Xantia V6 broken exhaust (update: and rough low rpm running)

[Mandrake]

Fair enough. I do agree that some cylinders are running too rich and have been for a long while. I'm not sure why but I have two theories which both involve the oxygen sensor control loop:

a) One (or maybe two if its coil related, given its a waste spark system) cylinders are misfiring, perhaps only under load, causing un-burnt oxygen to be detected by the oxygen sensor. The control loop then richens the mixture to "correct" the lean condition causing all the correctly firing cylinders to run rich. If this persists the long term fuel trim will be skewed positive and cause the mixture to be too rich even when in open loop for example during hard acceleration or during a cold start before the oxygen sensor warms up. The stumbling and poor performance I see could be a combination of misfiring on one/two cylinders and stumbling under an over rich mixture on the remaining cylinders.

I think this is my most likely scenario, especially given that resetting the ECU (thus clearing fuel trim memory) improves the running for a short period of time, until it "adapts" again.

2) There could be a leak on one intake runner causing one cylinder to run lean, again the oxygen sensor control loop will sense this and drive the overall mixture richer causing the other cylinders to run too rich. I don't think this is the problem, but having said that I did use a second hand gasket last time I refitted the intake manifold and there is a small possibility that this is the second time I've refitted the same gasket as I had a couple of second hand but good gaskets that I may have mixed up. I really should have used a new gasket and will next time I take the manifold out. A small vacuum leak shouldn't cause any problems at wide throttle though ?

[KP]

Have you thought about getting the injectors tested as well?

Also an ECU reset?

Sorry haven't read the whole thread

[Mandrake]

Have you thought about getting the injectors tested as well?

I've done an injector balance test myself a few months ago - all injectors were identically matched. Compressions are excellent and nearly identical too BTW.

Also an ECU reset?

Not sure what you mean by reset - fault codes can be cleared with a Lexia (there aren't any codes to clear) and learnt fuel trim and knock retard tables are wiped when the battery is disconnected for more than a few minutes. I'm not aware of any other "reset" that can be done on this particular ECU ?

Sorry haven't read the whole thread

Don't be sorry, that's a good few hours reading. I wouldn't wish that even on my enemies given the number of blind alleys that this thread went down.

[KP]

The battery disconnect method i was meaning to clear the flash learned memory on the ecu

[Mandrake]

The battery disconnect method i was meaning to clear the flash learned memory on the ecu

In that case its something that has been done dozens of times over the last few months, usually due to disconnecting the battery while doing other work, but sometimes to specifically clear the ECU. (Although I now usually unplug the ECU connector to do that to avoid disturbing all the other systems)

[Mandrake]

The coil pack and fuel pump from KP arrived today, and the coil pack went straight in after work.

Before fitting the coil pack the car was driven about 10 minutes and was its usual sluggish misfiring self, much the same as it had been for the last month or two.

I initially fitted the coil pack without disconnecting the battery or ECU as I didn't want to introduce the additional variable of resetting the ECU.

There was an immediate large increase in power and improvement in driveability, particularly noticeable at low rpm at small throttle openings where before it was flat and gutless, now it had quite good pick up.

At some rpm/load combinations it was hesitating a bit so I suspected that the ECU needed to be reset to clear out any erroneous fuel trim data that had been learnt while misfiring. (Eventually it would sort itself out but it could take a long time)

The gearbox was also exhibiting some "rpm flareup" on up shifts, which again I put down to the auto-adaptive system having adapted to the engine misfire which was no longer present. (The engine ECU would have been over-reporting the developed torque to the gearbox ECU)

I disconnected the engine ECU for 10 minutes to reset it, and did an auto-adaptive reset of the gearbox, then went for another drive.

Straight away the gear shifts were a lot smoother with no rpm flare up on upshifts, and the hesitation I'd been seeing was gone. I took it for an "Italian tuneup" to try to de-soot the spark plugs that would have been previously misfiring which seems to have been successful, as far as I can tell there is no misfire now.

Performance is not 100% - there does still appear to be what seems like a fuel starvation problem under wide open throttle, especially at lower rpm, but throttle responsiveness at light to medium throttle is excellent now with no hesitation or misfire as far as I can tell, even when fully warmed up.

Will it last ? Who knows.

Considering its 15 years old the second hand coil pack is in very good condition. There is a tiny surface crack in the middle coil frame and that's all. The old one of mine I removed has a massive crack right along the base of the cylinder 1/5 coil that I'm sure wasn't there before, along with large cracks on the other two coils as well.

The next step is to swap the fuel pump over which will hopefully resolve what I think is a minor fuel starvation issue. That will be several weeks before I get a chance to look at it though.

Assuming today's big improvement wasn't a fluke or an intermittent variation I've made some progress.

[Mandrake]

By the way, I've also discovered the cause of the "intermittent: engine speed sensor open circuit or short circuit to ve-" fault code that has been logged by the gearbox from time to time - it is caused by turning on the key when the engine ECU is unplugged. So in other words its a red herring and nothing to worry about

I noticed it when I had the engine ECU unplugged to reset the ECU, I turned the key on to wind the windows up and noticed the sport and snow lights for the gearbox were flashing, connected the Lexia and found that fault code there which wasn't there a few minutes earlier.

So obviously when the engine ECU is powered down or disconnected the gearbox ECU detects a fault on the speed sensor data line from the engine ECU - a good example of how unplugging stuff can set phantom fault codes!

At least I know what it is now...

[Stempy]

Fingers crossed then, hope the gearbox can take the new found horsepower

[Mandrake]

Fingers crossed then, hope the gearbox can take the new found horsepower

The thought did cross my mind It seems fine so far though, and I really thrashed it during the Italian tune-up...

Mind you the lumpy torque from an intermittent misfire at low rpm is probably hard on the torque converter clutch, as there was a small amount of clutch shudder when trying to pull away at low rpm in 2nd and 3rd before, that seems to have more or less gone now.

Performance this morning on the trip to the train station was excellent, lots of pull, very responsive on the throttle even when cold, and no lack of power below 3500 rpm when cold as was happening before. I did still notice it occasionally flatten out a bit after a few seconds of wide throttle as if it was leaning out, so I think I've made the right call with getting the fuel pump as well as the coil pack.

It would be just my luck to have two independent intermittent faults dogging me for months...

[Mandrake]

Ok, you all knew this was coming didn't you...

Nearly three days of uninterrupted excellent performance after swapping the coil pack, short trips, long trips, hot weather, cold weather, hot engine cold engine it made no difference, it was running really well, although perhaps with a bit of a wide throttle flat spot now and then.

Yesterday morning it was also ok, yesterday after work, terrible. I went for a 10 minute drive, almost immediately I could tell something was wrong. The initial 30% opening range of the throttle was still responsive but just nothing beyond that. And by that I do in fact mean nothing. 2nd gear up a hill starting at about 2000 rpm and trying to accelerate, anything past about 30% throttle just made absolutely no difference to the acceleration at all. I could push it down to 80% (not quite enough to kick down) and it wouldn't pull away any faster than about 30% throttle opening. Didn't really seem to matter whether I was in 2nd, 3rd or 4th.

There was no vibration or shudder, I couldn't feel it missing, it was smooth and quiet, in fact too quiet - none of the usual rawty engine note you get when accelerating, the engine note sounded subdued like you weren't accelerating even though you had the loud pedal well down... basically no symptoms at all except a complete lack of response to the throttle past a certain opening.

I turned around, drove back to where I left from, on the way back I noticed it was intermittently a lot better but under acceleration was "surging". (that's the only way I can describe it) Parked for 15 minutes engine off waiting to pick someone up, drove off again and the problem was mostly gone. Performance still wasn't what it should be but it wasn't totally flat like a pancake as it had been minutes before...

So what the heck is going on guys ? Surely this must be fuel starvation ? I can't believe I have another faulty coil pack, it has to be something else. This was the worst it has ever been, these are exactly the same symptoms I used to get occasionally on long hot weather trips where the performance would just fade away gradually over an hour or so leaving it feeling flat and unresponsive, (and then recover after being left a few hours) except this time it did it from cold.

I have my fuel pump to fit yet so I'm hoping I might get lucky there, but the symptoms this time were rather extreme so I'm not sure if even a dodgy fuel pump could explain it...again, I'm hampered in the diagnosis by lack of data. If I could monitor the fuel rail pressure and the exhaust back pressure in real time while driving the car while its symptomatic I could immediately rule out (or identify) two possible causes...its so intermittent and unpredictable that only by having instrumentation on the car while its playing up will it ever be found I think.

[addo]

MAP readings should reflect back pressure to a fair degree. With fuel supply pressure you need a gauge isolator for safety. These aren't too expensive.

[Mandrake]

A gauge isolator ?

[Hell Razor5543]

A gauge isolator is used to isolate the gauge from harmful media;

http://www.beccontrols.com/gauge-isolators.html" onclick="window.open(this.href);return false;

However, I would also go looking for failing solder joints. They (as you know) do cause intermittent problems, and sorting them will eliminate a probable cause of your problems.

[Mandrake]

A gauge isolator is used to isolate the gauge from harmful media;

http://www.beccontrols.com/gauge-isolators.html" onclick="window.open(this.href);return false;

Ah I get it. Some sort of diaphragm so that the gauge line only sees air pressure...

What would really be ideal would be an electronic pressure gauge with a remote sender - then you just have a small pressure transducer stub connected to the fuel rail, the rest of the run back to the gauge is just a pair of wires that can be routed under the corner of the bonnet and through the door jamb into the cabin for display - trouble is they seem to start at over £100 so are WAY too expensive for a test that may only take 10 minutes to confirm or disprove the fuel pressure drop hypothesis, and probably never be used again...

Edit: rather than a whole system including gauge I wonder if I could get just the pressure transducer by itself so I could run my own wiring and plug it into my multimeter for display ?

Something along these lines, although with the right fitting for the Schraeder valve:

http://www.ebay.com/itm/Pressure-transd ... 0799543734" onclick="window.open(this.href);return false;

However, I would also go looking for failing solder joints. They (as you know) do cause intermittent problems, and sorting them will eliminate a probable cause of your problems.

In the fuse boxes you mean ?

[Hell Razor5543]

Well, based on Dunks' experience, I would think that that would be a good place to start. All it takes is for there to be one bad joint on a rail that supplies several devices for those devices to go wobbly! Even with the best will in the world you cannot make a solder joint that can handle the punishment in an engine bay without it failing over a long period of time. IIRC your Xantia is 15 years old? If nothing else you will have eliminated a possible cause for some of your problems. You did also mention that an intermittent miss happened right after you worked on the engine bay fuse box. Coincidence? Doubtful.

I would still fit that fuel pump, and (if you haven't already done so) check the sender kit in the tank. I don't know if you get the same micro organisms in petrol that you get in diesel, but what harm can it do to check anyway.