Today my A4's ECU kicked the bucket

Status
Not open for further replies.
Joined
Jun 2, 2003
Messages
23,591
I was driving in almost 100F heat today. I suppose the heat killed my ECU. The engine runs just fine, but only after I reset the ECU. As soon as I start up before resetting the ECU, the car bucks like a mule and throws a CEL. A VAG-COM scan revealed an "internal control module ROM error." In short the ECU is toast. A new one is $1,400, but luckily one of my Audi friends lets me have a good one for $300. That's the closest I have come to breaking down with this car during the past 10 years and 138k miles! The other close call happend when the gas pedal broke off.
 
You might want to open the old ECU and see if there's leaking or bulging electrolytic capacitors inside of it. These can be replaced. For $300 I'd do it.
 
I'll make sure to check on the flux capacitor, Doc.
wink.gif
 
I have a serious disorder, leadfootitis, and I didn't know that the gas pedal was made from injection-molded ABS. After 70k or so miles of me driving in my normal manner, the gas pedal finally broke off right at the hinge. There was still a little nubbins left, so I could still drive with tippy toes. Thanks to my jedi power I've since been able to resist the urge to floor it beyond the point of resistance.
wink.gif
 
Well, Pablo, the Euro version of my Audi doesn't have a Hitachi ECU, but a Bosch one. The Euro '96 A4 also has a variable assist power steering, while the US version has just plain power steering. Of course, the price was much lower in the US. I wouldn't be surprised if the plastic gas pedal were also an exclusive feature of the US model.
 
ECU failure (without external causes such as a short circuit or connecting the battery backwards) is pretty rare. My mechanic, who has been in business for over 10 years, says that he has NEVER had to replace one.
 
The brake pedal is all metal. Believe me, that was the first thing I checked when the gas pedal snapped off.
 
Come times the air conditioner pump clutch had a diode that when the clutch turns off the magnetic coils generate a spike of around 60 volts ,sometimes that causes ecm failures, especially if the ecm goes out again,I hate electronics on cars.
 
Steve, that's interesting. Any idea how to make sure that's not happening in my case?
 
quote:

Originally posted by moribundman:
I was driving in almost 100F heat today. I suppose the heat killed my ECU. The engine runs just fine, but only after I reset the ECU. As soon as I start up before resetting the ECU, the car bucks like a mule and throws a CEL. A VAG-COM scan revealed an "internal control module ROM error." In short the ECU is toast. A new one is $1,400, but luckily one of my Audi friends lets me have a good one for $300. That's the closest I have come to breaking down with this car during the past 10 years and 138k miles! The other close call happend when the gas pedal broke off.

You've been jump starting any cars?
 
quote:

Originally posted by moribundman:
Wow, ten whole years. He's seen almost all.
wink.gif


Think about it for a minute--most cars didn't get
ECMs and fuel injection until the late 80s. Then they had to get out of warranty before he ever saw them for anything more than routine maintenance, which takes about 3 years (and I was mistaken--he opened up his business in 1992). So newest cars he would've been likely to see for repairs would have been 1989 models that had come off warranty.

When he started his shop, most cars coming in were still carbureted. I don't see how his having been in business longer would've resulted in seeing much more in the way of ECM equipped, fuel injected cars.

But none of this changes the fact that ECM failure is RARE. Trade journals even say..if you replace an ECM, check for external problems because they almost NEVER fail otherwise.
 
Does this ECM have the capability of being "flashed"?? Most do have some ability to be reprogrammed. My plain Jane DC PCM can be flashed twice before it has to be sent back to the man for a total purge (most reputable sources require you to turn in your old PCM before receiving a new one, since they have to program your VIN, mileage, and other info into the new one).
 
Alan, no I have not been jump starting any cars. I don't do that since it's not recommended.

brianl703, you told me a mechanic who's been in business for ten years hadn't see a single dead ECU. You won't be able to convince me that my ECU is not defective, no matter how much you argue that they barely ever fail. I know there is a company in Germany who does nothing but diagnose and repair defective ECUs, because car makers generally just replace defective ECUs. If a company can survive by doing nothing but fixing ECUs, then there must be some failing! What do I care if it's rare? If I win a million dollars in the lottery, I will still take the money!
wink.gif

FYI, even a 1982 VW Scirocco has EFI and an ECU, so those components have been around much longer than since the late '80s.

Gary Allen, the ECU for the '96 A4 with the manual transmission is plug-and-play. No coding is required. I don't know if the ECU can be flashed.
 
I'm not saying that your ECU isn't defective. It's just that if your luck were applied the other way, you might want to buy a lottery ticket..
smile.gif


There are companies who rebuild ECUs and make money doing because many ECUs do fail because of external causes, which will damage the replacement ECU too. Perhaps there's an intermittent short between two wires that shouldn't be there, or maybe a solenoid is shorted internally, or...

But for an ECU to up and fail with nothing else wrong? Very rare.
 
Yes, brianl703, we all can speculate. I can only go by what the VAG-COM scan revealed in form of two specific error codes, which are SPECIFICALLY pointed out in the Bentley manual as a defect requiring a new ECU. A quick inquiry at my trusty German dealership confirms "The ROM module does on occasion go **** up."

I installed a new (used) ECU, it was plug-and-play, no coding required, and the car is running fine.
 
That's very interesting about the ROM module being known to die.

That reminded me that certain type of ROM, called an EPROM (Electrically Programmable Read Only Memory) has a finite lifespan--I've heard 10 years is what was guaranteed on those EPROMs that specified a lifespan.

If it is an EPROM, there will usually be a sticker on the ROM chip which, if removed, will reveal a window through which you can see the actual chip. Exposing this window to UV light of the correct wavelength and intensity will erase the chip. EPROMs without the window are available to save costs, so lack of a window isn't a guarantee that it's not an EPROM.

EPROMs have a limited lifetime because the bits are stored as charges in an insulated layer and given enough time, the charge leaks away, causing the EPROM to revert to it's unprogrammed state. When the chip is exposed to UV light, the UV light makes the charge leak away in a few minutes.

The alternative to EPROMs, mask-programmed ROMs, do not have this issue because they are built with the correct programming inside of them--there is no charge to leak away over a period of years. Some automakers used mask-programmed ROMs and some used EPROMs. I don't know which, although I believe I read that Ford used mask-programmed ROMs. Mask-programmed ROMs are more economical for higher-volume applications, with EPROMs being more economical for lower-volume applications.

This is all very interesting because, back about 10 years ago, I was kidding around with one of my friends telling him that his fuel-injected car will eventually lose the contents of it's EPROM chips. (At the time I drove a carbureted car..)
 
Status
Not open for further replies.
Back
Top