Blogs Car Info Our Show Deals Mechanics Files Vehicle Donation

Major injector/ECU/fuel problems

1996 cutlass ciera, 2.2, GM PFI, 186K miles

I replaced all the fuel injectors as there was gas in the oil, (i had already replaced the FPR and checked everything else) and I got a great deal on 4 brand new ones.

I crank her a few times and she hydrolocks. I pull plug #4, and crank it again. Fuel shoots out in a geyser. Yep hydrolock. I pull the remaining plugs and crank her again, the remaining 3 cylinders were full of gas too.

I then pull the fuel pump relay, and clear out all the gas. (summarily making the garage a toxic area) and rent a noid light set to further diagnose whats going on…

I crank the engine, and the noid light refuses to flash, or even activate, regardless which harness its plugged into. Now I have to note I have the ICM unplugged due to something else, but I doubt that would keep the noid light from working?

I then check voltage at each of the harnesses, (key on) I read 14.5 volt constant on one side, 1.5 volt constant on other side. Each harness was the same. (key off) I read 12.4 volts constant and 0.15 volts on the other.

I then pulled the inj coil fuse (which I assume is the driver?) and I get key on or off 0.05 volts on one side and 0.00 on the other…

Is my driver or ecu bad? The old injectors were squirting fuel into the oil, or so I am sure as the fuel pressure when tested even with a new FPR leaked down in about 2 minutes. Now with healthy injectors the fuel is an ocean!

Did you try it with the ICM plugged in?

You might want to pull the fuel pump fuse, pull the plugs and turn the engine over until it stops spraying fuel out of the plug holes to unlock the hydrolock. BUT, there is a reason why oil and gasoline mixes inside the cylinder(s) and needs to be fixed or you could experience the same thing over again. In the end you could have an even more serious issue…

A borehole camera would probably be more useful then the noid light. A 12V bulb attached to two wires will do the same job.

Sounds like a suicide mission to me, doing all this experimenting while fuel fumes are all over the place. A little spark and you become a statistic.

No, not yet but today I will.

As I said above, I’ve already done that.

The fumes have already dissipated, Im more concerned about the readings at the injector harnesses im seeing. And the fact that im not seeing any noid light activity while it appears my injectors are getting a constant voltage that started all this

noid light activity while it appears my injectors are getting a constant voltage that started all this

Sounds like the ICM has failed. The injectors seem to be commanded a solid battery signal (or full open injector)) rather than the pulse-width modulated signal they are supposed to get. Rather than using a noid light, you might try a DVM with a PWM signal input. Cranking the engine should not show 100% PWM.

PO said the ICM was unplugged.

Missed that kurt… Serious bad wiring problem!

digital volt meter… okay, but whats the PWM acronym?

icm plugged in, no difference except diff voltages at injector harnesses

(key on) is now - 0.81 constant and 0.00 on other side. Its the same as (key off) noid lights still refuse to light and spare injector when plugged in doesn’t even click.

This is weird as before i discovered the fuel in the oil the car ran great as always… Im starting to suspect a bad driver more and more

Pulse Width Modulation. Switching ON to OFF Higher On to Off gives a higher PWM reading. 100% is on all the time 50% is half on half off.

Maybe I missed it, did you try to start the car after you unlocked the hydrolock?

No way to do that, every time I plug the fuel pump relay in, all 4 cylinders fill up with fuel. That’s why ive been trying to use the noid light, its like all the injectors are being held open or all 4 injectors I bought new are bad, or I damaged all 4 putting them in, which I doubt lol.

That is like looking for a zebra when a horse is most likely. Now way would ALL be bad and no way would you screw up all 4 installing them. Its just too unlikely.

All 4 injectors should not open when you insert the relay, ever. You have a wiring problem.

1 Like

Yeah, agree completely. The odds of those last two are nigh on astronomical. The wiring has no damage, and wouldn’t just “be bad” after sitting off all day, after running for years unscathed. That’s why Im thinking bad driver in the ECU. Ive heard 96 ecu’s are notorious for driver failure. :frowning:


I unplugged 2 of the injectors (1 and 4) and reconnected the fuel pump fuse, and cranked the engine. Cylinders 2 and 3 shot out fuel.

So, the injectors are shorted to ground when plugged in. (explains the battery being drained every time I go to work on the car)

Now, I checked the wiring diagram, and there are 4 driver circuits on two different ECM connectors. So, the chances of all 4 wires being shorted to the same/a power source outside the ECM is near zero. Now, I did install a new ICM when this all happened, but after, I reinstalled the old one. So both ICMs were used with this diagnosis. But, remember the ICM was unplugged for the first half.

So its one of these possibilities:

  1. All 4 new injectors are bad
  2. All 4 injectors, each with their own source wire, are shorted in the harness
  3. All 4 injectors were installed incorrectly
  4. Two different ICMs are shorting the ground
  5. The ECU injector control is shorted
  6. There is a short in the ICM circuit wiring, even unplugged

Which one doesn’t seem insane?

From what I can tell in the wiring diagrams, the ICM has nothing to do with the injection system. It simply receives signals from the PCM to fire the coils. The PCM receives ignition signals from the CKP and CMP. This seems to be the case for the 1996 Ciera anyway, assuming I’m looking at the right diagrams. The ICM in other 2.2L models received the CKP &/or CMP signals.

The fuel injectors should not be receiving any power with the key off, or with the key on engine off (except for 2 seconds during pump priming). The injector positive voltage is supplied along with the fuel pump positive. This power comes either through the fuel pump relay (triggered by the PCM during key on prime or engine cranking/running), or the fuel pump oil pressure switch (which activates when oil pressure is above 4psi as a backup to the relay). If either one of those is stuck closed, you will get constant power to the pump and injectors.

Even with constant voltage at the injectors, the only way they can be triggered is through their grounds, which is normally handled by the PCM.

There could be two problems here. Injector/fuel pump constant voltage and injector shorts to ground.

I poured over the wiring this morning and you’re right. It must be because its a leap OBD year the power comes from the fuel pump circuit, and not the ICM. It turns out the DC low range setting on my meter is bad, (I got 2.8 volts from a wrench) and when I used a higher range to check for voltage at the harnesses, I got zero voltage. Didn’t see that coming. So that means the injectors are shorted to ground, and I doubt all 4 signal wires are grounded outside the ECU.

Im still going to replace the meter as now I can’t exactly trust it. I’ll then retest, but still… Am I correct in thinking even more now the ECU’s driver is shorted?

You should check all 4 anyway. Unplug the injector connectors and PCM connectors 1 & 2 (1 is black, 2 is clear). Then check for continuity to ground on all four injector ground wires. If any one shows continuity to ground, there is a short. If they check okay, then reconnect the PCM and check continuity to ground at injector connector ground pins. If there is any continuity there, then you have PCM problems

BTW, there is a 5-pin inline connector between the injector connectors and the PCM connectors (engine harness to fuel injector harness connector). If there any shorts in the ground wiring, you should check that connector to be sure nothing untoward has happened to it.

Couldnt do much today, but i did jot down all the pins. I also disconected the ecu, and tested the injector harness for continuity. I didnt get a short to the body from the harness, but i did to the engine? The valve cover, the water neck, etc.

I dont recall thats normal? It should’nt be shorted to the engine, should it?

If an injector ground wire is disconnected at each end, there should be no continuity from the wire to any ground, whether to body, chassis, or engine, which are all hooked together as a common ground. If you didn’t get a short to the body, but did to the engine, perhaps your test probe wasn’t making a good connection on a painted body surface.

Did you check all 4 injector ground wires? If any or all are actually shorted to ground, you’ll have to trace the wires through the harness(es) to find where the problem is. Perhaps chafing somewhere. If you disconnect the injector harness from the engine harness (the 5-pin connector I mentioned above) and check continuity again, you can narrow it down to which harness has the problem.

I used the inner fender, the large brace in the engine bay, and a strut stud, still possible though I suppose. I did’nt have time to test all 4 leads, but I will this morning.

You wouldnt happen to know where the two harnesses meet, would you?