*full disclosure of possibly irrelevant red herring*
My car was broken into a while ago. gauges smashed/fuses blown/bias against cops once again confirmed.


The Trans Am has been sitting the garage for way too long. I've decided to start working on it again to pass emissions. It seems to be running way too rich and one fan is always on, which I seem to remember indicates limp home mode, but I thought limp home mode set the SES light on and/or a code. Neither of which symptoms I'm getting.

Symptoms:

1. As mentioned one out of two fans is on as soon as ign goes to on
2. While my AutoProm is able to connect to the 7730, the data it gets back appears to be junk. The virtual dash just rapidly fluctuates all values between 0.08 and 0.10. The values list is equally jumpy. I can attach a log file if anyone wants to see that bit.
3. Upload/Download to AutoProm works just fine
4. The parking lights turn on/flip up the headlights. Yeah.
5. Car starts instantly, but runs pretty rich and doesn't have the smoothness it should
6. RedHerring #2 MSD BTM no longer works when wired in. I have to use stock ignition.

Questions:

1. Would a blown ECM cause any of these symptoms? How can I verify that I am in fact not in limp home mode?
   
2. Is there a simple way to test that ECM is running code from the AutoProm? Change an idle variable or something?
   
3. Is there any way to isolate and test the ECM related variables? I don't mind picking one up on eBay, but I don't want to be that dumb mechanic that just starts replacing parts without understanding what's going on.

 

nebulous,
sorry to hear about your car man... thats just rediculous.
I am having a couple of the exact same problems. One that stands out is the fans coming on as soon as you turn the key. I think you have diagnosed it a lot more than i have at the moment. Tomorrow I am going to go to the JY and see what i can dig up. I heard that a Baretta has the same ECM as my car.
Anyways I think your problems could definatly be ECM related. I am pretty sure I caused mine when I forgot to hook up the main ground wire while starting the engine. The strap hooked to the firewall started smoking and since then nothing has worked right. I have been though EVERYTHING i can think of, checked things at least a dozen times, and the ECM is my last item before I take it in and say "fix it"

EDIT*
Found a handy link if you are thinking about getting another ECM.
https://www.thirdgen.org/ecminterchange
Lots of cars to choose from.

 

Had some time to do further diagnostics today and it does seem like my ECM could be to blame. I unplugged the ribbon cable from the AutoProm - which to my understanding would force the computer to run in limp home mode(correct me if I'm wrong) - and it ran exactly the same as it did with the AutoProm plugged in.. Since the last time my ECM went haywire was shortly after the engine swap when I forgot to connect the engine ground, I checked it. A bit loose, so maybe that has something to do with it.

Is there any way to more definitively test the ECM itself?

Does anyone know what part on the unit is damaged by a lack of proper grounding? If it's a few diodes, fuses, or caps I'll just replace them. I'm not a afraid of a little soldering. If a surface mount IC gets blown by some voltage spike, that's another story. But it leads me back to question 1.. I'd love to know how to test the unit itself.

Failing those two, what's the best place to purchase a 7730?

 

The only way to truely test an ECM is with a test bench.

Do you have a stock MEMCAL, or an EEPROM you can program a bin to, to verify that the ECM does or does not read from the EEPROM/AutoProm?

Usually when in LHM, there is no data sent through the ALDL, so it may be that something is not right with the AutoProm or the ECM's ability to read from it.

 

and with the 7730 being so well documented, you could just force factory test mode with a 3.9K resistor on the ALDL diagnostic terminal and monitor the outputs.

 

If I understand that mode correctly it just tests the hardware functionality, but not the ability to read from the PROM.

 

it needs to read the PROM to know what algorithm to execute.

 

@Six_Shooter: Yes, I do have a stock memcal and a box full of microcontrollers/eeproms/programmers/etc. I should point out that I removed the AutoProm variable by plugging the stock memcal in directly and the car seemed to run the same. I have not as yet tried to get ALDL data when running directly off the memcal, but I'll try that as well. Not that it would be definitive, since you mentioned the ECM doesn't transmit ALDL data in LHM anyway. These symptoms so far seem to add up to the AutoProm working correctly, but the ECM flipping out for some reason(probably intermittent ground from loose ground cable) and going into LHM.

@Saar: It seems I fail at Googling. When I searched for various iterations of "1227730 factory test" I get this very thread.

I guess I'll just pick up another 7730 from eBay(unless someone has a vendor recommendation) and drill a new self tapping screw into the firewall for a more solid ground from the passenger side head. I just dislike replacing parts to see if they were the problem. I'd rather understand exactly why things are broken.


Does anyone know why a bad ground tends to blow these ECMs up? Any way to protect them other than ensuring ground connections are solid?

 

Id say rockauto to buy a ECM, no core charge and great warranty.

1 day before my warranty expried, I was able to send it back for a new one.

[for me it was the canister purge valve causing me years of problems and fried ecms that some still worked half cooked, some didn't.]

 

Invoking the factory test code is a little more involved then just the 3.9K resistor. The ignition input to the ECM needs to be above 16 volts while the battery input is less then 10 volts.

To see if the ECM is running from the PROM do this: while observing the SES light do a key-on, engine-off. The SES should turn on briefly, blink off, then turn on solid. If it does the blink off, the ECM is running from the PROM.

Can repeat the test after a minimum of 10 seconds of key-off.

RBob.

 

@Gumby: RockAuto does indeed rock. Unfortunately it looks like they do have a core charge for the 7730s they're selling now. What was wrong with the purge valve? I assume they have a flyback diode across the coil to prevent voltage spikes on the line when the coil field collapses. Did that burn out or something?

@RBob:I'll double check when I get home, but I'm fairly certain that in key-on, engine-off configuration the SES light remains on continuously. No SES light when engine is running however.

 

only noticed it when I put my bare hand on the valve and got burnet cause it was shorting out inside hot. Which was frying a section of the ECM. which on some the ECM still worked, some didn't.

it was a nagging problem the entire life of the car with me, but original ECM lasted 6-7 years like that.


on top of there is no code for a bad CCP valve, it just fries the ECM as a code.

 

Grumby,
Can you tell me more about this CCP valve? Where it is located? I'm in the same boat as Nebulous as far as having bizare computer/communication issues.

 

I should point out that with a hardening of the ground strap connection and a new ECM, I'm getting both serial data and closed loop. Passing emissions on the other hand... That's a different matter all together.