The never ending problem.

Hi,

Thanks everyone for the warm welcome, further to my welcome post I have an issue with my own car which I have have been chasing as and when I have time for the last 2 years.

I'm sorry this is a long post and thank you to anyone who spends the time to read it.

The car is a 2006 saab 93 2.8 Turbo V6.

The issues started in the cold weather just over 2 years ago when my car started making a reasonably loud knocking noise on start up and under acceleration when cold, although I'm uncertain, it also seemed down on power.

After digging round testing etc the noise was pin pointed to cylinder 6, removing fuel or spark from that cylinder stopped the knock completely. Unable to find anything wrong with the electronics at the time and the noise sounding mechanical, myself and another qualified tech deemed it to be the engine (105k miles full history).

I was lucky and found a verified 60k engine at a breakers and swapped the engine. All was well apart from it seeming a little down on power which Iater I found to be a boost issue with the car only boosting to 5-6 psi instead of around 8.5.

The old engine was stripped after replacement however I was unable to find any issues with the engine bearings, pistons etc.

While looking for the boost issues I checked the actuator adjustment / pressure, the turbo, the commanded vs actual boost, pwm signal to the solenoid, I replaced the recirculation valve, I borrowed a known good MAF etc, as I could find nothing wrong I even tried reflashing the ECU with the same standard software already installed in case of corruption however still no change.

Whilst looking at this issue the cold weather started and the knock on the new engine returned although not as severe and now on cylinder 4. With this knock it started very similar to the previous knock however it has changed over time, at first moving from 95 octain to 99 octain fuel reduced the knock however now the knock is no longer only when the engine is cold but now when the engine is hot or cold and mostly only at idle and when setting off, it is also less severe, once the rpm's increase the knock is gone, although watching the knock retard on the scanner shows cylinder 4 to be most prone to knock / retard.

Having exhausted every avenue I could think of and only owning a basic single channel oscilloscope without much training or experience with it I decided to replace the engine ECU having found a cheap used one online.

The ECU was coded in and tested, instantly the power returned and some and after a few basic checks I can confirm this ECU is remapped although I'm unsure what map is on it. The car drives better and now boosts to 11-12psi.

I'm advised the ECU was taken from an otherwise standard vehicle and the breaker was unaware of the remap.

This leaves me in the position of not knowing if the old ECU was faulty or pulling the boost due to knock however the replacement has not cured the knocking noise.

I have also checked and confirmed the following recently,

Coils swapped with spares I have
All plugs replaced and regapped to 0.7mm (colour of the plugs is correct)
Smoke test (intake) carried out with no visible leaks
Gas analysis carried out, afr is very slightly off at 1.036 instead of 1 but not enough to knock (I don't think) and may have been an issue with the analyser as it has always previously been 1 when testing.
I have also done a coil test with an oscilloscope and cop adapter, the test was done at idle and all of the front bank appeared to be equal ( have videos of them if anyone wishes to see)
Timing chains replaced 20k miles ago when the engine was replaced, also replaced the oil pump and water pump

I have tried a used pre cat wide band lambda, maf and boost pressure sensor along the way however I have just ordered a new maf and wideband as they are old now and having new wouldn't be a bad idea

The car runs on both petrol and LPG and the symptoms are found on both fuels.

I'm really getting fed up with this car but I can't afford to replace it and I genuinely like the car, it was one owner and low mileage when I bought it 7 years ago and I hate being beaten and not fixing things.

Any advise welcome, thank you for your time

Andy

Hi Andy,

Do you have fuel trim values?

It could happend also that after you re-flashed ecu, it started to re-learn from factory values, to those values which makes engine work lean.

I am assuming here, that you have a lean condition due to ecu is reading rich (still don't know why). If so, you can add a resistor (1.8kohm) in series with ECT in order to make ecu think engine is colder and consequently add more fuel. That is only to test and we know what we are dealing with.

Hi Guafa,

Thanks for your response, I'll have to check the fuel trims as I can't remember what they were reading however this does also happen when the engine and ambient temperature is cold also ( as low as -5 Celsius) so I would expect the fuel at that time to be rich anyway.

Also when tested on the gas analyser wasn't far off lambda 1.

Will update with the trims as soon as I can.

Andy

Actually i am not to familiar with AFR and what is good or not. I mean, if you have 1.036 or 1.050 is it a clue? I don't know.

Do you mean AFR is near 1 at every condition? I mean open loop, aceleration

Ah, sorry I was only looking at idle ( although this is around where the knock is).

I think the best bet is to drive to work on Tuesday using only petrol and then advise on the trims.

Thank you again. If you have any thoughts in the interim please let me know.

Many thanks

Andy