3.3 Turbo EFI misfire ? Any ideas ?

[Ben V]

I have recently converted my C2 turbo to EFI.
Spec:
Motec
Motec CDI and coil
964 N/A cams set @ 1.3
GT35R turbo
ARP headstuds and rod bolts

The engine is currently on the dyno and has been mapped, I am pleased with results however I have an intermittent misfire which gets worse with heat.

At 6200 RPM the engine starts to miss a little on a 2nd consecutive test (when hotter) the engine starts to miss at 5500 RPM. It does sort of pull through it.....

I have changed, spark plugs 4 different sets, ignition system including coil, injectors, wiring loom, both crank an sync sensors. Motec in recording no errors at all!

I have checked the cam timing and valve clearances all see good.

The engine is producing 400hp @ 0.7 bar of boost (with misfire) and 499.7hp
Can anybody give me any other ideas?

[smurfbus]

I have changed, spark plugs 4 different sets, ignition system including coil, injectors, wiring loom, both crank an sync sensors. Motec in recording no errors at all!

Did this include plug wires? If you can put a strobo on suspect plug wires you can possibly see if it cuts out (not sure if you can see it at high rpms but on idle misfires can be seen as missed strobo light)

I just ordered MSD super conductor wires.

What is the spark plug gap? You could try to tighten it to about 0.028 to see if that helps. If it helps then widen the gap 0.005 each time until you get misfires again and then you have your sweet point just below that.

[Ben V]

Hi,

We thought the HT leads could be the problem, I was using the original spec Cooper core leads. In our quest to find the problem I replaced them with new magnacore silicone leads.

I have not gapped the plugs but have tried 4 different new sets, Bosch platinums original spec, then NGK race plugs both 8's then 11's, I am now back on NGK Iridium 8's. Are plug gaps that critical?

The misfire is fairly faint on initial run at 6200-6400 the with a second run imediately after it comes in at around 5200-5500 softly then clears and comes back at around 6000 ish then starts to pick up again towards 6400. It happens on low boost 0.7 bar but worse at 1.2 bar. 50-60% throttle it doesn't do it.

It is real head stratcher fortunately the motor is on an engine dyno and not in the car!

[smurfbus]

1.2 bar might be enough to be troublesome with stock spark plug gap. Hopefully more experienced guys will chime in but I would try smaller gap in the mean time. The super conductors I bought have very low resistance.

How is the AFR on those dyno runs?

[proffighter]

Have you checked ignition timing with the strobo and what values are you running?

AFR's would be interesting too...

[MJR]

Sounds like an ignition problem. Easiest check is to gap the plugs a little tighter. Also the coil may not be fully charging. You could try giving it a little more dwell.

As boost, heat, etc increase, it becomes harder for the spark to jump the gap. I use a scope to isolate these problems.

[Ben V]

So we have spent another long evening on the dyno, up to 81 power runs.

Tried plug gaps from .022-.035 no change in miss. We have had AFR's as rich as 0.79 but 0.82 seems to be best 499.7 hp with 1.15 bar and a miss!

The problem is definitely heat related 1st run is almost clear the miss get progressively worse by every consectutive test. EGT's even across both banks at about 750 degrees C.

The miss causes big blue flashes/flames from the exhaust. The miss only seems to occur when the manifolds are glowing dull red and hotter.

We are only running 0.7 bar of boost at present, seeing 409hp on the 1st run then 386hp with the miss.

Still no nearer finding the fault, any other Ideas?

[drmatera]

Do you have a good sized power wire for your coil? And voltage drop? Ground wires checked rechecked then checked again? And I don't mean make sure they are tight I mean loosen them inspect then reattach

[CaptainCalf]

Ben, I'm new at this and working on my own EFI so I'm interested in your issue and I hope you find the culprit soon!

This is a shot in the dark, but is there any chance you're floating a valve?
RC

[David]

I'm having a similar issue at about 5500 although I haven't spent any time solving it. I'm curious to hear what others find and hopefully I'll get off my butt and start looking myself.

I suspect it could be related to running 6 waste spark coils off 3 outputs from an M&W 4 channel CDI (the 6 channel wasn't available when I did the build.)

[turbo nut]

Does your ECU have data logging? If so look at it very close. Maybe post it if possible and someone may be able to help.

[sjf911]

Originally Posted by drmatera

Do you have a good sized power wire for your coil? And voltage drop? Ground wires checked rechecked then checked again? And I don't mean make sure they are tight I mean loosen them inspect then reattach

I agree, check the secondary +12V coil source under demand and also see if voltage corrected dwell is correct to give you adequate spark energy.

[David]

I downloaded data a few days ago, so I'll try to take a look this weekend.

[mark houghton]

An easy fix would be to go back to CIS and forget all this high-tech EFI/ECU/multi channel CDI's and wasted spark crap
Of course, you have to consider that my old-school school brain is taxed enough with the simplicity of CIS - having cut my teeth decades ago with an .016 feeler gauge, dwell meter, ignition points and a timing light....fiddling around inside a POS British roadster.
Glad I could be of some help. And now back to your regularly scheduled program....

[David]

I looked at the data log and I don't see any significant abnormalities. The only unusual thing I found is two spikes to about 8,000 rpm in two gears (4 spikes) during one acceleration with each spike lasting about a tenth of a second. I've never seen that before and they're in the 5500 to 6000 rpm range. So they're in the range but they only happened the one time in the 20 minutes of data storage. The advance, lambda, temp, MAP, etc all look normal.

I didn't drive to log this data but rather downloaded after a drive so next I'll go out and as soon as it happens again I'll download the data.

[Ben V]

Yes I do have logging and I have not seen any blips, errors or indfiction pointing me in any direction.

Mechanically the engine is rebuilt and can't think of anything that would give such a random misfire after 5000 which is definately heat related. The valves are new with new shrick springs. In have taken the cam covers off and everything looks in order.
I am convinced we have an interference issue somewhere........ The next job is to change the ref signal from the standard Bosch 60-2 with mag sensor to 6 timing pins with hall effect.

[TurboKraft]

Originally Posted by Ben V

I am convinced we have an interference issue somewhere........ The next job is to change the ref signal from the standard Bosch 60-2 with mag sensor to 6 timing pins with hall effect.

If you modify the flywheel, go 12 teeth. On a 60-2 flywheel, take out the tooth on each side of the -2 space, making it a -4 space. Leave a tooth, take out another 4. Repeat. Now it's a 12-tooth.

But I really don't believe this is your issue, not with a Motec M600. It can read 60-2 natively, and the hall effect sensor eliminates any high rpm timing errors (assuming reasonable sensor gap and good sensor wire shielding).

You've got a great coil, and the CDI takes care of all dwell issues. Per the coil's wiring diagram, the coil doesn't need big leads to it when running CDI.

The Magnecore wires are good. Are you using the original spark plug connectors or the silicone ones from Magnecore?

On the chance this is mechanically induced, were you able to log engine head temperatures when this happens? Or can you get the engine to that temperature and have a repeatable misfire, but not at lower temps?

[Ben V]

Hello, my thinking was using less pins, 6 or even 3 would give less chance for a problem. Although I agree it is a stab in the dark as there are no errors logged on Motec at all. So perhaps I should just try it with hall instead of mag @ 60-2 first.

To start with I was using the original Copper leads which were like new, just in case of interfereance I replaced them with new Magnecore leads, I am using the ends which came with them, identical shape to the originals albeit probably made out of silicone.

I have yet to run it again to physically log the head temp (I need to rob a Shell station first as I have already burnt 300L of petrol testing). I will log it. The miss has only 1 constant that is heat. The hotter the engine the earlier the miss comes,
1st run it doesn't miss at all, 2nd run after the headers and turbo have gone red it will miss fire at may be 6000-6200. 3rd consecutive run the miss will come in at say 5200 and pulls through it, 5800 and pulls through it , 6200 and pulls through it. Each time it misses it loses about 10hp. You can hold it at the miss fire points on the dyno and it continues to miss (blue flashes/flames from the exhaust) so I am assumming unburnt fuel / ignition.

The loom is absolutely top notch, in fact he does wiring for F1. I have had a conversation with him to confirm the sheilding and get his advice which is, make sure the coil wires from the CDI are twisted (which they are), sheilded and make sure the CDI unit is not too close to the ECU, which it is on the Dyno.

My only slight reservation is; we swopped the CDI setup out for the OEM bosch setup and got the same problem.

There must be something so simple I am missing, as I could probably build 2 more EFI cars with the amount of new bits I've tried!

Has anybody ever made a diesel C2 turbo.......... because I'm getting that way!

Thanks Ben

[sjf911]

I would favor swapping out your spark plug wires to completely new ones, including new connectors at both ends. You might be having issues with a crimp or connector as the temps go up especially at the spark plug end.

[Ben V]

I have completely new leads/wires, how about a soft valve spring ( with heat! ) we tested the new shrick valve springs prior to fitment! Could I have a soft 1? Clutching at straws now.....