A car problem...

[MSTRS:Me]

I tried an Aussie Holden site...what a prick of a 'forum' to try and post in, so easier to ask here...
We have a VS Commodore with the V6 Ecotech engine. Car has been fine, but the other morning, after sitting all night, when the key was turned the radiator fan came on and the engine would turn over but not start. Some investigation suggested that the injectors were not firing. Spark at plugs though. Disconnect battery, remove computer and check for damage inside. Nothing. Reconnect battery, still the same - turn over, no fire. Probably the CAS? Yep, says an auto electrician. Hmmm...easier to try another computer (if I can borrow one) so disconnect the battery again, pull out the existing comp and look for another. 2 days - can't find one. Start thinking about that fan going when cold. Thinking I'd found the temp sensor I pull off the wire (was actually the gauge feed). Reconnect computer and battery and turn on the key. Good, no fan. Engine starts just fine as always. So reconnected it thinking that all would die. Then I discover the temp sensor not what I'd thought. And the engine starts everytime. All that has changed is the battery was disconnected for 24hours or more.
What is going on?

[crazefox]

exposed wire maybe. Or faulty earth on temp sender?

[cowpoos]

I tried an Aussie Holden site...what a prick of a 'forum' to try and post in, so easier to ask here...
We have a VS Commodore with the V6 Ecotech engine. Car has been fine, but the other morning, after sitting all night, when the key was turned the radiator fan came on and the engine would turn over but not start. Some investigation suggested that the injectors were not firing. Spark at plugs though. Disconnect battery, remove computer and check for damage inside. Nothing. Reconnect battery, still the same - turn over, no fire. Probably the CAS? Yep, says an auto electrician. Hmmm...easier to try another computer (if I can borrow one) so disconnect the battery again, pull out the existing comp and look for another. 2 days - can't find one. Start thinking about that fan going when cold. Thinking I'd found the temp sensor I pull off the wire (was actually the gauge feed). Reconnect computer and battery and turn on the key. Good, no fan. Engine starts just fine as always. So reconnected it thinking that all would die. Then I discover the temp sensor not what I'd thought. And the engine starts everytime. All that has changed is the battery was disconnected for 24hours or more.
What is going on?

Hmmm...don't know.

But...if I were you..I would PM hdtboy.aka gav...used to work on holdens for a holden franchise..and makes alot of sense till you buy him a few beers. [he's a mechanic]

[yungatart]

Thanks Poos...blardy car!!!!

[Quasievil]

Hmmm...don't know.

Now thats a First people !!

[Delerium]

disconnecting power may have reset the cars ecu and sorted the problem

[MSTRS:Me]

exposed wire maybe. Or faulty earth on temp sender?

Perhaps. But the only wire I touched was one to a capillary? tube that feeds to the temperature gauge in the dash. The sensor that talks to the computer appears to be on the valley, behind the thermostat.

disconnecting power may have reset the cars ecu and sorted the problem

This has also been suggested by another auto sparky. I can't understand why 10 minutes disconnected or 24+hrs disconnected would make a difference, but he assures me that there is a time limit memory in the ecu. ie leave the power off for long enough and it will clear itself.

Trouble with these theories is that I'm not left with answers and am hesitant to trust the car will behave.

[Static]

basically the engine ecu would have had a fault code pop up that it though was worthy of not letting the car start and stores it untill it is taken to a dealership or the like that fix the problem and clear the code from the ecu but disconnecting the battery for long enough in theory should clear any fault codes stored. well this is how audi's work anyways, id suggest take it to a holden dealer and get them to check for any fault codes?

[cs363]

disconnecting power may have reset the cars ecu and sorted the problem

I'd say this is most likely and from the symptoms described the original fault was probably from a faulty sensor.
If you have a problem again check the trouble codes on the ECU, that will pinpoint the problem. Heres a list of codes:http://www.aussieholden.com/forum2/showthread.php?t=239

How to read the codes here: http://www.techedge.com.au/vehicle/aldl160/vn_aldl.htm (I believe the VS uses the same 12 pin connector for the Diagnostic port, if it's different from the one in this link don't use these instructions!)

[1 Free Man]

say no more my son You have a Holden!! there's your answer on a nut shell.
Go trade it on a FORD.

[no-coast-punk]

Sounds like an immobilizer problem. Those keys (after a certain year, which I'm assuming yours is) have a radio transponder in them that talks to the car. The keys are also uniquely coded to a particular ECU. Which is why swapping another computer won't work.

GM's rolling code algorithm SUCKS and they sometimes lose synch. Doing a hard reset (no power for a few hours) can bring things back in line. The ECU can actually draw enough power off capacitors and what not inside the car to keep memory for over an hour with no battery.

GM's immobilizer allows everything to run except the fuel system. The fans coming on is part of a default failure mode. Did you notice if the mileage in the cluster had been replaced by "- - - - - -" That's also a good indicator of a hardware fault.

I would have your battery/alternator load tested. Typically these weird things happen when system voltage gets a bit low. A car with a marginal charging system gets parked for a few days, first time the ECU starts it up may see 8 or 9 volts and have the internal memory scrambled as a result.

It was probably a one time fluke.

If the charging system checks out ok, and the car never sits for more than 48 hours or so between drives. And it still does this. I would have the transponder ring in the ignition cylinder checked out.

[MSTRS:Me]

That sounds almost right NCP. Except there was plenty of power in the battery, and not 9 hours before the car had just been Napier - Manfield and back. I realise that any fault code has been wiped by having no power to the ecu, so I can't get it checked by a dealer, which is why I hope you are right with that 'one time thing'.
This morning all is well. Let's hope I don't get a call later from Yungatart "That bloody car..."

[no-coast-punk]

A weak alternator or battery doesn't care how much the car is driven. A dead cell in a battery will cause the max. voltage the thing can hold to be ~11v. A battery with a dead cell can still generate sufficient cranking amperage and the car will start normally. However to create this amperage the battery will have dig deeper into its reserve power and system voltage may drop below 9v even though the starter is getting plenty of amperage and the car seems to be cranking normally. That low voltage can play hell on computer memory. Same situation with a weak alternator. If the voltage regulator has been borked, the system will never get above 11v no matter how long the car is driven. The car will also not demonstrate any of the classic "low battery" starting systems. A big chunk of weird computer problems I have ever dealt with can be traced back to charging/battery problems at some level. Like... most dead fuel injectors die because they overheat. They overheat because they aren't getting sufficient voltage and have to draw more current to do their thing. Have the alternator and battery load tested. Every car with a battery more than 2 years old that I touch for any reason gets a load test for this reason.

Also, make sure that the alternator is tested for a failed diode. This is an easy test if you have access to a multimeter that can read A/C voltage. Take an A/C reading across both battery terminals with the car running. Anything more than about .05VAC is too much. I've seen cars running over 1 volt of A/C voltage that started just fine, but had all sorts of intermittent drivability problems due to memory getting corrupted.

The bottom line is that modern electronics are EXTREMELY sensitive to voltage fluctuations. Many of the old assumptions about 12v electrical systems no longer hold true. A battery/alternator issue these days will cause problems LONG before the lights get dim when you try and start the car.

I remember dealing with a Ducati 999R once. The owner had the frame powder coated a different color. One of the ground straps contact points hadn't been prepared properly. This lead to a 0.5 volt drop on the ground side of the electrical. This means all the electrical was running .5v lower than it should have been. The bike had all kinds of strange issues until the grounding issue was cleaned up.

Probably 70% of the crank/no start cars I deal with have 0 fault codes in the computer. The cars that do have faults stored were cars that died while driving. Most computers do ZERO fault code monitoring during the first few seconds of run time. This is because most engines do strange things on startup that would otherwise set faults. Constant on computer memory is also expensive and creates other problems. The very basic computer routines to make the engine start and run are always going. The routines for fault detection actually need to be "booted up" just like your home PC.

[MSTRS:Me]

Battery etc have been checked and are all good.
Holden specialist technician believes the immobiliser was triggered by an incomplete circuit between the key and ecu. He thinks the key is OK, so I have replaced the metal contact shroud around the ignition switch.
Now it's finger's crossed that has fixed the problem.

[no-coast-punk]

Excellent.

Sounds like the man knows what he's doing.

Lemme know how it turns out. It's nice knowing if I'm right... or an ignorant windbag.