996.2 Strange set of faults, help please! (Resolved)

[YPVS]

Hi There

Agreed on the ECU, but actually it is the exact same part number.

I'm confused now. You said the car didn't come with an ECU and the one bought second hand and fitted was a Swiss regional DME. Now you say the part number matches an ECU which the car apparently never came with. :dont know:

Would a DME from another car not need coding to the alarm and key?

Part numbers here...

https://www.bellandcolvill.co.uk/deroure?qs=1&preview=1&make=3&model=25&groupid=2721

 

[berni29]

Hi There

As far as I know all of the >2002MY 986/996 NA cars have the same Bosch ECU with part number 996.618.604.00

Also I believe that there are only two variations of the model specific mappings that are flashed onto the ECU's and that is USA or ROW. So no difference as long as it comes from a non USA spec car.

And yes you are correct, it has to have the matched alarm module and ignition key transponder with it to work.

So in that respect swapping the set in and out of the correct MY 986/996 is a simple task.

Berni

 

[berni29]

Hi been very busy this week, only just had the chance to do the diagnostic's.

 

[berni29]

More

 

[deMort]

Ok .. certainly some issues here ..

Bank 2 is wrong .. its too rich .. was this listed at the same time as the test on Bank 1 ?

Something wrong there from the graphs .

lambda sensors are wrong or are reacting to a wrong mixture .

Bank 1 adaptations are almost at max whilst bank 2 don't seem to be adapting .. RKAT at 6.09 .. max limit is 6.9 .. issues with both banks basically .

STFT is wrong .. they should be pretty much matched bank to bank .. they are not .. you have a fueling issue going on there .. dme , maf , injector leak , misfire .. pretty much everything can cause this .

Engine load is hard to work out as i don't know those figures .. is it 4/5 or 4/100 .. 4/5 is obviously a massive load on the engine whilst 4/100 isn't .

My tester and i would expect about 23 % engine idling load on the maf .. maybe upto 50 % when driven .. these figures make no sense to me i'm afraid .. i need a context as to the limits .

Short on time tonight so i'll go through this lot again tomorrow .

 

[berni29]

Hi DeMort

Thank you for taking the time to look!

Engine load was just over 25% but the car was not completely warm. This is from the Durametric.

When you first start the car from cold it idles steady as a rock, no misfires whatsoever, then after about 30 seconds. I think it then starts to use the data from this the lambadas etc and it all starts to go horribly wrong. That makes me think the fuel pressure and injectors are ok, but the metering is wrong when it goes into the closed loop.

I have sprayed cold start around the induction system and vacuum pipes but it's seems to make no difference. I guess a smoke test is in order.

It looks like at least the bank 2 pre cat 02 sensor is toast. is that enough to cause the bad running issues?

The MAF reading of 21kg/h sounds within limits, but I could not find a lot on this.

As soon as the car goes into Drive or Reverse the wheels immediately fall off (so to speak). The engine goes from bad to worse.

All the best, and thank you again

Berni

 

[deMort]

It's great to have some data to work with so thank you for that .. i expect this will be a long post but please read it carefully along with my last one.

Some of this might sound like i'm teaching you to suck eggs but i'm going through what i can think of

I've looked in detail at the readings .. most of which are fine so ill concentrate on what's not ..

Engine load at 25% is correct .. when reved i would expect closer to 40 % but with not driving the car it might not go that high .. test it anyways and see what it says .

Maf at idle should be about 18 kg/h .. 21 is ok but you need to rev the car and see a rapid increase in the reading .. 100 plus basically .. this is only a snap shot though as testers are not fast enough to give a true reading ..

Ignition on but not running and look at the maf voltage .. it should be 1 volt .. anything above or below 9.0 - 1.1 is a knackered maf .

The Maf is also a critical component of the fueling .. you have replaced / swapped out with no difference so other than a corroded wire which would be unusual i'll have to assume this system is ok .

You also need to look at atmospheric pressure .. this is a value and a component that's in the dme .. it should be 1024 mb but anything around 1K is fine ... assuming your not up a mountain that is !

This value alters the maf reading in the dme .. it adds or subtracts fuel according to what height it thinks the vehicle is at due to air density being lower at higher altitudes .

Bank 1 lambda sensors are reading correct but the RKAT ( adjustment at idle ) is almost at max .. the car is opening the injectors for longer to give more fuel ... you need to do a smoke test as its consistent with an air leak for that bank atm .

To test the post cat sensor bank 1 then rev the car then let it idle .. it should read low then high then stabilize .

Bank 2 is also at fault .. it's not adapting when bank 1 is .. so why ? .. its reading a rich mixture and a high one at that .. both pre and post cat sensors are reading this .. it will damage the cat at this rate ..

So .. any chance of a blocked or damaged cat on B2 ? .. hit it ( cold ) with your hand .. any rattles ? .. good exhaust flow out of the tail pipes ?

lambda sensors when dead tend to read about 0.44 and don't alter .. yours look like they are reading correct for the fueling ..

Are these all new lambda sensors ? if old then both B2 sensor replacement is a start but we have issues on both banks atm .

Camshaft angles ... you need to rev the car and look at the actual angle versus the specified angle .. they should be close at all times ..

looking at the reading for them you posted ... i think i would recheck the cam timing .. bungs out , stick the tool in and the same for the other side against crank at TDC .. pin it to be 100 % there .

Could it be timed 180 out perhaps ? cam shafts as you know have to be in a certain position when timing these up .. that is certainly a possibility atm .

I have no idea the running issues with this or even if a car will run if this happens so it's just a thought but one i would certainly want to confirm as being correct !!!!

As this is what i would call an induced fault .. by that i mean it could be something that won't happen with a car on the road then we are also back to correct components .. if the dme is the correct part number for the car then at work i would reprogram it with the latest software .


What else would i check .. hmm ..

Smoke test of the inlet tract ..

Fuel pressure .. regardless of whether i thought it was wrong of not just to eliminate it ..

I might pull off both cats to prove no restriction .. it will be loud at that point but will rev up if so .

The fault is certain at closed loop .. so the lambda sensors are now in control and the car is adapting to what they say .

 

[deMort]

This is basically an EDIT for my previous post ..

I've been sitting here thinking how can a car run rich on one bank and weak on another ..

I'm no engine man these days .. it's been many years since i've rebuilt one so this sort of stuff doesn't come naturally to me .. i work out faults these days and this is not a normal fault i've ever seen ..


Now then .. many years ago and more than i care to remember i did my Porsche Bronze test .. part of this was to set the cam chain timing on a M96 engine ..

I got it 180 out and it was only when rotating the engine and rechecking i realised my mistake and reset it correctly before the test time finished ..

The engine rotated with no issues from memory .

My feeling atm is this might be the issue .

I would certainly say check it even though it's a lot of work if only to eliminate it .

 

[berni29]

Hello there again

Well, we did a smoke test and found that the large plenum jubilee clips were not tight enough. It has not made any difference though.

We changed out both o2 sensors because one was dead. No difference in running, however we are observing the following behaviour.

We get 0.445v or so on both the o2 signal wires on the loom.

On starting from cold we get 0.5v approx on both o2 sensors until it goes closed loop then Bank 2 goes to 0.03v - 0.05v

We checked continuity between the 02 sensor plug pins and the ECU and all is good.

We are still getting the P0151 code.

Both short term fuel trims are at 25% unless you rev the engine in which case they go to zero then come right back to 25%

MAF airflow seems to be between 25-28 kg/h at idle.

We are going to drop the engine down a bit and check for further air leaks......

EDIT, did another smoke test, no obvious leaks............Ugh this is driving us nuts.......

Many thanks

Berni

 

[deMort]

Are both banks at 25% or is one at -25% ?

Both at 25% with a lambda reading at 0.03 is a massive ... massive lean mixture to the point the car can not open the injectors for any longer to compensate and bring it back into lambda .

0.5 volts in open loop is correct , it's also the voltage when a lambda sensor dies .

When they go closed loop the lambda sensors are active and telling the dme the mixture by means of a voltage ... in this case a lean one and then some .

I can honestly say i've never seen a " reported " mixture this lean .. i'm struggling to see how the engine would even run .. i would expect you would hear a sucking noise from an inch or so diameter pipe that's not connected its that lean .


Sooo .. that's what the lambda sensors are saying .. be very interesting to see what a co machine reads on each tail pipe .

Ok so that fault code .. its below .. i'm not sure if this is for one bank or both or indeed which bank it is but if you read the very bottom line that's what Porsche say to do in the event of the other checks being ok .

Car running that lean and the fault code .. i would go with that .

IF one bank is -25 % though with the other at + 25 % then i'm back to camshafts as mentioned .


Just an fyi .. when checking wiring it's not just for open circuit .. you need to check for a short to positive and a short to negative on every wire tested .. you may well have continuity but if it's shorted to earth it will kill the signal and cause this issue .


EDIT .. if you want to chat on the phone sometime or even better via facetime i'll be happy to do so .

 

[berni29]

Hello again

Thank you for your advice last night Mr DeMort. We were thinking more systematically today.

We did the pin to earth and bulb tests today and everything checked out.

We also pressurised the intake manifold (by using a device that connects to the throttle body. The system would not pressurise. We rotated the engine to what seemed like a good spot. I could hear nothing from either exhaust when injecting air (via a long tubed floor standing bicycle air pump).

However we could hear a hissing from the back of the engine nearish the SAI it would seem, indicating an air leak you would think.

So seeing as the airbox and MAF were off the car we started her up to check the state of play diagnostics wise.

In open loop, both pre cat o2 sensors were at 0.46v, then in closed loop lo and behold both started to behave as normal. Pic below. Both STFT at 15-20%, car idling nicely. Using a hose cannot hear the suspected induction leak though. Having said that its noisy so it might get drowned out.

With the MAF back on, same 0.46v on both pre cat o2's but in closed loop both went to 0.05 - 0.06v. STFT at 25%.

On revving gently to 3k both STFT stayed at 25% but on snapping the throttle shut came down to zero with the engine idling roughly, them over the next 15 sec or so came back to 25% with the idle of the engine improving and speeding up a little as they did so.

I cannot explain the o2 sensors both behaving in unison now apart from an intermittent wiring fault.

The STFT at 25% must most likely be due to a significant air leak, which could also cause the 0.05- 0.06v reading on the o2 sensors as I understand it.

Fuel pressure gauge arriving tomorrow.........

Many thanks again

Berni

 

[deMort]

Ok .. you might be getting somewhere here .. i didn't think of the SAI but it makes sense ..

SAI .. air injected into the exhaust to fast warm up the cats .. imagine if this was open to atmosphere .. fresh air constantly getting into the exhaust .. the lambda sensors see this as a weak mixture and try and compensate ..

25% is the max it can compensate by .. this is the injector duration being open for as long as possible .. and it's still too weak .. as i said last night a MASSIVE air leak to cause these readings .

The car .. if this is the case is actually running correctly and getting a false reading via the extra air in the exhaust .. its making the car run far too rich !!!!


To test .. hmm .. quickest option .. ignition on and clear a fault code .. if no codes then unplug the maf to induce a fault code , replug it in then clear the code .. this will reset the adaptions to 0 .. ign off .. unplug both Pre cat lambda sensors .. start the car and see what it does .. it will be running without lambda sensor adaption at this point and running on default values .

Whilst this can hide other faults its a good indicator i feel rather than trying to get at the SAI for now .

Your graph shows correct lambda switching followed by a huge lean reading btw .

EDIT .. it's possible the fault code is because the readings are so far out the dme thinks there is a short in the wiring .. readings like this just don't happen on the road or a normal fault .. induced as i mentioned is the only way .

I'll also add for anyone following this .. Induced means a second hand part fitted that is faulty .. something that can't happen to a normally driven car .. not a mistake as such .

 

[berni29]

Hello Again

Well, being systematic won in the end. And believing that whatever you are left with must be the truth.

Cutting to the chase the car had the wrong injectors in. My brother (and thankfully not me) ordered the wrong ones and somehow did not check the new against the old even though it is something he would usually do in case of packing errors etc.

Original injectors are back in (ugh job) and she is running perfectly subject to a road test. Both o2 sensors are behaving and the STFT are a +2%-3% which is fine for now, and the engine was cold and only run for a few minutes.

Case closed. Ready for the next one!

All the best and thank you again!

Berni & Bro

 

[deMort]

I can honestly say i would never have guessed at the wrong injectors fitted .. kudos for finding the fault .

 

[Griffter]

Don't be too hard on yourself DeMort. Your first reply mentioned fuel pressure and injector duration. Not a million miles away!

Well done Berni. Been following this with interest.