WinFlash flashing suggestions 16188051

Had (2) 16188051 PCM's brick while flashing with WinFlash via benchtop. During the same time I flashed one 16188051 at least twice on benchtop and then once in the car - so my cables and software must be OK.

On my laptop, charger plugged in, I turn off screen saver, anti-virus, speakers and wifi, even turn off my cell phone. The 12v battery is fully charged, twice now halfway through the flash..it stops and I get the error message. Once it fails to communicate there's no second chance no matter how many times you retry to program.

Just sent one PCM to CATS for repair and have a Cadillac LT1 PCM showing up Thursday that needs to be flashed. I have no choice but do the flash to get a car running.

Moates web site has this:

"BEWARE: these ECUs are notorious for “bricking,” or being killed by a bad flash program. Please ensure that there is adequate battery voltage at all times. It is a good idea to put a charger on the vehicle while programming as a drop in voltage for one second (fan kicking on, headlights, turn signals, subwoofer, etc.) can be enough to kill the ECM."

Would you think keeping a charger on the battery (12 volt 2amp) during a bench top flash might be too much voltage?

the PCM is going to draw around a half to one amp with no other loads being placed on it, a 2-amp charger should do just fine as long as it doesn't have any kind of "pulse" mode. it should either taper off charge current near 14.4ish volts(temp dependant) or drop into a float mode and maintain 13.3ish volts(also temp dependant), either case is more than adequate for ensuring a safe flash.

there's a linear regulator in the ECM itself that should ensure a decent voltage as long as input voltage is above ~12v. too much voltage should be no problem. winflash itself checks ecm voltage before beginning the flash routine and throws an error denying the flash if it's too low, but it doesn't check it while flashing is happening.

but, i'd be worried about a shitty ripply power supply on your benchtop causing problems in a bench setup. even though the ecm is hardened against power fluctuations, the flashing routine may be sensitive to power quality. a battery provides very linear dc output with no risk of distortion. i'd be tempted to run your benchtop setup from a real battery or a proper linear power supply rather than a switched mode one. there's NO way the ecm will draw enough power to kill a reasonably healthy battery, especially out-of-vehicle with no accessories.

seriously, i have flashed 8051 ecms thousands of times in-car with the stereo on with varying battery voltage and NEVER bricked one. i've even done it a bunch of times with the stereo cranked, thinking 'who cares'. never run into problems. these things shouldn't be spontaneously screwing up a flash on you.

if you want to be sure your data line's integrety is ok, run flash read operation a half a dozen times on an ecm before a FLASH. the read will probably fail checksum if you have a bad connection or whatever but isn't risky. i'd also call it good if it can pull a datastream for 10 minutes or so without any checksum errors.

last thing is don't use an ancient bin when you're programming, some original $EE bins had a TSB on programming failures, and there was a fix in later bins. it does bootstrap the flashing routine from stuff from your bin file, so using a really old bin before the TSB was released may result in increased failure rate due to the old flashing code, although i have a theory that winflash contains its own copy of the flashing bootstrap code and uses it instead, i can't really confirm. any calibration number from the left hand column on my site should be cool as a base.. http://fbodytech.com/bin/index.html

I would suspect either a poor benchtop power supply or a flaky cable that drops or corrupts the data every now and then. You say battery so you're using an automotive battery on the bench????

A car with a charged battery and a good electrical system shouldn't have any issues supplying enough voltage to the PCM while reflashing even without a charger.

steveo

Thanks for the help, been using a fully charged 12 volt battery as my power source. Ensure that it is charged before any use. When you say "run flash read operation" do you mean via WinFlash?

Been using my laptop without any datastream capabilities.

By the way I downloaded a custom created bin from fbodytech that had the car running great (bin attached). The problem is the street rod with LT1 driveline was driven into a shop for frame bracket welding last Saturday. After the welding..the car won't run...hope I narrowed it down to the PCM, the replacement will be here Thursday and requires the attached bin.

(No they didn't disconnect the battery)

Sorry lionelhutz...your response came in while I was responding to Steveo

Thanks again

the AN28F512 requires VPP to be 12.0volts +/-5% during a flash. assuming it is setup the way the P66 V6 is, then VPP is switched via transistor, a FET of some variety, so voltage drop shouldn't be too significant. acceptable range is therefore 11.43ish-12.6ish volts. I'd keep it on the higher side of that whenever possible.

I imagine in the event of battery voltage nearing/above 12.6, there is perhaps a zener or varistor on the VPP circuit ready to bring voltage back down to normal levels.

OK guys:

Finally got this PCM working after the modification:

This topic got me up & running:

http://www.gearhead-efi.com/Fuel-Inj...able-image-for