCC32, I'll try to socket the E side and clean flash it, even though I am pretty certain it has nothing to do with the communication link.
Hopefully I can make this one come back to life.

[Tom H]

if there is no communication to the computer the only way to bring it back to life is by pulling the T side chip and bench flash it with the programmer.

After checking and possibly reseating the PLCC32, I'll try to socket the E side and clean flash it, even though I am pretty certain it has nothing to do with the communication link.
Hopefully I can make this one come back to life.

I detailed a way to reload the PCM without removal of the chip. Look for bootstrap...

The conformal coating is there for a couple of reasons. The coating prevents condensation from forming on the board. This condensation can result in oxidation of the tracks, breakage--> brick. Also, depending on the drive of each of the parts, the leakage can result in errors on the bus --> brick. All the other failure modes relating to sockets --> brick. If you are going to uses a socket, try to find one with a thicker gold plate. Vibration will cause fretting corrosion between IC lead and socket. Just a ton of failure modes. Above it was indicated that the ribbon can fail, I had this also due to handling. This was easy to fix for bench work, but very hard to fix to put it back in service.

Suggest working with an untouched PCM, working through the header and reloading that way. Once the conformal coating is broken, all bets are off. The coating is a PIA for the manufacturer, results in higher scrap, cost of chems, no rework or bed of nails test once complete. They put it there because it was necessary.

Hope this helps -Tom

[winchy2]

Thank you Tom. I agree with you on the conformal coating. Maybe, after I do some checking under a magnifying glass, I can brush in some Urethane or something like that.

Also, looking for the bootstrap info, I got to your other thread on the hardware research you are doing. Pretty amazing stuff! As an EE myself, I love this type of work. Keep it up! :thumbsup: