legacycentral bbs

Car seems to be overfueling and won't run. Actually it will barely starts before it coughs and sputters to death. Have been testing different things and then I saw that my injectors always have 12V across them- not just the + voltage they are supplied with, but the across them!

What in the world is making the computer ground the injector signal at all times? It doesn't even matter if the key is in the ignition...

Any help would be great. I am going to start by swapping the ECU since I have a complete parts car but any other tips as to what sensor could cause this to happen would be great.

Car doesn't read any error codes if that helps.

-Mark

Uhh... That's weird.

Does your battery drain when the car's off too, then?

Is your ECU powered when the ignition is off? One way to check is to connect the test mode connectors and see if the solenoids and relays are cycling.

What happens if you unplug the ECU's 4 connectors?

What happened just before this problem started happening?

That's a good point, my battery was drained when I had left it sitting for 6 weeks. Problem started just before I went out of town for the summer.

ECU doesn't appear to be powered all the time since test mode only responds when the ignition is in on position.

If I disconect the 4 connectors injectors have 12v+ but aren't grounded. They are the way they should be.

Swapped the ECU from a computer in a car I ran last year - no change to problem.

Before the problem started car ran fine. Just finished a FMIC setup with a 60 trim T3 turbo.

After setup car ran well till I spun off the compressor wheel bolt on the turbo. (Due to blow off valve not "blowing off") I took the turbo off and fixed the problem. Later that day I blew an air hose during a high boost run just as I let off - seemed my blow off still wasn't working just right??. I limped the car home where I replaced the blown hose.

Next day I attempted to drive the car to work. At first car felt fine, boosted well and felt normal. About 10 min into the drive if I pulled up to a stop light the car would die on me, but, would start right back up and run ok - but a bit worse each time this scenario happened. Seemed like if I gave it too much gas it would stumble.

Just made it to work when the car died as I pulled into my parking spot. After work things just got worse, car would hardly run and I had to try and keep it at 3500 rpms or so or the car would die. Eventually it simply wouldn't run anymore.

Checked my plugs after trying to start and they were rather wet. Checked for spark and spark seems good.

Anyways it is all very wierd to me....

Huh. Weird.

So both your stock ECU and the spare ECU held the injectors open, even when the car was turned off?

And disconnecting the ECU makes that stop happening, indicating that the wires aren't grounding somewhere on their way to the ECU.

I have no idea, man.

Does the Check Engine light come on with all the other idiot lights when you turn the ignition on before trying to start the car?

Yes, both computers held the injectors open. I could swap the computer in my brothers car out to I suppose, I know for a fact it works - but so did my spare I was driving that car not to long ago.

Yep, if I disconnect the harness to the computer the injectors close.

Yes the check engine light does come on when key is in the on position and while cranking. It goes out if the engine catches for a few seconds.

Is the battery in good shape right now? It's charged up and all? I guess if you measure 12 volts across anything on the car it's gotta be, but I thought I'd ask.

What happens if you disconnect just the 26-pin connector of the ECU?

What about just the 22-pin connector?

If you have a parts car, swap over the ignitor. Maybe one of the power transistors has failed.

That's whats in the ignitor, power transistors, right?

Yes, the ignitor is basically two big power transistors, but it's just responsible for ignition, not the injectors.

The injectors are driven directly by power transistors inside the ECU. That's why I can't understand why two ECUs would do the same thing.

ah, my bad. continue....

What if there is some type of problem with the harness connections for the ECU...perhaps the ECU is fine but your voltage from the ECU is being bled off or fed to ground somewhere along the way ? If I udersterstand correctly, 12V is supplied to the injectors from elsewhere, and the ECU gives somewhere between 0(ground) and 12V to the other leg to make 12V to 0 variable across the injectors.

I'd check the output from the ECU using the straight pin method once you identify the correct wire(s), then begin to trace toward the injectors to see if the voltage between your traced wire and ground suddenly changes. Perhaps the connector has failed and you'll find the problem right there.

You've almost got it, professor. The injectors are fed constant +12v on one side, and the ECU has control over the other side. It either leaves them floating (leaving the injector closed) or pulls them to ground (opening the injector).

The thing that puzzles me is that he says that without the ECU connected the injectors aren't open. So wouldn't that mean that the grounding must be happening inside the ECU?

Yes, that is correct, if I disconnect the harness the injector is no longer grounded. I am at school in a FEA lab right now but when I get home I'll test the individual connectors that you mentioned.

I suppose my answer was quite lame, however if the connector is corrupted, the wire(s) to the injectors could be finding steady ground through another (correctly) grounded wire coming out of the ECU harness, as such connecting to ground on the other side of the ECU, so pulling the ECU would cut the ground. Using this logic a wiring failure seems less likely, and a connector problem, more likely. I'd take a hard look at the pin diagrams and see if there is anything looking funky in the area.

Is there any sensor input malfunction that would cause the ECU to calculate full-open on the injectors ? I would think not, that this would be detected as out-of-bounds data and trigger codes instead, as such output would toast the injector solenoids in short order.

come to think of it, is it really possible to determine whether you have a constant 12V when the ground is cycling rapidly ? how did you determine this ? I would think only an o-scope could do that correctly, and lessor meters would be driven crazy and be unable to give a real reading

Umm... this was determined because I have 12v across the injectors at all times car running or not, key in the ignition or out... My car doesn't even run right now.... my poor suby...

It's a good question though -- did you just use a digital multimeter? Because it is possible that the power is cycling rapidly rather than just being constantly on.

You're leaving your battery unhooked now when you're not trying to start the car, right?

Yes, using a digital multimeter. The battery reads 12.41 and all injectors read 12.09. Unfortunately I can't find an old analog one around that works.

Pulled connections on ecu and found that the 26 pin one stopped the grounding but the 22 pin one didn't change anything.

What else, oh good suggestion on unhooking battery.

I am going over to my brothers now to try my old ecu in his car. I'll see what happens.

-Mark

Well pulled my ecu, plugged it my brothers car and all works well on his car? Therefore the ecu is good for certain

Here is something interesting I have found. With the 22 pin connector unplugged from the ecu I have 11.56v across the injector, with the 22 pin connector plugged in I have 12.07v. From the positive side on the injector to the battery I have 12.40v anytime.

Perhaps with an anolog meter I would see that the computer is pulsing the injector?

Whatever the problem is it is from some signal in the 26 pin connector that is for sure.

-Mark

Well, the 26-pin connector is the one that includes the injector wires.

You should measure 12 volts from the positive side of the injector to the negative terminal of the battery at all times. The injectors are always powered.

Now here's something kind of disturbing: the ECU's power supply (both main and backup) comes through the 22-pin connector. Without that connector plugged in, the ECU can't possibly power up.

Dude, are you pulling our collective legs? I don't expect so but I have to ask.

What I meant was that there must be some signal in the 26 pin connector that is making the injectors come on - besides the fact that it grounds the injector.

The computer was tested good as I just put it in my brothers car and tried it out.

When I am measuring voltage at the injector yes of course it should always have 12v to the battery ground but not across it.

I was just commenting on the fact that there wasn't quite as much voltage across the injector as there was from the positive side of the injector to ground. Perhaps the digital multimeter gives a lower reading if signal is pulsing or perhaps it's just the loss through the wires. There was however about .2v difference across the injector when I plugged in the 22 pin connecter as well - I checked that a couple of times.

What if I get out the meter again and go over each signal from the 26 pin connector, one of those must some how be bad. Any info on what I should be looking for?

Oh, sorry forgot to respond to your last question. No, I am not pulling your collective legs, I really wish I was and that I could go out and drive my car but unfortunately I can't. I understand why you had to ask though...

Very confused that's all.

Maybe unplug the ECU's harness connectors and check for continuity between one of the injector pins and each of the ground pins?

http://www.graphics.cornell.edu/~v/ecupins/

Ok, here is the latest.

With the ecu totally unplugged measuring with a multimeter I get the following readings.

Injector 1, 2, and 3 to either injector ground - 0.3M ohms.
Injector 4 to either ground - open.

Result - all injector plugins have 12v and aren't grounded (as to be expected as they aren't plugged into the computer)



With the ecu have only the 26 pin connecter plugged in.

Injector 1, 2, 3, and 4 to either injector ground have 168k ohms.

Result all injectors plugins show 12v across them.


I have no idea if that makes since.

This reminds me of the type of problem you see with a weak connection...you read 12V but it doesn't do anything, won't carry enough current to activate anything or light a light. That can be a problem with multi-meters, they can show the voltage but if you applied a load, it would vanish. If I were you I would rig up an LED that can stand 12V with test leads. You should be able to see it pulse visibly when the injector would be pulsing. A regular bulb is too slow to glow so use the LED.


i think there is a chance that nothing is really wrong here, just confusing meter readings, and the real start problem lies elsewhere.

You could be right, I'll certainly check it out. In any case it just seemed to make sence since my car seems to be seriously over-fueling.

I'll let you know shortly what I find.