Stuttering, hesitating 550, high LTFT, some solutions.

I started the OBD location thread(it is there ...etc) about the 3-pin OBD plugs on the euro 550's , but that soon developed into describing the problems I have with my 550. So I was in fact polluting my own thread.
As this is something a lot of people can be helped with in the future, I decided to open a new thread.

Originally I started with a hesitating, stuttering 550, with the occasional SDL and even CEL, which is very rare on a euro 550 (on a US version it lights up when you sneeze I understood). Symptoms mainly on low rpms, but idle was ok.

My OBD scanners , one on each Motronic plug, showed a consistent lower MAF value on one side, and a flatline O2 sensor on one side when stuttering.
The MAF difference between one side and the other was around 20% ( for example 16 and 20)
The long term fuel trim values were ok on one side, around 1, but around 13 on the other. So a lean condition.

Problem with the 550 V12 is that the signals are crossed: MAF value of the left sensor shows up on the right ECU and vice versa.
In the end it turned out that even Throttle Position, and Fuel Trim values where crossed-signal!
As I had two scanners connected, I simply switched their places on the board I had attached them to, just to make things simpler to understand visually. Now my left scanner shows all left values.

Here is a list to check if you have a hesitating car, all of which I checked several times :

- Fuel pump displacement out of FPR: 700 ml in 30 seconds on the left side. This is low compared to what the WSM says (1 liter), but not low enough to cause a lean mixture on low rpm's.
(My car was bought with freshly "repaired" pumps, but the garage had installed cheap rip off pumps of 3 bar instead of 4 bar Bosch pumps. I originally had like 150 ml flow...)
- Fuel pressure, with vacuum line connected and disconnected. Should be a minimum of 50, and not drop fast after stopping the engine. Was ok on mine.
- MAF values are lower on right MAF (= left ECU), but do not really drop while stumbling, so the MAF's do not seem faulty. MAF's react normally it seems.
After the major and a lot of testing, the MAF values are equal.
-TPS (both) values show normal behavior, and reflect the pedal position , even during stumbling. Max value of 72% is normal, starting with 0%. Backprobed for real V values, left and right must be within 0.07V
- with the help of Ian and his schematics , 20.000 km away from me, I checked resistance on injector harnesses, engine harnesses, ECU, all ok. There was a small resistance that could not be explained, but not likely to be the cause .
- checked ECU's for bent pins, obvious corrosion etc. Cleaned the contacts with deoxit.
- did the complete major with retiming of the cams, was not really off, but is timed slightly better now.
- smoke test, more about that later

replaced or repaired:

- 2 Fuel pumps (4 bar), replaced the rubber collars
- 2 fuel filters
- 4 O2 sensors , both up and downstreams. They are 23 years and 130.000 km old, so does not hurt to do that even if they still work
- 2 crankshaft sensors, same story.
- 2 temperature sensors in the V during major, "while you are in there "
- 2 MAF sensors, bought a used one, and 2 new copies.
- 2 air filters
- 12 plugs
- All gaskets from the intake manifold/plenum, during major
- cleaned the IdleControlValves

All of these things did not help, car kept on stuttering, with high LTFT on the right side.

Now for the solutions I found:
- the return line to the pump baskets I had to make myself as the garage had left them out. Just as the venting straws...
It turned out one return line I made was partially blocking the entrance to the basket. The low rpm problem was actually a fuel -moving occurance: when exiting a roundabout, with 20 liter left in the tank, the pump basket was empty because entrance was partially blocked, and the fuel was forced away by the centrifugal force on the roundabout. So the fuel pressure must have dropped at those moments.
I repaired that, added the vent straws, and the stuttering was gone!

Then the Fuel trims.
I did several smoke tests, to find 3 vaccuum leaks.
The first 2 were the axles of the throttle body valves. I took those apart, and revised them with 7 mm high bearings. The original were 5 mm high, and were running dry, which had caused play of 0.1 mm. Not much of a vaccuum leak to begin with, but after repair no more faint smoke, and fresh bearings. And after rebuilding the throttles, the MAF values are equal .
But the biggest leak was at the back, on the solenoid that used to open the muffler valves. (I have aftermarket mufflers without the valves). It takes some time to find that, especially if you look at the engine for leaks...
So all vacuum leaks repaired. But still the fuel trim got no better than 11. Almost acceptable.
I am waiting for three secondary air valves to arrive, maybe the exhaust draws fresh air , confusing the O2 sensor. This is to be continued.

Then what is really puzzling me:
Today I switched the MAF's.
As a result, MAF values again differ 20% between left and right. Beh...
But the LTFT is ok now, with still 1 on the right, and 5 on the left! It dropped from 11 to 5 in 1 km of driving.
What could be happening? Two slight problems compensating each other?

 

Well, yes. And myself, for being so naive...
But as the engine appears to be structurally ok , and it is sort of fun to fix problems, in the end I expect to have a well sorted 550.

Now about cross-data, it turns out that the plug on the left side Motronic, shows all data from the right side of the car.
Not only MAF values, but also STFT and LTFT, and TP and O2 sensor data.
That is why I simply switched the place of each scanner on the board I have them on in the car , so now my left scanner shows the data from the left side of the engine, while being connected to the right hand plug. It makes understanding data while driving a little bit simpler, you do not have to translate in your mind...

It is well possible that the plug does not come out of a motronic where it is located next to, but is in fact going into that motronic when no scanner is connected.
And that it is in fact connected to the Motronic on the other side. But that is not really important.

Now about the strange positive effect the MAF switching had.
I really do not understand how the MAF values were near identical, but the LTFT was clearly not,
And after switching the values differ so much, but only the high LTFT dropped to normal levels, while the other LTFT did not change.
In case of a misreading MAF, or 2 misreading MAFs , one too high other too low, switching should have affected the original good LTFT as well...

I think it is safe to assume the original right MAF was underreading, causing the high LTFT, while the other was ok.
But after switching, the underreading MAF now has no effect on the fuel trim anymore, as that did not rise.

How about this: as the air intake goes through 2 MAFs, but ends up in one plenum, is it possible that the default opening of one throttle valve is off? Closed too much?
The good or maybe overreading MAF senses less air, so maybe a higher vacuum because of a more closed valve, that gets filled up by the other side air ? So fuel trim in the end is ok, even with a lower MAF value?
The underreading MAF lets in more air than reported, but loses that air to the other side because of the higher vacuum.
And in the end the underreading MAF accidentally is right as the non-measured surplus air is not used on that side?

 

Above a certain rpm, the actuator on the back of the manifold (vacuum operated) opens 6 butterfly valves inside the plenum, so the air can take a different route.
But I forgot if it is 6 or 12 valves...Or maybe the valves are normally open, and close above 3k rpm...
Could well be that the air is not mixed, I just thought I had found a possible explanation...should have made pictures when I had it apart...
But if it is not something like that, it gets even harder to explain the results of the MAF swapping...

About the plug, yes both have only 3 wires.
When I spray brake cleaner into the intake of the right side, all data on scanner(+Motronic) left would change , also the fuel trims ...

 

Thanks for clearing that up (12 valves etc) , so my theory of shared air may be the possible explanation.
About that Y-axis, yes, throttle position is the most logical.
Anyway, will test some more by swapping both MAF's for spare others, to see what happens.

About the plugs and wires, I will take a look later on, but I think the 3 wire end is plugged into the scanner

 

Ok, so that sends my theory to the bin I guess, no air mixing. I now remember seeing a baffle inside the top cover, preventing the mixing.

On the OBD plugs, yes I had to remove the 2 wire part to expose the 3 wire plugs into which I plugged the 3 pin to OBD2 converter cables, on both sides.

So what else could cause the strange behaviour of the combination of MAF + LTFT, where one specific MAF is underreading on one side, but not anymore when installed on the other side? Higher resistance on the wires to the ECU? Ian, remember that strange resistance on that side half a year ago, which we could not find the cause of?
Could that be hurting the values of that original MAF, and when swapped left to right, the MAF could finally supply the correct values ?
And the MAF that took his place , is overreading it a bit, which compensates for the extra resistance?

As there are no vacuum leaks anymore, injectors have been cleaned, fuel pressure and supply is ok, spark plugs are all equally well colored, compression is ok , and all within 10% on the entire engine, all sensors have been renewed (apart from cam phase), there is not much more to do other than replacing a Motronic, but I do not think that is necessary at all.
Hunting for the cause of that resistance again?
Or play with the position of the throttle valve on the right trying to balance it again? I guess that is the only thing left to do.
Will report back on this in a few days.

 

Haha, yes it is a lot of info, 8 months of trying to find all the faults.
And there were a lot of them...

I indeed sprayed brake cleaner into the intakes to enrichen the mixture, but that is with airfilters etc in place. So only fumes pass the MAF, no actual liquid solvent on the MAF itself. The MAF's I cleaned with special MAF cleaner, but no results. ( dirty MAF normally happens with the hot wire type, not the element, so no real surprise)

About that resistance, you are correct, at the time we dismissed it, but when all else is done, you get back to the whole thinking process, looking for the mistake somewhere...

At the moment the MAF's are out of balance, but the LTFT is acceptable. Maybe I better leave it this way, car drives good now.
I wonder how many cars drive around in similar circumstances, never knowing it is not 100% perfect...