Page updated on 06-18-2018

How often do ECMs go bad?

01-27-2006, 01:05 PM
I have a '94 Metro 1.0L 5-sp with 60,000 miles. The car has been amazing, needing only regular maintenance, four new tires and an alternator belt these past six years I've owned it.

About two years ago, it began idling roughly when you first start it (in any weather), have to hold the pedal part-way down those first two minutes, after that warm-up time it is fine. I took it to a shop, they "cleaned" the EGR-something but couldn't fix the problem. No biggie, I can live with that.

This past month, there is a new intermittent problem. The car will sometimes start and run fine. When it doesnít start, there is a rapid LOUD clicking noise from the fuel pump relay. I replaced the relay, no change, the new relay clicks too. I bypassed the relay with a heavy gauge jumper wire.

Like I said, the problem is intermittent. Sometimes it will start and run fine. Now that the fuel pump relay has been bypassed (and I can hear without that loud clicking) Ė when it wonít start, Iíve noticed that the fuel pump doesnít turn on (make a humming noise) when the ignition key is turned on. Obviously, sometimes the fuel pump is getting the power it needs and will start, sometimes not (replace the relay in this condition and the relay clicks rapidly and loudly).

Iíve taken it to a shop, but of course its working fine when I do. They speculate that it could be the ECM, and suggest that replacing it is more that the car is worth. A neighborly backyard mechanic also thinks itís the ECM gone amuck.

Someone on the forum mentioned that removing an ECM is a BEAR. Iím dreading taking the ECM out to send in for a make-over (I canít even see it under the driverís dash), and wanted to ask for more opinions. Does this diagnosis seem reasonable? How often do ECMs go bonkers on Metros?


01-27-2006, 01:43 PM

"Like I said, the problem is intermittent. Sometimes it will start and run fine. Now that the fuel pump relay has been bypassed (and I can hear without that loud clicking) – when it won’t start, I’ve noticed that the fuel pump doesn’t turn on (make a humming noise) when the ignition key is turned on. Obviously, sometimes the fuel pump is getting the power it needs and will start, sometimes not (replace the relay in this condition and the relay clicks rapidly and loudly)."

The only thing your ecm does regarding the fuel pump is to ground the relay. Therefore if the fuel pump is still not working while you have the relay bypassed then you can pretty much rule out the ECM. Do note there is an additional realy that might not be working. Also for $1 you can get a test light, use it to check for power coming into your relays - if you have power at your relays then check for good power and ground at your fuel pump. If you have good power and ground there then you know your pump is toast - if you don't have good power/ground there then you will just have to repair the wiring. (

Thanks to DieInterim, this and other valueable Geo info can be found at .

01-27-2006, 06:30 PM
I think bypassing your relay for anything other than a quick test is not safe.If in the unfortunate event you were in a collision you would still have fuel pressure and possible broken fuel lines.

01-27-2006, 08:05 PM
I am the person who said removing the ECM from a '93 Geo Metro was not

I did this on a '93 (same year as my Metro) in a Pull-n-Save yard and the car
was up on wheels welded to the frame - so the vehicle was up higher than is

Even so, it is hard to get at.

On that car, the ECM was up under the driver's side dashboard to the left of the
steering wheel bolted to essentially the inside of the wheel well with three
Phillips-Hex head bolts.

The real problem was twofold.

1. All the wiring from the Fuse Box is right there in your way.
.....I ripped that out in the Wrecking yard - but you wouldn't want to do that!

2. The three Phillips-Hex head bolts were hard to get turning with just screw
driver! Mine was large (!) but couldn't loosen them. Lucky that a nice old
fellow came by and loaned me his 10 mm socket wrench with a long extension
rod. THEN - it was EASY.... Still hard to get at the Phillips-Hex head bolts, but it came out then...

Go to my thread to see what it looks like...

I think all my problems stem from a Fuel Pump going TU.

I have a thread in this forum on that, also.


01-27-2006, 08:09 PM
check the ground connections on the back of the intake and behind the overflow tank, if these are loose or dirty conection will cause same problem

01-28-2006, 12:15 PM
Crvwrr69 is right in 90% of the cases is the fault of the ground connection, without a good one you will start to have problems, ECM s usualy outlast the life of the car and engine unless you start poking with the screwdriver in it or flood the car.

01-28-2006, 11:06 PM
You have me wondering now if that grounding thing isn't the reason my wife's
Geo Tracker CEL is on always, but shows no code when the socket is jumped
as per Chilton Manual instructions....!

I know this is the Metro Forum, but does anyone know where these gounding
wires would be on the 1993 Geo Tracker 1.6 L jobbie?

What the Hell, it's worth a try! I was going to order a used ECM and try it
out. Maybe it's just a bad ground!


01-29-2006, 12:01 AM
Once you have exhausted all of the recommended actions above I would do the following -

It truly sounds as if there is a bad triac or a cracked solder joint in your ECU. Although it is possible for the logic to be messed up within the ECU it is more likely that you need to replace a triac.

What is a triac? A triac is a three terminal semiconductor for controlling current, or in common terms, a relay switch.
I would check the solder joints on all of the triacs and all connector posts too. If solder joints look fine then we determine which triac controls the fuel pump and replace it. We are talking about a couple of dollars for the parts and a 50% chance this solves your problems.

I am willing to help you if you so desire.


I'm a bit confused by this. As far as I'm aware a TRIAC and in general SCRs operate by controlling the duty cycle on a sinusoidal type waveform (ac). Since the supply in the car is 12V DC it is not clear to me how it works. Also on the pic Doctorbill posted I can only see only one device which looks like a power semiconductor device and I suspect that this a voltage regulator.

The rest of the stuff as far as I can see are the micro, a bunch of logic ICs and a bunch of transistors, resistors and caps as a diode as I recall.

Where are the TRIACs and how many are there?

01-29-2006, 01:48 PM
Hmmm.that's an interesting design - I would have thought locating the power device on the external device would have made more sense as less energy would be required to send the control signal rather than the main signal down a wire - but of course that would be more sucecptible to interference. I guess they are just using the control sgnal form the logic to switch in the main signal to the pump via thepower device which makes sense.

Good point about opto-isloation and yes the same thing can be done in a variety of ways. Most consumer electronics tend to have designs where cost savings are a key factor and many circuits are not necessarily the best design or the most robust. In thise case thought it is rather excessive to want $1500 for something not worth more than about $50 in parts.

01-31-2006, 04:23 PM
Thanks for the responses, I really really appreciate them.

The problem hasn’t resurfaced yet, statistically, it won’t be more than one week before it acts up again.

I have checked all the grounds I could find, removed every bolt and cleaned every contact with some sandpaper and WD-40, none of the contacts looked bad, but who knows.

Thanks for the chart BeoWolfe, when the problem returns I will check the other relay and troubleshoot with the chart.

I fear that Blake’s suggestion might be the real culprit, either a failing triac in the ECM or some other internal ECM problem. It’s curious how the fuel pump relay clicks so rapidly when the problem is acting up, so rapidly it sounds like a buzzing (10 or 20 times per second). I took the plastic cover off the relay last time, and could see that magnetic clapper slamming back and forth as hard and fast as it could. The sound was really loud. If the ECM controls the relay by grounding it, there is something curious going on there to cause it to trip on and off so rapidly. I will measure the voltage at the relay the next time to see if it is constant, high or low. Hopefully cleaning some grounds have fixed the problem.

I hoped that the unique relay buzzing might have sounded familiar to someone, but this might be an oddly rare problem to solve.

Thanks again everyone.


02-01-2006, 12:17 PM
The control signal may be oscillating. Without seeing the circuit it is hard to say but if the micro drives the power device via a line driver that incorporates other components like a low pass filter and these degenerate the signal can start doing funny things like developing + feedback and so creating an oscillator.

The best way to see if this is happening is using an oscilloscope to trace the control signal all the way from the micro to the final signal going to the pump. A meter can be helpful but an oscillation will show up as a reduced voltage. Nothing like actually seeing the waveform.

You may want to check out a online auction like ebay for a used scope - very useful tool for troubleshooting electronics.

Add your comment to this topic!