Tom M'Guinn

Ah, so that's a bit different. What is it under load?

WPO

A bit of an update.

We did try the KLR bypass by installing a jumper wire between pins 9 & 16 - however this did change the advance so much that the car didn't even start. Jumper wire removed -> car starts normally but of course with previous problems.

Then we started to check things which have been supposedly checked by the shop already: TPS. And of course the ohm values measured were not within range so this indicated a bad TPS. Tomorrow we'll get a new TPS and I'll post the findings here accordingly.

Really interesting if this has been the culprit all the time...and this wasn't tested by us as the shop told yes of course TPS has been measured classic...

Edit: I don't exactly know what the advance is when the header glows. I'll check this from my buddy also - if TPS doesn't solve the issue

WPO

New TPS installed and same problem persists.

What on earth is wrong with this car? We're running out of ideas

WPO

It moves around 30-degrees without load and then it less than 10-degrees under load. So it's 20+ degrees late and hence the mixture burns in the header which becomes glowing red.

The numbers aren't "exact" but approximate i.e. eyeballed. Anyways it is in "safe" mode (or whatever that is called) as it gives 0.2 max boost and advance is stuck.

odonnell

It really seems like the KLR is pulling out timing due to perceived knock under load. Maybe someone can chime in with the actual amount/authority it has over timing. The fact that jumpering it keeps the car from starting may also be a symptom.

WPO

Is there a way to bypass the KLR? Jumpers between pins 9 & 16 didn't work as it moved advance so much that the car didn't even start?

Knock sensor has been replaced and wires have been beeped so they're not broken.

This is really an insane problem Most likely it's something really simple and stupid but that needs to be found first.

harveyf

I would remove the knock sensor from the block (but leave it hooked up electrically) to isolate it from any vibration and test again on rolling road. You may have some sort of vibration that the knock sensor is interpreting as knock.

WPO

Knock sensor is currently hooked up but not bolted down to the block so unfortunately that is not the culprit.

Perry 951

Does the car still have the catalytic converter in place? If it is, disconnect from the rear section of exhaust and see if the substrate is melted or missing. If melted, there is your restriction. If missing, it could be lodged in the muffler causing a restriction.

WPO

Yes cat is there and it's been checked also and it was found to be OK.

I think we'll start crying soon as the car is possessed with some unknown gremlin Or then we're just missing something really obvious? Pretty much everything has been checked?

V2Rocket

now i know the ignition rotor has a wide edge to allow for a range of timing...and that 944 spark timing is electronically controlled and is theoretically independent from cam timing...

but i can't help but wonder if your camshaft is timed 2 teeth off.
2 teeth would be 18 degrees on the cam gear. believe there is just enough valve-piston clearance to allow such an error to still run.

Tom M'Guinn

What is the history of this motor. Was it running ok, then started to act up, or is it a fresh build and/or have a new flywheel?

The part about the car not starting with the KLR bypassed doesn't make sense. Are you sure you did that right? Remove the big connector from the KLR and short pins 9 and 16, so that the DME feeds it's own ignition pulse back to itself without going through the KLR. I can't think of any plausible way doing that would advance the ignition enough to keep the car from starting. At the far edge of hypothetical, I suppose the TDC pin on the flywheel could be way off causing the DME to have wildly advanced ignition, but if so,I can't see how the motor would ever run well with or without the KLR. I'd try bypassing again and pay close attention to pin numbers.

Have you tried checking the blink codes yet? There's a very real chance they will help. Discounting your KLR bypass test for the moment, your problem doesn't sound far from the factory service manual's description of what happens when there is a fault detected in the "knock/charge pressure control" system. The FSM says: "If a defect has been found, the engine will operate on a safety system (basic charging pressure up to approx. .3 bar overpressure and 6 degrees later ignition), which means significantly reduced performance" --in other words, limp mode. If you look at the WOT map posted above, and subtrat 6 degrees of timing, it matches up with your symptoms pretty closely.

Perry's suggestion is a good one too, and could also explain your issue. I'd just check the codes first since it's so easy and may help point you in the right direction...

Perry 951

When/How did this problem start? I've read everything that has been checked/replaced but not much about what happened to arrive at changing turbos and checking sensors.

Tom M'Guinn

As you say, the spark will jump when the DME tells it to, regardless of where the rotor is, but I suppose the cam could be so far off that the spark is jumping to the wrong post in the cap, effectively changing the timing by 90 (or I guess 180) degrees. Hard to imagine it would run much if at all that way though...

WPO

The engine was running fine earlier. The problem started when the car was tracked and after that it lost boost as per the description given.

Turbo was reconditioned by the shop it was in - even though there wasn't an actual problem with the turbo. TPS was changes as the ohm values were not as specified in Clark's instructions. The new TPS has the exact same values though? Reference and speed sensors were changed as it was thought that the issue was caused by bad sensors. AFM was measured and found to be good so it wasn't changed. Knock sensor was changed as it was thought that the problem is caused by that - which it could be - but to no avail.

Actually we did test the KLR in a wrong fashion: the pins were shorted and then KLR was attached in the connector. We'll retest this by just shorting the pins and KLR not connected. Somehow we thought that this was the way as there were no real instructions apart from the shorting of those two pins So thanks for pointing that out

We'll check the blink codes also and I'll post here what the findings were. That will be on Monday at the earliest as my buddy is on a small trip this weekend.

Thanks for everyone so far. Let's tackle this gremlin and get the 951 back on road