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

[Mandrake]

I'm confused - the air intake temperature sensor is on the opposite side of the engine to the timing belt, so how could they damage it while doing the timing belt ?

Its in the back of the air filter box, on the right hand side looking into the engine bay. At least it is on the Xantia.

I was going to reply to your previous post and suggest that your cutting out was a separate issue, looks like it was!

[larppaxyz]

I'm confused - the air intake temperature sensor is on the opposite side of the engine to the timing belt, so how could they damage it while doing the timing belt ?

Its in the back of the air filter box, on the right hand side looking into the engine bay. At least it is on the Xantia.

I was going to reply to your previous post and suggest that your cutting out was a separate issue, looks like it was!

Sorry, i edited my message. I was talking about intake manifold pressure sensor.

[Mandrake]

Sorry, i edited my message. I was talking about intake manifold pressure sensor.

Ah, that makes much more sense. Yes a faulty or unplugged MAP sensor will cause stalling, especially on a cold engine for some reason.

When the engine is hot it will usually run without the MAP sensor, but only if the throttle butterfly is clean, and TPS and oxygen sensor are working properly, then the ECU can adapt and synthesise the lost data to allow the engine to run, but the throttle response can be jerky and there will be a tendency to stall and difficulty starting.

The MAP sensor you just replaced which has failed, where did you buy it and for how much ?

Many months ago I tried to replace my MAP sensor with a £20 pattern part (not Bosch) one from eBay - the sensor that arrived looked poor in build quality and was dead on arrival ! The engine would not even start with it connected. I sent it back and got a genuine Bosch sensor for about £65 and that worked perfectly, however I discovered that there was nothing at all wrong with my original factory fitted sensor, so I now have a spare...

Could it just be that your MAP sensor has failed by itself due to being poor quality ? If your original sensor is the factory fitted Bosch one and its working fine now you've fitted it back in there was probably never anything wrong with it, just like mine...

To test it measure the manifold pressure with the Lexia in parameters measurement - with the key turned on and engine not running it should read local ambient pressure in millibars, eg around 1000 millibars.

Idling on a hot engine it should read between 350 and 450 millibars depending on whether there is an accessory load. If it passes both of those tests its ok.

[larppaxyz]

When the engine is hot it will usually run without the MAP sensor, but only if the throttle butterfly is clean, and TPS and oxygen sensor are working properly, then the ECU can adapt and synthesise the lost data to allow the engine to run, but the throttle response can be jerky and there will be a tendency to stall and difficulty starting.

But disconnecting sensor while engine is running, always makes engine cut out. I have tested this before It could be possible to start it without sensor, i haven't tested that.

The MAP sensor you just replaced which has failed, where did you buy it and for how much ?

I'm pretty sure it's problem with wiring or connector, not with sensor. However, it's a no-name part, cost me around 30 euros few years ago. I didn't make any difference, so i kept old sensor in my glovebox knowing it's in working condition. I now swapped it back, because it was easy and i wanted to get back home.

Could it just be that your MAP sensor has failed by itself due to being poor quality ? If your original sensor is the factory fitted Bosch one and its working fine now you've fitted it back in there was probably never anything wrong with it, just like mine...

It's possible, but it just would be pretty amazing that it failed just when they disconnected it from connector. Now that i think about it, it's possible it was damaged by physical force.

To test it measure the manifold pressure with the Lexia in parameters measurement - with the key turned on and engine not running it should read local ambient pressure in millibars, eg around 1000 millibars.

Idling on a hot engine it should read between 350 and 450 millibars depending on whether there is an accessory load. If it passes both of those tests its ok.

I'm pretty sure that if it gives any values, they are fine, but when i twist it or move it, it shorts out. So could be wiring or internal damage. I will take a look at it after i'm sure there are no more problems with my car.

So when are you going to clean your ICV?

[Mandrake]

But disconnecting sensor while engine is running, always makes engine cut out. I have tested this before It could be possible to start it without sensor, i haven't tested that.

Yes, disconnecting the MAP sensor on a running engine will instantly stall it, however leave it disconnected and try to start the engine a few times and it will start on the 3rd or 4th attempt (might require a little bit of throttle too) once the ECU has set a fault code for it, (which takes a few attempts at starting) it will ignore the MAP sensor and rely entirely on the TPS sensor and engine rpm for the base injection calculation, helped by feedback from the oxygen sensor once that warms up.

I know because I ran my car with no MAP sensor for a couple of days.

So when are you going to clean your ICV?

I've cleaned it today (but it wasn't very dirty and already turned freely) and done some other work as well like spark plugs but the end result is disappointing - no real change, maybe even slightly worse. I'm too tired tonight to write a detailed post but I will post more details of what I did tomorrow including some pictures.

[Mandrake]

As promised here is an update on the ICV cleaning and other work that I've done in the last few days. The bad news is that the "invisible caravan" is back again.

As previously described, last weekend I blanked off the spigot on the lower butterfly that normally goes to the charcoal canister purge solenoid, the performance of the car immediately jumped up dramatically and stayed that way consistently for four days, so I thought I was onto something - a problem with the mixture regulation, possibly due to a vacuum leak near the charcoal canister.

After reading the thread in the 406coupe forum and the success larppaxyz seemed to have cleaning his ICV I realised the problem could be that the ICV was sticky and not able to compensate for changes in vacuum when the purge solenoid opens.

I couldn't remove the ICV without also removing the manifold, (at least not until the hose clips are replaced with jubilee clips) I wanted to check the condition of the spark plugs (the previous plugs fouled badly) and also wanted to replace the inlet manifold gasket - which had been reused twice and may not have been sealing perfectly. So out came the manifold.

First the good news. This time the condition of the spark plugs was excellent, the insulator tips were white and clean with no evidence of fouling or carbon build up whatsoever:

Rear bank:

Front bank:


They've only done a bit over 1000 miles which isn't much but those following this thread will remember the previous plugs ended up looking like this after around the same mileage:

Rear bank:

Front bank:


Massive difference! The only two things that have changed between then and now is the type of plugs, and the coil pack was replaced, although the coil pack was only replaced well into half the mileage that the plugs did.

The previous plugs which fouled were Bosch FR7DC+ the ones that came out clean were Bosch FR8KDC. I guess it really goes to show that plug specs DO matter. Compared to the ones that fouled, the FR8KDC (which is the factory spec) are a heat range hotter, have dual electrodes versus single, have copper core (vs yttrium) and are resistive vs not resistive.

I speculated previously that the colder plug was the main reason for the carbon fouling, I think these photos confirm that ? It's possible that an ignition misfire was contributing too, and could explain why one or two cylinders in particular were much worse than the rest, but even the cleanest ones still showed signs of carbon fouling while all 6 plugs this time were absolutely spotless despite half the plug mileage occurring with the faulty coil pack still in place. (Perhaps they fouled a little bit but burnt it off again when the coil pack was replaced)

Because I'd gone to the trouble of getting some new plugs before lifting the manifold I swapped them over anyway as they're the same type. One thing I did notice however is a prominent "corona ring" on the number one spark plug (near right looking in the engine bay) which was not present on any of the others. You can see this below on the left compared to one of the others:



I've been suspicious of the strength of the spark on cylinder one before and possible leakage in the boot, when I inspected the boots on the coil pack I discovered that the boot on the cylinder one spark plug was swollen at the end causing it to be a very loose fit on the plug, also the inside of the boot was not clean but had oil contamination that looked a bit like a nicotine stain. It's not easy to see in the picture but the one on the right has nearly a millimetre bigger hole due to swelling:



So I swapped the boot with a spare one I had off one of my spare coil packs. As I've noticed problems with dielectric leakage in damp conditions before I thought I'd give some Silicon "dielectric grease" a try in the spark plug boots, so all the spark plug boots got a small wipe of Dow Corning DC4 silicon grease in the hole and also on the spark plug insulator which is supposedly designed for the purpose.

I then moved onto the ICV - after removing it I found that it wasn't sticky or jammed in any way. It did have quite a bit of carbon build up though, so I cleaned it thoroughly with carb cleaner and then lubricated it with LPS 1 and refitted it. The valve inside rotates freely when you swing the ICV in your hand and clicks off the end stops.

I then refitted the manifold etc with a brand new gasket. As I'd had the charcoal canister purge line blanked off previously I decided to leave it blanked off so as not to introduce too many unknown variables.

Upon refitting everything I went for a test drive - disappointment! The much improved performance that I'd had for the 4 days leading up until then was gone. It wasn't horrible but the "urgency" and responsiveness it had was gone leaving it feeling a bit flat and lifeless. I drove the car between Wednesday and Today and it didn't vary much but just stayed mediocre.

The question is have I done something wrong, have I disturbed something, or is it just the intermittent unknown fault rearing its ugly head again ? I was as careful and thorough as possible but taking the manifold out disturbs a LOT of wires, pipes, hoses etc...chances are I've just disturbed the mystery fault again...

Today I thought perhaps the boot I swapped for cylinder one was faulty (all the rest of the boots were left alone) so I swapped the boot again today for yet another one and went for another drive - no real change.

I then did my oil and filter change that had been planned for a while. For a short time after the oil change it ran better but after only a few minutes it was running worse than before the oil change. After driving into Glasgow and back today the car has gone into one of its lethargic moods where it really does feel like it has an invisible caravan attached. No get up and go at all really unless revved over 3500.

I can't even tell for sure whether it's misfiring under load or not. The idle is sometimes very smooth but at other times especially when hot its dropping out slightly as its idling as if its misfiring at idle, but I can't see any reason for it to do so...

The idle speed is still sometimes surging then trying to stall multiple times after a hot start and also dips and surges when pulling to a stop as well.

So despite my best efforts being very careful and thorough in my work, finding that the spark plugs were in perfect condition, etc the car is now running much worse than it was on Wednesday and I once again have no idea what is going on or even what direction to head in. Every thing I do seems to make it worse for no logical reason.

[addo]

Have you dummied up the coils to test each plug and lead? You only need to make a ground tether with crimp eyes and affix it to both the coil under test and engine ground.

Has the battery terminal been replaced yet?

[Mandrake]

Have you dummied up the coils to test each plug and lead? You only need to make a ground tether with crimp eyes and affix it to both the coil under test and engine ground.

I can't quite visualise what you're suggesting ?

The problem with any static test of the spark (manually firing the spark with the Lexia and watching it jump an improvised gap) is that its never shown any problems, even when the car has been running poorly. Last time I did this test on the coil pack that I now consider to be faulty all 6 outputs of the coil pack could produce a spark that could jump in excess of 20mm, which is over 60,000 volts.

I don't have a complete loss of spark, at most I might have a weak spark that misfires under throttle load, but even that I'm not sure of now after seeing how spotlessly clean the insulator tips on the plugs were - that suggests that the combustion is successful most of the time otherwise they would foul.

Has the battery terminal been replaced yet?

No, because I can't find anything suitable to replace it which won't be a backwards step from what I have now... Nothing that's available is a good fit, I've already bought two different clamps that I'm not happy with.

One thing I am going to try is driving the car with the ICV disconnected - when the engine is warm unplugging it actually increases the idle speed to about 1000, so it won't stall. (it's having to close almost right down at 27% OCR reading to regulate the idle, and the mechanical default of the valve when unpowered seems to be about 40%)

The reason I want to try this is the discovery with the Lexia that the ICV valve doesn't stay put during driving - for some reason the ECU opens the ICV almost wide open (in excess of 80%) when you accelerate then closes it again when you ease off. Given the small size of the ICV valve compared to the main throttle plate it makes no sense why it would provide additional air flow with the ICV when accelerating, so I'm curious to see what effect it has if this is not happening.

[Mandrake]

Right, a bit more tinkering today.

First thing I did was check for faults on the Lexia - our old friend "Intermittent: Oxygen sensor open circuit" was back again...

I cleared that and later on during driving the oxygen sensor voltage was showing perfectly normal switching so it seems like the slow warm up time of this oxygen sensor is what sometimes trips this fault code - it behaves like the heater in the oxygen sensor isn't working with a very slow warm up when idling. (but the heater measures ok )

My guess is the ECU expects to see switching within a preset amount of time after starting and if it doesn't see it this fault is logged. If I drive off soon after starting the exhaust heat gets it up to temperature quickly and switching so everything is fine, however if I cold start then just let it idle for several minutes (which I did yesterday) it may trip this code.

I'm still in two minds about whether I need to go to all the hassle of swapping the oxygen sensor or not as I'll have to take it out, send it back for warranty replacement, be without any oxygen sensor for a week, and then end up fitting the exact same model of sensor again, when the problem could be that its not quite the right model sensor for the car. (The plug didn't fit without a small modification)

The next thing I did was to unplug the ICV connector - the car starts perfectly when cold with it unplugged as it seems the "rest" position of the valve when not energised is about 40% which is where it normally sits at idle with a cold engine anyway. Since the ICV is opened fully during acceleration the idea was to see what difference preventing this made, and the answer is pretty much nothing, performance was still very flat and poor. When the engine was warmed up it was idling at about 1000 rpm with it unplugged, but apart from that there was no noticeable change to when it was plugged in, except that it wasn't dipping and then surging when pulling to a stop as the idle speed was higher than normal and not controllable by the ECU.

Looking at some of the Lexia data during a test drive I noticed a couple of things - one is the maximum injection time under wide throttle doesn't seem particularly high to me - about 8ms, which isn't that much when you consider idle is 2.4ms, I'm sure I've seen figures as high as 12ms before but even under full throttle I was only seeing 8ms.

The other thing I noticed is what MIGHT be knock retard of around 3-5 degrees under acceleration. I've talked before about the three different timing figures shown on the Lexia - "base advance" which I guess is the baseline figure taken from the timing maps based on rpm and engine load - torque reduction, which is a figure that is subtracted from this either during idle (timing based fine idle correction) or during a gear shift, and "advance" which is the final advance figure that is actually applied.

What I noticed for the first time today is that while driving under load the advance figure was a few degrees lower than the base advance figure even though the torque reduction figure was sitting on zero.

So for example at 50% throttle at 2000 rpm there was a base advance figure of 24 degrees, torque reduction of zero, but actual advance was only 19 degrees, with the difference presumably a result of applied knock retard...

This difference wasn't always there either, it came and went under different conditions. Also it seemed to me that the timing advance was rather low - most of the time in the 2000-4000 rpm range with anything above about 20% throttle the advance was on average about 20 degrees, and sometimes it was dropping as low as 16 degrees. Does this strike anyone as unusually low advance ? When it was showing 16 the base advance was 21 degrees, so something (knock retard ?) was applying an extra 5 degrees of retard.

If that 5 degrees of additional retard shouldn't be there (on 99 fuel at least) that would explain quite a bit as a 5 degree retard will cause a surprisingly big power reduction.

The next thing I did was disconnect both battery terminals and clean/check them, I also cleaned and checked all the terminals/bolts on the earthing plate beside the battery then reconnected everything. I then started it (already warmed up engine) and went for a drive.

Not surprisingly there was a massive improvement in performance, all of the performance I lost on Wednesday when I did the plugs was back again. I wouldn't say it was perfect at low rpm but in sport mode it was back to being very responsive and zippy.

I even took it on a short blast on the motorway and with a couple of minor hesitations aside it was flying and seemed to remain good through the whole drive.

Before anyone jumps to conclusions and says "Aha! Poor battery connections!" I really don't think it is. Based on repeatedly experiencing the same jump in performance when the battery is disconnected I really do believe that we're just back to the "good performance for a while after a battery off ECU reset" that has plagued me for months. (Unplugging the ECU instead of removing a battery terminal has the exact same effect without disturbing the battery terminals...)

Something happens after a while of running under some as yet unknown conditions that causes the ECU to learn "bad data" which starts messing up the adaption tables (fuel trim, knock retard etc, not sure which one) and causes the performance to drop. A reset clears this data and it runs well again for a while which can be between minutes, hours, days, or even a week or two.

Either that or maybe the ECU is faulty. Perhaps I do need to try the ECU swap after all. At least I've probably ruled out the work I've done over the last few days - there can't be anything wrong with the plugs or plug leads, coil pack etc if an ECU reset restores almost normal performance for a while!

[lexi]

An ECU swop would make good reading and possibly a first on a Forum V6 .........providing you get get one for 50 quid

[Mandrake]

I wouldn't be buying one just on a hunch Alex, too many parts have already found their way into this car on hunches and sure bets! However I may be able to borrow one for a test...

If it wasn't for the damn immobiliser coding to the CPH a swap would be a 2 minute job. As it is the CPH has to be swapped too, which is not as easy to get at.

[Northern_Mike]

Right, a bit more tinkering today.


Before anyone jumps to conclusions and says "Aha! Poor battery connections!" I really don't think it is.

Why don't you replace them and see? It's the cheapest, easiest thing you could do, yet you persist in avoiding doing it. I don't understand..

[Mandrake]

Right, a bit more tinkering today.


Before anyone jumps to conclusions and says "Aha! Poor battery connections!" I really don't think it is.

Why don't you replace them and see? It's the cheapest, easiest thing you could do, yet you persist in avoiding doing it. I don't understand..

Mike its nothing to do with trying to avoid it, I simply haven't been able to find any that fit properly, as I've said until I'm blue in the face. I've already spent money on two different styles of terminals recommended on here that IMHO are duds and are not suitable.

Once I cut the old terminal off there's no turning back, if the new one doesn't fit and make a snug reliable connection I'm up s**t street without a shovel. Having only one car I have to think carefully about any surgery that runs the risk of immobilising the car. If I had another car I'd be a lot more willing to "wing it" in matters like these.

There's also the fact that I have no evidence of a poor connection at the battery terminals, I've done voltage drop tests with a running engine between the battery negative post (not terminal) and the engine and body multiple times with NO evidence of any poor connection at all. Any resistance that was high enough to affect a running engine would cause serious problems with cranking, and I have NO problems with cranking speeds, nor have I ever had any.

If I could get the correct terminal and cable that I knew for sure would fit, and cheaply, I would do so. However I'm not willing to throw money at a new terminal that may or may not fit, may or may not help, when the evidence so far says its not the cause of the problem.

I can assure you that unplugging the ECU by its own connector instead of removing a battery terminal has the exact same ECU reset and temporary performance improvement effect, I've done it both ways so many times now I know this for a fact.

[Mandrake]

Hmm. I've noticed over the last week, and can't remember for sure whether it was happening before the last couple of weeks worth of tinkering and changes, that I'm getting an intermittent stumble when taking off.

It will be idling perfectly smoothly, seemingly on all 6 cylinders, I'll put my foot down quickly at a junction and I have about a 20% chance of it "stumbling" before it pulls away, eg it coughs slightly and the engine revs dip momentarily before recovering.

If I didn't know about all the work that had recently been done to the car I would immediately point the finger at an under load misfire such as a weak spark. EG - when you open the throttle suddenly the combustion pressures increase, the required spark voltage goes up and the cylinder with the weak spark fails to spark at the plug and the cylinder misses. If by chance the cylinder that misfires is the next one due to fire in the firing order at the instant you press the accelerator you get an ugly stumble where the revs drop before the following cylinder files and recovers the situation.

If I snap the throttle wide open quickly at idle it sometimes stumbles momentarily even not under load and it has done that on and off for a LONG time now, regardless of the various work I've done that throttle snap stumble has never completely gone away.

Anyone had this stumbling problem before ? Everything points towards ignition problems IMHO yet we have new plugs, spark plug leads, different coil pack etc... what else can it be ? Weak coil primary current from the ECU ?

Oh, and the sudden drop in power below 3500 rpm when cold is back again too. So when accelerating up a hill cold in first gear it struggles a bit lacking in power up to 3500 rpm and then suddenly roars off at full tilt when it hits 3500. This has been intermittently coming and going for months now too and I suspect its related to the overall power loss and stumbling.

I'm guessing everyone is all out of ideas, except scrap the car ?

The only thing I can think of is measuring the coil primary current ramp and injector current ramp waveforms - but that requires a decent quality current clamp probe which are on the wrong side of £80-£100. Something that would be nice to have but would probably only be used once or twice...

[addo]

Well, ECU etc swap if you don't want to try the battery terminals first. But I would 100% change them myself.

Another thing to try is pulling the fan relays (without a battery disconnection event) as this will reveal current supply inadequacies. Don't boil it though.