94legturbo

iTrader: (6)

My car ('94 Legacy Turbo) has developed a problem where at roughly 5-5500rpm there is a split second voltage drop of around 8V which in turn knocks out the ECU and injectors/coils which cuts the car out momentarily and stops you from revving beyond that.

I have discovered this whilst doing some data logging of the ECU by using a handheld oscilloscope where we monitored the reset line and 5v and 12v power lines. We found that when it cuts out at around 5500rpm, there is a significant spike on the 12v power line where it drops by 8v knocks out the ECU causing the processor to reset.



With a bit more thinking, we monitored the 12v supply from the cigarette lighter and then from the battery itself to determine if the power drop was localised around the ECU but it turns out that the whole system suffers from the drop!



Looking at the readings above we saw that there were smaller spikes before the main one split by four cycles, making it sound like it could be related to one of the cylinders. Using this data, we went out again and got some readings from the CAM sensor to determine where in the cycle the issue was and it came back as always occurring on cylinder number 4.

This would therefore be a good place to start looking. However, I am confused to what could be causing it really, because as i have replaced every coil pack, injectors, spark plugs, ignitor module and ECU.

I'm wondering if the wiring has an issue somewhere, where its arching and shorting out somewhere. But this only happens on load and at these specific revs too, if you rev it in neutral its fine.

If I were to measure the resistances of the wiring, i wouldnt expect to find much as the car works fine at lower revs, its only when it hits around 5500rpm that the problem occurs. If there was a resistance problem that i could measure under the bonnet, I would be surprised that the car would run? If it is a short, then whatever its shorting against isnt fused as i havent blown anything.

The list of things to try I currently have are:

1, Data log again tracing the ground as we forgot to do this.

2, Try changing the battery/running without alternator to rule them out

3, Measure the resistances in the wiring,

4, Swap coils around to see if the problem moves to a different cylinder in the cycle

Apart from this, I'm a little stuck.

Has anyone ever experienced anything like this before, or got any suggestions of where else to look?

In a nut shell, I have a spike drop of around 8v which is knocking everything out. This is across the whole electrics, back to the battery. It occurs within cylinder #4's cycle each time, but only at 5500rpm.

saiklon

Nice investigative work there.

One idea that came to mind is that somehow you have a tuned (inductor -capacitor) circuit centred at 5500rpm which shorts out the power much more when its resonant frequency is reached. Since the coil is an inductor, maybe you could check the wiring between the ignitor and coil no 4 and swap the coil around as well. (Even though you've replaced coils already)

I'm just throwing out a wild idea here, it sounds a bit far-fetched to me as well.

94legturbo

iTrader: (6)

Thanks for the suggestion. I have been trying some things this afternoon and have ruled out both the Battery and Alternator.

I have also measured the resistances between the Ignitor unit and each coil, but they all seem to follow a pattern and it doesnt look like there are any anomalies on #4 in particular.

These were my results against each pin of the Ignitor:

Yellow wire:

#1 - 0, 1.5, 1.5, 1.5
#2 - 1.5, 1.5, 0, 1.5
#3 - 1.5, 0, 1.5, 1.5
#4 - 1.5, 1.5, 1.5, 0

Red/blue/green wire

#1 - 0.6, 0, 0, 0
#2 - 0, 0, 0.6, 0
#3 - 0, 0.6, 0, 0
#4 - 0, 0, 0, 0.6

Having had a look around the wiring, it looks ok from the outside - nothing obvious at least?

I have some spare coils that came from the car before I changed them, so I might try swapping #4 tomorrow for all the spares to see if it makes a difference.

Are there any other circuits that you can think of that might be inductor/capacitor related? thanks

Wooffy

That's a fair sized drop. Take a look at earth leads chassis to engine and body. They may look tight but there may be enough corrosion to cause the drop. Also I'm assuming you are knowledgable enough to measure for voltage drop using a multimeter, as you could have a break/weak point midway down a wire.
Also check fuse connections as they can have enough tarnish/corrosion on them to cause the drop when that system is put under load.

94legturbo

iTrader: (6)

Thanks. I've cleaned up all the earth leads in the engine bay, and have an earthing kit too, but have a feeling there may be one underneath somewhere, maybe between the gearbox/engine and chassis directly, so will have a hunt for any more and clean them too.

Someone else has mentioned the connections under or around the fuse box, so i will have at them and the fuses tomorrow.

I am fairly alright with a multimeter as long as i know which wires i need to be checking

94legturbo

iTrader: (6)

Right, well today I had a look in and around the fuse box and all the connections etc are all clean and free from corrosion.

I also swapped coil #4 for a spare and interestingly it moved the point of the cut out to around 6500rpm from the original 5500rpm.

I thought it had fixed it at first because it revved through 5500rpm quite freely, but alas it was still there.

When it hit the wall and cut out, I got a definite CEL flash on, and a second time when i tried but when I got back to check the codes, there was nothing stored, which is mighty strange, as every other time its always stored a code if it flashes on, no matter how quickly.

I tried another coil pack and the cut was around 6000rpm, one time at 4500rpm, then just to see, i popped the original one back on and it went back to around 5-5500rpm again.

The fact that it differs, makes me wonder if it is something to do with the coils or the wiring to them?

Wooffy

Not trying to go through the sucking eggs routine but I'll post up how to trace voltage drop so everyone knows.

Firstly you need a multimeter and set it to volts DC. The circuit you are testing HAS to be under load. Now here comes the bit that confuses a lot of people so I will explain in the form of an example.

You suspect the live lead to the starter motor is faulty as the starter was turning slow with a fully charged battery. You replaced the starter and it's still doing the same. Switch your multimeter to volts dc and place one of the probes on the positive battery terminal and the other on the positive terminal on the starter

(this is what confuses people as you are probing the same wire. The volt meter doesn't measure voltage per se it measures pressure difference and this works when the circuit is loaded as it always is when you check positive to negative)

Next you get your assistant to crank the car over whilst you observe the readings (this is putting the circuit under load) on a circuit such as the heavy amp loap of the starter 0.2 volts is acceptable but it's worth striving for as low a reading as possible. If the reading is for example 1 volt then that circuit has dropped 1 volt and is only being supplied with 11 volts. The drop can be cause by dirty connections or wires being broken or having been subject to excessive heat.

This way of checking can be used on earths and any other circuit as long as you put the circuit under load. It allows you to pinpoint faults that trying to check for resistance can't.

You can even test long lengths of circuit by starting at both ends and then working your way back bit by bit probing through the insulation till you find the affected section.

Now all you need to do is breakdown the loom into sections (metaphorically don't pull it apart) and start testing.

94legturbo

iTrader: (6)

Thanks for the procedure, do you think it may be loom related then? Its about the only other thing I can think of now..

I also worked out why it wasnt storing the codes - because the ECU was resetting.. whoops.

saiklon

I would re-do the wiring for coil #4 from ignitor to the coil as the next step.
For the test you can also power the coil directly from the battery.

So, run a new trigger wire from ignitor to coil #4 and a temporary wire to the +ve battery terminal.

94legturbo

iTrader: (6)

excellent, thanks. So the wire that comes from the ignitor is just a signal, and the other one (yellow) is a live, presuming that it earths through the block?

I'll give that a go then, cheers.

94legturbo

iTrader: (6)

Ok so I tried rewiring #4 yesterday but it didnt change anything

what to try next?

saiklon

Since you can get the cut to move to 6500rpm by changing the coil maybe you could live with it? Or try different coils till you get one that will cut at a higher RPM.

I once had an experience where I put totally wrong plugs on the car and they would misfire so bad that the relays within the car would click. I know you've changed plugs but it might be worth swapping the #4 spark plug and try again with the 6500rpm coil pack installed.

360ste

iTrader: (1)

I agree it sounds like a resonant frequency issue due to the resistive, inductive or capacitive tendencies of the circuit. Have you looked at the plugs for signs of corrosion on the threads or the connector for the coil pack? You could get a cheap set of plugs and fit them SHORT TERM to see if this fixes the issue.

94legturbo

iTrader: (6)

Much as I would like to, my car is needing a remap as i've changed a load of bits and the setup is wrong at the mo. The mappers dont want to do it until this problem is sorted as its causing them issues to map around.

also I would be worried that even though it could be at 6500rpm for a while, it may even move in other gears? As I have some track time booked in september, Im keen to get it sorted once and for all.

thanks for the suggestion on the spark plug, i'll try anything now

Jolly Green Monster

iTrader: (4)

Try add a ground wire from neg battery terminal to alternator body/bracket

94legturbo

iTrader: (6)

The plugs have literally done around 60 miles and look like new still. The coil packs all look fine by eye taking each off too, the contact bit which touches the spark plug is clean without corrosion, so no obvious signs.

I think you're right on the resistive/ inductive/ capacitive elements of my issue though, I'm thinking of ringing an auto electrician tomorrow to discuss whether they think they would be able to track the problem or not..

Thanks for the suggestion, I have got an earthing kit and one of them is to that position. I have also cleaned up all the earth contacts i could find around the engine bay (including the kit ones) to no avail.

94legturbo

iTrader: (6)

Well I've spoken to an auto electrician and he said it sounded like quite a weird problem, but he would have to take a look at it. So currently its booked with them next week on the 11th.

The thing is, I dont know what capabilities auto electricians have. I dont mind paying them if it sorts it, but I'm not keen on spending lots of labour charges for them to find nothing etc, especially if i can get hold of another loom and swap that -which is an option. But again i dont want to go to the expense and effort of swapping the whole engine loom if they can track and sort the problem more easily.

On another note, when i disconnected the alternator to rule it out, i didnt remove the belt. Its a bit of a long shot, but if the alternator was still generating power from spinning and arced out at a particular rev range to something nearby, could that cause the loss of 8V, or would it need to be connected to the battery etc for it to cause a problem?

94legturbo

iTrader: (6)

The car visited the auto electrician last week, but he couldnt find anything and was confused by the symptoms. The only thing he suggested that I hadnt already looked at was the immobiliser, but I'd be surprised if it could cause such a specific problem?

The last couple of times that I've been testing the car, I get a CEL as it cuts out, but then cant get the codes because the ECU resets. I am doing some more data logging this week, where i will trace the ground, and hopefully log the error codes before they disappear.

Fingers crossed it might tell me something.

saiklon

The CEL is most likely showing because the ECU is restarting at that point due to the low voltage. There are no actual errors which is why when you check for stored codes you don't find any. Any genuine error would not be lost during the sub-second loss of power to the ECU that is happening.

94legturbo

iTrader: (6)

oh, that would be annoying if thats the case.. I'll give it a go anyway just incase there are any codes, but not hold my hopes up..

94legturbo

iTrader: (6)

Its fixed!! Woohoo!

The problem turned out to be a loose ECU ground to the back of the engine block. So after 6 months off the road, a quick 2 minutes tightening said bolt and we're back in business! Frustrating, but I'm dead happy.

The loose bolt was making the ECU ground go open loop and it is mounted next to cylinder #4 hence why i think it got interference in the firing cycle.

Thank you to all who helped me along the way

JdcTypeR

iTrader: (19)

I was just reading through this and was about to say I had this problem and it was the earth at the back of the block passenger side. It used to just cut out randomly.
Im Sorry I didn't read this thread sooner mate as it could of saved you the headache.

94legturbo

iTrader: (6)

no worries

alls well now and i'll now know for next time