Car dieing on start

[Dickhead]

I have an Astra 2.2L MK4 Convertible, recently a problem occurred where I'd turn the engine on, and it would just die at idle. I got standard somewhere, and an ECU reset fixed the problem (took the battery out for half an hour). Anyway, I managed to get home and the same problem occurred, an ECU reset fixed it again. I just got diagnostics done at the local Holden dealer and the tool couldn't communicate with my car, now they're saying it's either the ECU or the immobilizer. Any other ideas on what it could be? I'd be charged $1450 to replace the ECU, and $450 for a new immobilizer, and they're not even sure if those are the problems. Oh yeah, they reset the immobilizer and now the car is running, but there's a chance the same problem would start happening again.

[Vargus]

I'd be charged $1450 to replace the ECU

Wtf? I can buy ECM's for $200, aftermarket, moddable ones for $600.

You need to go to a dealership or a garage that isn't going to fuck around with you.

What year model is it? My local Autozone had a code scanner that could read my 03 VUE's codes. No reason why any competant garage wouldn't have a scanner that could read yours.

[Professor_Skullsworth]

^ but if there is a bus or module down, no scan tool could communicate with it.

From what you stated(op), it sounds like it could be an immobilizer problem. If you had a chrysler/dodge/jeep i could tell you a lot more,...but im not sure how the gm/holden style immobilizers work exactly. good luck

[Dickhead]

It's a 2003. Oh yeah, I just remembered he mentioned that the data serial bus could be. Argh, this sucks. I just don't wanna get ripped off. If it's simple shit I'd attempt to fix it my self.

[Dickhead]

The problem randomly occurs, so I managed to drive it to another garage the other day, a few days later he called up and said he fixed it, he said a ground wire to the ecu was fucked, I managed to drive it back and when I got home it wouldn't turn back on. A few hours later, I turned it on, took it for a spin, came back, it turned on once and died, but the weird thing is after the engine cut off, there was a long whirling sound coming from the engine bay, after that the car wouldn't turn back on, the engine would just struggle. Also, accelerating on idle won't keep it alive, it will still die, the engine light also flashes.

Any ideas?

[MasterYoda210]

If it is the ECU, they are pretty easy to replace by yourself from my experience.

Just go to a scrapyard, find one there (make sure it is exactly the same though, look up on t'internet) then from what I've experienced, there is just one massive multiplug plugged into the ECU, so unplug that, remove old ECU, insert new ECU, reattach plug.

If your is at all like mine and changes ignition timings and the like based around your driving, it will be a bit crappy for the first few hundred miles until it relearns your driving style. Idle should just be pretty standard though, so that should be fine.

If it's the immobilizer, I have no idea.

EDIT: Not keeping alive when you rev it implies to me it is some kind of ignition problem also. Could be your timings? Also, not to scaremonger, but potentially the whirring for ages when it died could be the snapped cam-belt or pulley whirring round with no resistance. But definitely get your timing checked. Does it smell of petrol or anything? Because my old 214 threw its timing, was a bastard to start and keep running and when it did start it reeked of petrol because it was making far too rich a mixture, and it would struggle under load.
And finally, I don't know how GM/Holden/Opel/Vauxhall engines control idle, but my car has a stepper motor on the inlet, and that whirrs when the car stops to re-set itself for when the car starts, so maybe the ECU was trying to reset something like that, but failed?

But that's the best I can offer, I'm no mechanic, just going of my own experiences.

[truckfixr]

Don't buy any parts until you know for certain what the root cause is. You can easily spend a shitload of money without fixing the problem.

Ok now, give us a little more information. Does the problem occur on the initial startup after a cold-soak? Does it occur only after the engine reaches operating temperature? Will it restart if you leave it sit for an hour or so (without disconnecting the battery)?

I could be (and probably am) way off base here, but I have a sneaking suspicion that you may be losing fuel pressure. I recommend that you verify your fuel pressure when the problem occurs.

[Rocko]

One thing you can do yourself (if you've got a voltmeter) is to confirm good power and ground on all the modules. 2.2 cars should have an ignition module on the top of the engine, a body module buried somewhere (probably not relevant to your problem), and a Powertrain/Engine module in the engine bay. If the power or ground is bad, it typically causes the engine to die right after you start it.

Just remember that when you test those terminals on the modules, be extremely careful, they're pretty fragile and you can fuck your connections by jamming crap into them. It varies by car, but the power wires are usually pink and the grounds are usually black. Power should be very close to whatever the battery voltage is.

When it dies, does it sputter at all or just shut off like you turned the key off?

[Dickhead]

Fuck, of all things it was the key/remote. The battery was flat, so it wasn't picked up by the immobilizer properly. I spent a shit load at the mechanics, when it was just a $4 battery that needed replacement.

EDIT 2: Well, it's fucked again. The problem occurs after a long drive, if it sits for a while it will turn on fine and drive fine. It doesn't smell of petrol, what do you mean by timings? It dies in a peaceful way, the revs will just drop, kind of like when you switch the engine off. Oh and the engine light flashes when the problem occurs.

[MasterYoda210]

It doesn't smell of petrol, what do you mean by timings? It dies in a peaceful way, the revs will just drop, kind of like when you switch the engine off. Oh and the engine light flashes when the problem occurs.

I don't know how well you know how an engine works, but basically, the camshaft rotates in time with the crankshaft, so it starts to let in air and petrol at the top of the first stroke, fires the spark plug right at the top of the third stroke at maximum compression, etc.

If this precise timing is even a little bit off, like it opens valves too early or late, or ignites the mixture at the wrong time, it will make the engine run badly.

As I say, all I go off is my own experiences, not any formal mechanical training, but when my old 214 started dying on start up, or when it was idling at lights, or if i tried to give it the slightest bit of power, it turns out the timing was retarded, that is the camshafts were opening and closing things after they should have been.

But again from experience, when timing is at fault the engine will struggle a lot, and not die 'in a peaceful way' as you say. I'm now led to agree with truckfixr.

EDIT: Thinking about it, how old is a mark 4? I assume Australia simply uses the same 2.2 engine as Europe, and I've heard things from the boy racers that there are common faults with larger engined 2001+ Astra's in regards to stalling. I'm almost sure (I had had about a gram of MDMA at the time, so my mind was racing) he said it was a fuel pressure regulator.

[ComradeAsh]

Why would the actual key have a battery in it?

[Rocko]

Why would the actual key have a battery in it?

Because remote controls don't work by magic.

[Dickhead]

It's a 2003 build, with a 2.2 Ecotec. If it was any of those problems, wouldn't the computers still be able to communicate with the ECU? A number of different computers have been tried.

[MunkeyQ]

Why would the actual key have a battery in it?
They always have a battery inside to power the RF, or sometime IR, transmitter. Normally a 3 volt lithium CR2302.

[Rocko]

It's a 2003 build, with a 2.2 Ecotec. If it was any of those problems, wouldn't the computers still be able to communicate with the ECU? A number of different computers have been tried.

If the ECU doesn't have power, you won't be able to communicate with it.

[Dickhead]

If the ECU doesn't have power, you won't be able to communicate with it.

The units are receiving power from the terminal, it just won't communicate.

[ComradeAsh]

They always have a battery inside to power the RF, or sometime IR, transmitter. Normally a 3 volt lithium CR2302.

The fob does, but I didn't think the actual key did.

[Rocko]

The units are receiving power from the terminal, it just won't communicate.

Your scan tool gets power from the battery, not from the module. If the module doesn't have power the scan tool will not be able to communicate with it, but it will be able to turn on.

[Irukanji]

The fob does, but I didn't think the actual key did.

Most of the newer holdens have the immobilizer built around the key. Maybe that's what they meant?

[Dickhead]

Oh okay, I suggested to the mechanic that the ECU might not be receiving power, and he said that wasn't the case because the scanner was receiving power. Is the ECU probably dead?

[Hyper-dimension]

I had this happen to me on my old Nissan pickup. It was simply a half-loose connection at the alternator. Check these.

Also, have you replaced your spark plugs, wires, rotor, and distributor cap lately? Could be a distributor-based issue.

[truckfixr]

Oh okay, I suggested to the mechanic that the ECU might not be receiving power, and he said that wasn't the case because the scanner was receiving power. Is the ECU probably dead?

As Rocko already stated, the scan tool can getting power does not guarantee that the ECM is powering up. The power terminal at the datalink connector is usually supplied through a different fuse than the ECM .

Even if both the ECM and the scanner are powering up, there are other issues which can prevent communication between them. Before condemning the ECM, you need to verify the integrity of the data buss circuits. The circuits need to be checked not only for continuity, but for shorts to power, ground, and to each other. Also, verify the integrity of all power and ground feeds to the ECM. All power and ground circuits which feed the ECM should be tested with a test lamp, not a digital voltmeter(A digital meter doesn't load the circuit). DO NOT use a test lamp on any other ECM circuit (while the ECM is connected). Doing so can damage your ECM.

[Rocko]

As Rocko already stated, the scan tool can getting power does not guarantee that the ECM is powering up. The power terminal at the datalink connector is usually supplied through a different fuse than the ECM .

Even if both the ECM and the scanner are powering up, there are other issues which can prevent communication between them. Before condemning the ECM, you need to verify the integrity of the data buss circuits. The circuits need to be checked not only for continuity, but for shorts to power, ground, and to each other. Also, verify the integrity of all power and ground feeds to the ECM. All power and ground circuits which feed the ECM should be tested with a test lamp, not a digital voltmeter(A digital meter doesn't load the circuit). DO NOT use a test lamp on any other ECM circuit (while the ECM is connected). Doing so can damage your ECM.

Fuckin this. ECMs don't go bad very often, and a lot of technicians will condemn them, replace them, and then get screwed when it hasn't fixed the problem. Rule out everything but the ECM before you commit to that expense.

Use the voltmeter to check for continuity and shorts in the communication wires, and use a test lamp for the power wires. Here's a tip: connect the voltmeter so you can see how much power is available, then also connect the test light. If the voltage drops more than 2 tenths of a volt (or whatever the light is rated to) then you've got another load on the circuit, which means resistance has developed somewhere. Modules need full battery voltage to function properly, so a relatively small amount of resistance in the power circuit can keep them from operating.

A lot of this is tough to manage without wiring diagrams, though, see if you can find a manual at the library or something that will have detailed diagrams of the power circuits and scan tool communication wires.

[Dr rocker]

In theUK, the 2.2 has a timing chain. It is renowned for having a poor oil feed and failing. I would check out that it is not worn / stretching with the heat of a long run, causing a timing issue. I would then look at oil starvation to the head causing a cam / valve problem, then start looking at ecu problem, which could be coincidental or a conection fault.

[Dickhead]

I took it to another Holden dealer, this one wasn't dodgy. They managed to communicate with the ECU and figured out the problem was with the transponder code, they changed that and cleared the codes, it's been good so far.

Thanks for all the help, I avoided an ECU change as much as possible even though it was recommended by quite a few places. If it happens again, I'm gonna be fucked.