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

[lexi]

Counting dealers I think I'm about owner number nine

You don't own that car though..........it owns you

[BW]

I'm a very regular reader of this and the other Xantia V6 blogs. I've had mine for 13 years now, and it's certainly had its moments over the years but just to even out the seemingly endless problems most seem to have, mine has not had even a tiny fault in the past year and recently sailed through its MOT with good emissions and no advisories. It has covered 130000 miles plus now, but only does about 5000 a year these days. (Touching lots of wood!)
Definitely worth persevering because potentially you have such a wonderful car. I never cease to be amazed at the technical prowess and knowledge of some on here.

[Mandrake]

Thanks BW. Sadly I don't think this one has been cared for like yours will have, if you owned it from new! I really do think a lot of its problems are due to neglect of standard maintenance like oil changes etc...

I don't think I'm dealing with just one fault that explains all symptoms either, I think there are three distinct issues.

I made some progress last weekend I think.

A few weeks ago I changed the fuel pressure regulator for a 3.5 bar one and although it helped with the "lean" issue I was still seeing it sometimes reading "lean" under load at low rpm. During that process when I refitted the spark plug leads on the top of the coil pack I put a good helping of silicone dielectric grease in the boots - big mistake in hindsight! (although there was already some in there before that too)

I noticed that performance got worse when I did that (although I didn't connect it to the grease at the time as I had the fuel rail out and changed the regulator as well) and since then it has seemingly had a bit of an intermittent misfire even at idle. I was noticing it quite badly last week where it would miss at least once a second at idle with the engine showing a slight shudder on its mounts.

So on a hunch last weekend I cleaned out all the silicone grease from the boots and terminals at the top of the coil pack - at least as well as I could with kitchen roll and a spray of carb clean. (not really ideal though - does anyone know of a cleaner that will dissolve/clean out silicone grease effectively ?)

I put the wires back in and instant big improvement! A lot more low down pull, and more importantly the random miss at idle now seems to be gone - the idle is nice and smooth now.

Not all of the symptoms are fixed - I can still tell the timing is getting retarded under load at low rpm when the engine is hot, but I believe that is an entirely separate issue.

So what happened ? Far too much grease applied such that it filled the gap between the mating contacts. Air has a breakdown voltage of about 3,000 volts per millimetre, silicone dielectric grease is approximately 20,000 volts per millimetre.

In most spark plug connectors the contacts are not under strong spring compression like a low voltage connector - so the spark actually jumps a small air gap of a fraction of a millimetre, probably between about 0.1-0.5mm, especially if the connectors don't mate well together. This is normally no cause for concern - after all older cars with distributors had the spark jumping a much larger air gap in the distributor as well!

If we pretend the contact gap is about 0.5mm then that's 1,500 volts lost if the spark has to jump through air, which is not significant when over 40,000 volts is available. But 0.5mm of grease is 10,000 volts, that's a BIG loss in spark voltage which will tend to make the engine misfire especially with a lean mixture or under load. Also, the tiniest movement of the spark plug boot (they don't actually fit very well and have quite a bit of "slop" between the rubber and the plastic elbow it pushes onto) will change the width of the gap between the contacts and thus greatly alter the voltage drop through the grease, whereas if it was an air gap between the contacts the change in voltage would be insignificant.

So you end up with a large reduction in spark voltage that has a tendency to change randomly with small movements of the boot/lead with vibration etc. Doh!

So the moral of the story is don't apply ANY grease inside the boot or plastic elbow or let it get onto the contacts - the only place you would want to apply grease on those top coil pack connectors is a thin wipe around the OUTSIDE of the elbow on the coil pack so that when you push the boot on it helps seal air tight between the rubber boot and the plastic elbow but WITHOUT getting between the metal contacts. Lesson learned.

I've also noticed that since then I'm not seeing any "lean" running at WOT at low rpm any more - but I don't know whether that was the tweaking we did to the camshaft timing or whether it was the ignition misfire which may have been causing un-burnt oxygen to pass through during a misfire event. Either way it's not happening any more! (albeit with slightly higher fuel pressure, I'm not sure whether it would read lean again if I changed the fuel pressure regulator back to the standard one)

The timing getting retarded under load at low rpm on a hot engine I believe is related to either low oil pressure (worn bearings combined with thin oil ?) and/or worn hydraulic lifters - I'm going to try some 10w-50 shortly to see if that works around the problem.

Whether the noisy lifters are directly affecting the engine performance on those cylinders and triggering knocking, (by affecting valve lift and duration of some valves) or whether they're just making enough mechanical clatter to trigger the knock sensor directly I don't know. But I have a hunch that if the thicker oil stops the lifter noise the timing retard will also go away. A year ago when I changed from 10w-40 to 5w-30 that could have been yet another example of two steps forward one step back! (Never put thinner oil into an old worn engine that is suspected of having neglected maintenance...)

Finally, I think the stumble when flooring the throttle at idle (or trying to stall tapping it lightly at idle) is a third problem unrelated to either of the above problems - and from what northern_mike said his does this too but doesn't have any other performance issues, which would tend to confirm its an unrelated issue.

I think this problem is related to the calibration of the idle screw and TPS/ICV - the throttle butterfly and ICV are both closed too much to get the idle speed down to normal perhaps due to a small undetectable vacuum leak. I may see if I can find a way to do a smoke test on the inlet to find a leak. If I'm right once that leak is fixed I should be able to increase the throttle plate idle screw slightly and the ECU will adapt the ICV opening as well and it won't stumble any more. From what I can remember this stumble off idle started a long time ago back when I first cleaned the dirty throttle plate - a cleaner throttle plate may have made a small vacuum leak more obvious as the dirty throttle plate would tend to compensate for a vacuum leak by reducing the (normal) leakage past the throttle plate when closed.

When you open the throttle just enough so that the ECU no longer registers the throttle in "idle" position, it then starts using the throttle position to calculate fuelling and under estimates the required fuel thus causing a lean stumble. I've already confirmed with a scope on the TPS and injector signal that the TPS is not faulty (although the calibration might be out due to the idle screw being set wrong) and that tapping the throttle lightly is causing the injector pulse width to drop and starve the engine of fuel (and that it doesn't happen with the TPS unplugged) - so something is a bit messed up in the relationship between throttle butterfly position, TPS, ICV etc.

Someone a long way back in the thread said something about worn throttle plate pivots that were causing a vacuum leak past the pivots - I wonder if something like that is enough to upset the balance at idle ? The fact remains that the idle OCR % is only 25-26% during a hot idle to reach the correct idle speed when it should be more like 30%, so extra air is getting in somewhere...

[Mandrake]

By the way, I came across a really excellent article on the inner workings and diagnosis of hydraulic valve lash adjusters (eg hydraulic tappets / lifters) here:

http://www.schaeffler.com/remotemedien/ ... omp_en.pdf" onclick="window.open(this.href);return false;

Very interesting reading for those that like to understand how these things work. It also covers variable valve timing actuator systems.

It would be interesting to know which of the many designs they describe is actually used in the ES9J4 - I assume it would be 3.1.2 - Hydraulic bucket tappet, as with twin overhead cams per bank I assume you're not going to have lifters or rocker arms. Anyone know ? The ES9J4 documents I have don't really go into that level of detail.

[RichardW]

Buckets

[Mandrake]

Thanks Richard, obvious in hindsight.

I assume part number 3 "Valve piston" is the hydraulic bucket. Yours for a mere £25 each from Citroen, so £600 for 24 of them.

I think I'll just try the thicker oil...

[Hell Razor5543]

I wonder how much these are?;

http://www.yoyoparts.com/oem/37545/citroen-094247.html" onclick="window.open(this.href);return false;

[Mandrake]

I wonder how much these are?;

I don't know, but realistically replacing any of them is only a mental exercise.

Apart from probable cost its a camshaft out job which I would only contemplate if I had the camshafts out to seal the bottom cam box seals.

If the change of oil doesn't do anything for them they'll just have to stay noisy and I'll cross my fingers that the engine still passes the emissions test come MOT time.

I have to say though, the improvement since cleaning all that silicone grease out of the spark plug lead boots has persisted all week, so it looks like despite all the multiple goings over the ignition system has had in the last year that the engine has still been experiencing intermittent ignition misfires. What's more a lower than normal available spark voltage would cause the engine to misfire more readily on a lean mixture than a rich mixture, so the better running with the rich mixture may have just been that the rich mixture was preventing an ignition misfire by not requiring as high a spark voltage.

And the "lean" reading on the O2 sensor may have simply been caused by an ignition misfire passing a lot of unburnt oxygen into the exhaust. I certainly haven't seen any lean readings on the O2 sensor since I cleaned the plug wires and I'm tempted to change the fuel pressure regulator back to normal to confirm the diagnosis.

It makes me want to thoroughly clean out any silicon grease that I had previously put in the spark plug boots as well - I did wipe the spark plug top electrodes clean of grease last time I had them out but getting the inside of the boot completely clean and free of silicone is difficult unless I could find a solvent for that kind of grease that wouldn't hurt the rubber - but from my quick searches it appears there is no good solvent for silicone grease and that manual cleaning is the only way to get it (mostly) off.

[CitroJim]

We didn't check for a broken woodruff key in one of the cams.... but then that would require rather more dismantling I suspect

Very unlikely as the hubs are a very tight almost interference fit on the cams anyway. They're a devil to shift even when you want to...

Counting dealers I think I'm about owner number nine

You don't own that car though..........it owns you

Many a true word spoken in jest!

[Mandrake]

Well I changed the oil on Saturday - no audible improvement in what I assume is tappet noise so far, however it does seem like the car is running quite a bit better on Sunday when we were doing a lot of driving around looking at houses. Overall mechanical noise from the engine especially when accelerating at higher RPM is a lot less. Power still fluctuates a bit during acceleration but I don't know if that's timing retard being applied, or perhaps the gearbox is getting ready to croak on me! The gearbox has done a couple of unusual shudders/jolts when slowing down on the overrun down hill in the last few weeks that have got me a bit worried...

Although the camshaft timing was proven correct I still think I'm right about some sort of valve train problem with this engine being the underlying cause of it's fluctuating performance and intermittent misfiring. Some time ago I read about the technique of diagnosis valve issues using a "vacuum waveform" - you can use specialised equipment to do so, but on a MAP engine like this you can connect a scope to the MAP sensor in AC coupled mode to look at the small (about 50mV P-P) pressure fluctuations that occur during the valve opening and closing cycles. I was recently reminded of this by the following article:

http://www.motor.com/magazine/pdfs/082007_09.pdf" onclick="window.open(this.href);return false;

Take particular note of Figure 1 that shows an engine with a misfire caused by a "tight" (not fully closing ?) valve on cylinder 5. (Third from the right) Each peak in the waveform is the point just before an intake valve opens - when it opens the pressure should drop (downwards slope) and reach a minimum before going up again. The minimum point (maximum vacuum) of all cylinders should be the same and there should be a general smooth downwards slope with only a small amount of ripple, if any. The exact shape will vary from engine to engine but it should be equal for all cylinders and roughly a sawtooth waveform. In Figure 1 in the article you can see a massive oscillation where it should be dropping steadily.

Here's what I measured with my scope:



From what I can determine, D and E are normal ideal intake waveforms with a downwards slope and not too much oscillation. F is also ok although there is a bit more oscillation than ideal.

A and B look quite suspect and look rather similar to cylinder 5 on Figure 1 of that article. You can see A and B repeating again on the right hand side and looking exactly the same. Those alone would suggest some sort of valve or valve train related issue.

However C I believe could be the biggest culprit - although I haven't seen any sample waveforms like this (there's not many published waveforms at all that I can find) it seems fairly obvious what's happening - the inlet valve is opening very late - that little horizontal section of line that shouldn't be there shows the pressure is holding steady when the valve should be open and the pressure should be dropping as the cylinder draws in air, then it suddenly starts to drop, but very late.

I think this could be a single hydraulic tappet on an inlet valve that has completely collapsed and is refusing to pump up at all, so you have maximum possible valve lash. Normal valve clearance on a non hydraulic tappet engine is only a fraction of a millimetre, I've read that hydraulic tappets have up to 2-3mm of mechanical travel / adjustment range, so in that case there would be a single inlet valve effectively running with about 2mm of clearance, hence opening very late, closing very early, and making one a hell of a racket.

Possible causes for A and B waveforms from what I've read include - broken valve spring or burnt valve, seized hydraulic tappet that is preventing the valve completely closing at certain engine temperatures, (effectively like a "tight valve" on an engine with manually adjusted tappets)

So which cylinders are A, B, and C ? I don't know - all I know is that those three are sequential in the firing order, however the firing order alternates back and forth between the two banks so they won't be adjacent cylinders physically.

Normally you would connect a second scope channel to Cylinder 1 spark or injector to synchronise the waveform - then you know the first one is always cylinder 1 and you can read off the cylinder numbers in firing order sequence and identify the specific cylinder.

I could do that with my scope but unfortunately this engine does not have a cam sensor, meaning both spark and injectors fire on pairs of cylinders, so I could identify each waveform down to a PAIR of cylinders, but not know which one of the pair it was.

A bit academic though really, if there is a valve fault that's a head off job and I won't be doing that. If its just a faulty tappet I could possibly change that just lifting the cam out but that's still a big job and would require reliable identification of the offending cylinder, and I don't know how to do this on an engine with no cam sensor. Still, it's nice to have some idea of what it might be even if I can't fix it.

The waveform above was taken before I changed the oil, and there has actually been some noticeable improvement to the waveform after the oil change, also the waveform was changing from cold to hot, showing the worst waveform when the engine is hot, which fits the symptoms. I guess in a few weeks we'll see if the oil does make a real difference or not. I'm not holding my breath, but at least it has received its scheduled oil and filter change! It's been just over a year and 6000 miles since the last one.

[lexi]

There ain't much wrong with a car when you can do 6k in a year with it.

however it does seem like the car is running quite a bit better on Sunday

How many times have we heard that one? There is nowt wrong with your imagination

Have you had a shot of another V6 to compare? Have you given another V6 owner a go to see what they make of it?

Are you sure you don't just suffer from an affliction of constantly talking about your bloody car along with a tiny bit of OCD?

[Mandrake]

How many times have we heard that one? There is nowt wrong with your imagination

It's not my imagination Alex, I'm just dealing with a cranky intermittent problem...

Have you had a shot of another V6 to compare? Have you given another V6 owner a go to see what they make of it?

Nope, I've not had a chance to drive another V6, nor get another V6 owner to take this one for a drive. Nobody seems willing - the problem is either way there is no insurance cover while driving...

Are you sure you don't just suffer from an affliction of constantly talking about your bloody car along with a tiny bit of OCD?

Believe me, I'd love nothing better than for the thing to bloody well work properly so I can enjoy driving it and stop working on it, and give the forum some peace and quiet.

I'm pretty sure now that I'll never be able to fix it though, especially if its valve problems. Also it feels like the gearbox is on the way out for real this time. I'm just waiting for that to go bang next.

So I'm going to try to scrape through the MOT next month, use it to help us find and move into a house then get rid of it or strip it down for parts if I get another one.

[lexi]

I was just testing you

[Mandrake]

I was just testing you

I know, but contrary to popular opinion I'm very weary of this car and its foibles now, I'm now purely trying to get through the MOT to keep it on the road for just a bit longer, until its convenient to deal with replacing it, whether its running properly or not. Games a bogy, as they say up here.

[lexi]

There is a free roof rack for V6 here..........might be handy for removals.
The bangornomics philosophy is good. Drive till it drops.......it might do another 6k