11sORbust

I have nowhere else to turn...... Does anyone here have experience with flashing LT1 PCMs? I offered to help tune a friend's 94 trans am. Figured I would do him a favor, but it has turned into a disaster. Ever heard of a tunercat error labled "incorrect message sent" ? This error has disabled 3 PCMs now! LMK, if you can help!!

Grumpy

On a 8051?.
I flashed one, a min of 100 times without a hiccup. I made up a mini harness, and used a lawn mower battery to do the reprogramming, out of car.
Also, used a AKM cable.
I've heard all sorts of horror stories, about them, but using the above things went well for me.

funstick

it will always do that when the fans kick on. its best to umplug the fans and put a battery charger on the car and set it for 2amps. ever since i started doing it that way havent had a lockup since. also to note. make sure your using a quality cable. i have also had the same problem with everyone elses software.

11sORbust

yep, it's the 8051. Tunercat just sent a very helpful reply. I'm going to do what they suggest once I get the chance. It has been a nightmare, that's for sure! First day I tune his car I flash the PCM several times, recalibrating the maf tables and some transmission stuff (now has 4.10 gears). It got late so we called it a day.

Day two came and I brought with me a nice little bin that I worked on at home. Hook up everything and it was working perfect. It erased the flash memory and programmed it to 99.9% and then an error popped up " incorrect message sent" . Right after that my laptop crashed and I heard the dreaded cooling fans. Turn off the ignition and laptop, unplug the ecm for a long time. Then I plugged it back in and tried to program the PCM and it failed. That was the first PCM.

Day three came after much searching through the boneyards(and reinstalling the OS and all my software). One hundred bucks got me a roadmaster PCM. Great! So I go back to try this again(day 3). Read the PCM two times. Both bins looked perfect, no errors. Figure it's time to flash...... it erases the chip then right as it started to program, I see the same error (incorrect message sent) . Cooling fans pop on and I know it's done for. Check the battery voltage and it was 10.6 . I figure that is the problem so I get another PCM and tell my buddy to charge his battery before we try to flash the next PCM.

Find another roadmaster PCM and grit my teeth for day 4. This time I go to the extreme and turned off my cell phone. Last time it was the only device that was not turned off, it didn't ring though. Always made sure that there was no chargers hooked up....followed every direction. Pulled the battery out of my car this time. It measured 13.X volts and is in great condition. Used a turbo charger to bring my battery up to 14 volts. Disconnected the charger and tested my diag cable with datamaster, it checked out good. Then, I read the PCM. It worked perfect....even seemed maybe a little faster than before. Go to program the PCM and it's working. The memory gets erased perfectly, then right as it starts to program I get the same error.

That was today. So I'm going to do what TC suggested and if that doesn't work then I'll send my cable to Andrew. See what he can find. If that doesn't work then I'm up a little brown creek without a boat.

Guess the bright side is that I have a PLCC to DIP adapter,a few flash chips and some PLCC sockets coming in the mail. My chip burner supports AN28F512 chips, it just has a DIP socket though. In the end, I'll win!! Just need to step up my soldering skills, DOH.

11sORbust

Tunercat says not to have a battery charger plugged in.....

I am starting to think about a "flash bench".

87tpi7749

That is good advice, I burned up two scan tools at work, a Tech 2 and a Ford NGS scan tool. The service techs said that a small amount of A/C voltage from the battery charger will "ride" the D/C current right into the scan tool and under the right conditions bam! Scan tool dead. Although this tid bit doesn't help your reprogramming problem directly, it may keep your laptop alive. Now I use a jump pack on customers' cars when I scan or reprogram.

Grumpy

*Starting to*?.

11sORbust

Give me a break, I can only corrupt flash chips at this point!

11sORbust

Give me a break, I can only corrupt flash chips at this point!

jwscab

Yep, I heard all the horror stories, so I flash them on a bench with a benchtop power supply. Radio shack makes this 13.5V bench supply, it has banana jacks, and a cigarette lighter socket, so you can test car type stuff. I made up a small harness based on stuff found on the web, that has its own aldl adapter, use a laptop that is plugged into AC, and so far, no issues. Run nothing on the laptop, power save mode off, no screen sleep, no other applications. Super cheap insurance, as you know after 3 ecms. I think the supply was like 40 bucks.

11sORbust

Well, the PCM was flashed several times, in car without any problems. I even used a turbo charger to bring the battery volts to 14.0 before reading or flashing. I'm under the impression there is hundreds of guys flashing 'in car' without any problems. Don't see why I would NEED to setup a bench. After this, I might just flash PCMs on the bench......

Grumpy

BUT, NOW YOU DO ! !!

I'm continually at a lose of when people ask a question, they get told the answer, and then want to hemhaw around. For all the time reposting to this thread, you could have used that time toward build the bench. A couple power leads, grounds, ALDL, and your done.

While, jwscab, is using a power supply, I'd be leery of any A/C ripple during programming, but he seems not to be having a problem.

Build the bench, and then you'll have a stable KNOWN starting spot. You HAVE to have a stable foundation to build anything on.

junkcltr

iTrader: (1)

Too funny.

Bench setup is the way to go. Yes, in car can be done. That PCM requires +12v minimum to be able to program it. It doesn't take much to have the battery in the car dip to less than that for a few hundred milliseconds and then the PCM is locked-up.

11sORbust

That is one thing that kills me, when looking back at my old threads. I have been told the right answer but kind of ignored it. Guess it's human nature to only highlight information that matches one's intuition.

Good thing is that I have learned the hard way before and just constructed the bench (about 10 minutes ago). Might not look that snazzy but it should work great. I'm putting in some extra hours to test it tonight.

Should also have a tech 2 tomorrow, my brother in law works at a dealership. He also said there might be a diag tool that'll flash PCMs standalone style. You can bet I'll be grabbing every factory calibration possible.

Will post pics tomorrow!

funstick

yes ac ripple is a concern if your charging at 10 amps plus. but at just 2 amps. ac ripple is usually under .010 millivolts. the ecm and scan tool both have filtering. and yes i flash tons of pcm's in car no problem doing it this way. again a good battery charger. or run a second battery on jumper cables. either will work just fine

Grumpy

Got a link to justify that statement?.
.010 millivolts?, you'd just about have to be at Livermore, or Scandia to measure that.
A shorted alternator diode, can put enough ripple into the system to totally confuse a pcm. BTDT

There is some filtering, and zenier diode protection, but it has it's limits. At $75+ (min) for a pcm, IMO, it's just not prudent to guess about what works, or might work.

Craig Moates

I've fried LT1 PCMs without a battery charger, but I always set a 2A trickle charge on the system and reflash in-car without problem.

Then again, I always hold my breath when I hear those fans kick on. I always ask the guy if he has access to a spare PCM beforehand.

Bench setup is the way to go. Battery charger has been OK for me, but not without nervousness. Plus, no telling variability of charger A -vs- charger B, right?

87tpi7749

Can these bad or misprogrammed pcms be brought back to life? Do they just need to be reflashed with an original calibration file?

Grumpy

Nope, someone like AKM has to fix it. A chip needs to be removed, erased and then reinstalled, as I recall.

11sORbust

...

11sORbust

Source of information?..........

11sORbust

So, a few ms below 12v, that's all it takes? Have you fried any PCMs?

TrevorH454

With a bad ECM, any reason you couldn't have a zif socket soldered in and use eproms so if you do screw it up, it wouldn't be a big deal to replace?

11sORbust

IMO, a ZIF socket is overkill, if they even make them for PLCC. Not going to take much force to seat a surface mounted PLCC into the socket.

funstick

i dunno. i got a high resolution osciliscope. they do read down that far you know. the newer 100mhz stuff. also i put the decimal in the wrong place. 10 millitvolts. i was thinking on the wrong scale. it was late i was tired. if 10 millivolts cuase a programming error you got bigger problems. also bench programming has its short commings.

again unplug the fans use a good clean battery charger set it for 2amps. everything goes fine everytime. what needs to happen is the bootloader needs to be rewritten to stop the damn thing from halting program until it hits like 10 volts.

11sORbust

I was looking at my chip programmer. It'll flash 28F512's at 12v . Now I understand, the chip needs 12 volts in order to be reprogrammed(has nothing to do with the PCM). Wonder what happens EXACTLY when a chip is programmed below the recommended voltage.?!.?!

11sORbust

Thanks for the clarification. I have a little 3 amp charger that'll work perfect. Really hope that is the problem(low voltage). I'm buying one more PCM to try out. Kind of sucks, going through 3 PCMs. Guess I'll just socket all of em. Should have the PLCC-DIP adapter and sockets this week.

One thing that I'm confused about. Andrew only replaces one flash chip in this link http://www.mindspring.com/~amattei/flashfix.htm
Is that one enough to restore PCM communication? I plan to replace both flash chips....

Grumpy

And those are?.

funstick

if the ecm sets a code on the bench it wont reflash. and your bench power supply still might not have the power stability. useually currupt flash happens when youve got a crappy laptop. but unplugging the fans and hooking up a battery charger. works great.

funstick

the chips have mixed adress lines. one has part of this one has part of that. the bootloader which is what you need to be able to flash is most likely on that particular chip or can be put on that particular chip.

11sORbust

bootloader? I'm just planning to copy good OEM flash chips(for both boards). It would save a bunch of time to socket just one board!

Craig Moates

There's some stock binaries you could get from someone around here that would probably save you socketing another for the sake of getting the binaries off it.

Remember that the address lines are mixed around.

11sORbust

I have looked around and came up with nothing. The only thing I can find is TC/lt1 edit dumps, don't think that is the same thing. Funstick mentioned this bootloader program/routine. Sounds like I just need to burn this to get the PCM to reflash. Waiting for clarification.....

Do you know how are the address lines distributed exactly? I mean, are the tables split up?

junkcltr

iTrader: (1)

There is no "exactly" in electronics. Anything designed as exactly will not work reliably. Although, I can think of a few things that are designed this way and people just deal with it.

Depending on what factory, what time, and what die rev. determines exactly how the chip may react to undervoltage. A manufacturer will not spec. how a chip will work with out-of-tolerance specs. That is why they have specs in the first place......so that the chip will work.

Also, you need 12V at the chip. That means you need the battery higher than that due to voltage regulation and cable loss.

Not surprised at the A/C filtering for the voltage going to the chip.
100Mhz scope shows 10mV of A/C ripple??? Is that with 0mV of ripple into the PCM or 5000mV of A/C ripple into the PCM.

It doesn't matter how much ripple you measured at the chip. It matters how much A/C ripple was rejected from getting to the PCM. That is, maybe your battery charger only puts out 10mV of ripple.

Mixed up address lines???? You mean the code is scrambled (poor mans encrytion) or that MCU address line x is connected to EEPROM address line y. Think about it, if it was this way then programming the EEPROM would not work. The EEPROM is not like a RAM that is just a matrix of memory cells. An EEPROM usually is section specific. Or maybe just the lower non-sector address lines are "mixed up" . Anything is possible I guess.

11sORbust

Put an order in for a reman PCM, it was $114 . Then it dawned on me, what if the reman PCM is socketed? That would be nice.

junkcltr

iTrader: (1)

Manufacturers usually do not socket devices on purpose. Sockets and connectors are a failure point.

What really happened is that GM messed up and allowed the bootloader code to reside where the motor/trans code resides. You will find that even to this day some people still do this. It is a bad design practice. You will see things like "updating your firmware with a bad/corrupt file will render your hardware useless.....send back to manufacturer". That is, they make the thing boot to the application and then try to jump to the bootloader which loads a new application. Then the application gets messed up and can no longer call the bootloader. That is what GM did. Ooooopppps.......maybe that device doesn't do sector erases which means it would have to be done that way.
Anyway, just another way for GM to cash in on things not going smoothly. Overall, I think they have solid designs though.

11sORbust

No details on this "bootloader"? Me confused

clippjr8

Go to ION at this link. He can answer all your questions and take care of fixing your fried pcms. His prices are great also.

www.madz28.com

11sORbust

This helped some....
http://www.oz1bxm.dk/PIC/bootloader.htm

This program might work with AN28F512's????
http://64.233.187.104/search?q=cache...r+28F512&hl=en

11sORbust

Uhh,.......

Besides, I sent him am email and still waiting(over 2 weeks)...

clippjr8

Sorry to hear about the wait. I just call him and he is always ready to help. His number is listed on his site so you can call if you like. He will help you even if you dont want to buy anything from him. Anyways, good luck and i hope it all works out for you.

11sORbust

Thanks! I'm hopin that I'll be able to solder the sockets myself. At least I have a few PCMs to learn on.

funstick

the 12volt programming voltage issue is in the bootloader gm designed for the lt1. the mcu itself operates at 5volts. the memory on the board programs at 5 volts. i highly doubt the regulator on the board has a loss of 7 volts. the mcu with the bootloader on it has an enable disable programming voltage of 12.1volts rewriting the bootloader to run at 10 volts would most likely fix this problem forever.. with a battery charger you should be able to maintian 13.5to14 volts. also yes the ecm should regect upto but not to say for certian around 400mv of ripple current to within 10%. as for most battery chargers. a good commercial unit with broad power settings typically like my snap-on unit onlys generates around 10-15mv of a/c ripple at the 2amp charge rate. also the battery acst like a giant filtering cap. ive yet to be able to measure a/c ripple on my battery charger in anyother place in the car aside from the battery terminals or within sevral inch of them. .but then again im spoiled.

junkcltr

iTrader: (1)

You might want to read the datasheet more closely. It is programmable ONLY when +12V is applied to VPP. Just like most of the EEPROMs manufactured during that time span. Yes, if you only want to read it then it doesn't need 12v. Since the VPP current is low, the voltage drop would be low. That is probably why the PCM can be reprogrammed without the engine running.
I think you got confused with the power supply (+5V) of the chip.

Luckily, I never have to worry about a battery charger ripple. I tend to just use jumper cables to get my rigs up and running.

A "bootloader" is just a piece of code that receives the more code and stores it inside the chip. If you mess up the "bootloader" part then you can't store more info in the chip.

junkcltr

iTrader: (1)

Wow, how did you find that? Your investigation skills are certainly up to par. My language skills are a bit a lacking so I got bored with it pretty quick.
Most bootload code works like this: MCU is waiting around for a command via a port (e.g. RS-232 interface). It gets the "I want load a new software algorithm" command. The MCU then starts reading the RS-232 port and storing the info into EEPROM.

Sometimes the "new software algorithm" includes the bootloader code that resides in the EEPROM. If the download of the new EEPROM info gets corrupt with some of it being the bootload part.......then the MCU is now hosed because the EEPROM is hosed.

The better was to do it is always have the bootload code seperate from the "application" code (e.g. the stuff that runs the engine). The thing is that this takes more work so most people shortcut it and mash it all together.......and then say return to the manfacturer to get it fixed for X dollars. Just like the companies that charge for "firmware updates". Basically, they are saying that they have bugs in their software and they want you to pay for them after you already bought there hardware and have been debugging it for them.

Bootloaders also allow the "application" to be encrytped (all f'ed looking to the naked eye) and then the bootloader decrypts it as it is stored into the EEPROM. Lucky for us GM didn't do this back then. Are they now?

Ahh, one more reason why I think things like the DIY WB are a great thing.

11sORbust

Do you guys think if bootloader stooped programming it would cause the error message I received? TC has an error msg for low voltage, never came up.

How much space does the bootloader code take up? The 28F512 are 512k chips, right? TC dumps are 128k. FYI, you can convert that one link to english....it didn't help me comprehend it though.




Nothing special but it should work. I plan to use real PCM connectors soon.

junkcltr

iTrader: (1)

Your best bet to figuring out what caused the comms error is to ask the software vendor what errors will cause that error message to pop-up. I would guess that any type of no response from the PCM so the causes could be endless......just a guess.

The PCM chips are 64K by 8 bytes each (64 * 8 = 512kbit).
As a comparison, each of them are twice as large as a 730 ECM PROM (32k by 8 = 256kbit).

I got an LT1 engine with the PCM last year. Maybe this year I will get my emulator going with the LT1. Would be nice to emulate on the LT1 stuff, wouldn't it?

BTW, that doesn't look like a lawnmower battery!!

I collect junk so if you mess up one of those dead PCMs I'll give you a few bucks for it before you throw it out.

11sORbust

The vendor says that anything from playing drums on the steering wheel to incorrect com port settings. Yeah, the possibilites are endless. TC suggested to maybe play with the com port settings. My com port settings(default) was good enough to flash the PCM several times, back to back. I doubt that is the problem. Figured if the voltage was too low I would have gotten that error(instead of incorrect msg sent).

My plan is to reuse the PCMs. Once they are socketed, I'll laugh at error messages......hopefully.

Ok, so there are two 512k chips and only 128k is for memory storage?? You don't know if bootloader is included into that 128k TC dump? Or know if it's on one chip, or distributed through both? Like I said before, It would be really nice to just socket one board. My soldering skills are not that great. DIP sockets look like cake compared to PLCCs.

I'll have a copy of each flash chip soon(both boards). Might throw em out there if people are interested....

11sORbust

doh

11sORbust

Hmmm, no desoldering?!?!! Now I have two questions. Where can I download motorola's bootloader program and how/where to load it though the test conector.