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

[addo]

Much as a course like automotive mechanics may turn out some quotient of dunderheads, there is also an attempt made over the three years to instill understanding of a wide variety of concepts. The lack of enforced professional development in most countries, seems to be the reason qualified mechanics go stale and retreat to stubborn/circular patterns of logic and repairs (eg, the swap and pray). Lack of an aptitude to learn and adapt, with the customer picking up the tab for guesswork and poor practice. Twice this January past in my roving diagnostics capacity, I've had mechanics baulk at the cost of replacing the actual problem item, because they are worried the customer will flip - having already been pinged for a handful of unneeded sensors and hours wasted in labour.

Which brings us to CAS units. Manufacturing tolerances - even with reputable aftermarket - are such that there is nil adjustability required in these units, the only adjustable ones come in aftermarket kits where you may be fixing a bracket of unknown offset. If it's not OC, is fitted to a clean spotfacing, the slotted wheel is not damaged (some fools will try to lock through the "windows" with a screwdriver when breaking the crank pulley bolt loose) then it is going to produce a signal within the range of amplitudes an ECU can readily work with.

[Mandrake]

Interesting post on the state of the industry addo, but I'm not sure how it helps the problem at hand. If you're saying don't bother to test the crank sensor signal because there is no way that it can be misaligned then I'm afraid I disagree.

I'm quite familiar with Hall effect sensors (they're used in a lot more than just engines...) so I know how finicky they can be. As the saying goes, when you have elimiated the likely causes you must then start looking at the unlikely causes, and this car has not exactly had easy run of the mill problems so far.

With coil pack, plug leads, plugs and compressions all ruled out less likely causes need to be followed up. What would you suggest to check next if there is still an intermitent misfire that comes and goes ?

If I could get my scope easily onto the crank sensor signal where it goes into the ECU I would do so in a heart beat, but from what I can remember of the connector it doesn't lend itself to back probing...

[CitroJim]

The chances of it being the crank sensor are remote Simon. Generally, they're a very reliable component and I've yet to know of one failing. The V6 one is not special - it's the same as used right throughout the range.

It's mounted very solidly in a boss on the gearbox and it's not going to suffer misalignment.

Yes, it's hard to 'scope it. If you do then really what you need is an ECU breakout box or a special adaptor for the crank sensor.

Do you know if the gearbox has ever been off your car? I ask as the crank sensor does not read the flywheel as in a manual car as an auto does not have one - instead it has a thin and easily distorted flex-plate with inverted dragons teeth around it's dished periphery. I can see if the gearbox has ever been removed carelessly then it could perhaps be a little distorted but again I can't see that being the cause as it's can't distort too far before it interferes with the torque converter.

Sadly, apart from whipping off the gearbox there is no easy way to see...

Simon, tell me. How does the engine start? And how does it run when stone cold? Does it start instantly and cleanly and then idle smoothly or is it hard to start and lumpy when cold?

Did you ever replace the Coolant Temperature Sensor? Sorry if I have missed that you did...

[Mandrake]

Hi Jim, no idea if the gearbox has been off before.

How does it start ? Again, it is intermitent. It always starts easily, however sometimes it will seem to be misfiring and lumpy for the first 30 seconds or so, other times it will be fine. Whatever is wrong is wildly intermitent.

No I haven't done the coolant temperature sensor, I have one but I don't plan to fit it as I'm trying to avoid any further "swap and pray" when there is no evidence at all that its faulty. Its been tested numerous times on the Lexia under various engine conditions and has never shown the slightest sign of trouble. While it might affect starting I can't see that it would be causing a huge loss of power hot or cold that I'm seeing...

Can you describe which side of the gearbox the crank sensor is and where the wire runs ?

[addo]

I have here a CAS that one champion left rubbing on a cam pulley - it chafed to the bare wires but still runs. The one fitted to Xantias etc really is a tough unit. Dead crank sensors seen this year - one Clio Sport (date of failure unknown as car had been stood several years not running) and a 2.0HDi which was intermittent - metered OC but did not always give an incoherent signal.

The crank angle sensor is normally at 1200H on the motor as fitted, or to the block's vertical axis. It's held on by a short 10mm head bolt to its integrated mounting ear, in a bore about 19mm diameter that sits on a small spotfacing. The location/access conforms to André's First Rule.

The CAS can be scoped with a breakout adaptor made of three pin Bosch plugs, these are found on many locations but if you can't get maroon ones you need to crack away the housings so the indexing guidepins don't clash. If you can get near a scrapper the brake pedal plug/socket are the same indexing.

A simple test for whether the box has been out - look for two sets of witness marks where the serrated washers go against its casting for ground wires. A not-small number of cars end up at a mechanic's workshop after the owner has had a red-hot go themselves and made no progress at fixing - yet vanity seems to dictate they deny all such escapades. The forensics are simple, and it's just more frustration if they don't tell you what has been played with as you may otherwise miss an additionally damaged item.

[DickieG]

Here's a thought that I don't think has been raised in the search for a solution, IIRC the valves are self adjusting as in its use engine oil to hydraulically adjust the valve gaps. Could it be that one or more of them is unable to maintain pressure especially when left for a certain period of time hence the random misfire when starting and also when running, the diagnostic capability of a Xantia is very unlikely to be able to spot such an event and having experienced hydraulic valve failure on other engines the symptoms are very similar, the only obvious tell tale to point in that direction was was a slightly noisy tappet, however on a V6 with 24 valves and the general cacophony that they generate with six injectors clacking away could easily disguise such a fault.

[CitroJim]

That's a good thought Richard We’ve cantered down that road with another V6 in the past with similar issues - a misfire when cold - but that turned out to be the ignition coil...

Anyway, yes, if the tappet was flat then the valve would not open fully or hardly at all... Should hear it as there's a world of difference between injector tick and valve clatter; the latter being much more metallic.

And that's something that wouldn't show up on a compression test.

Simon, the Crank sensor is down the back flank of the bellhousing and not a million miles from the diff. You should be able to see a flat brown three way connector somewhere under the throttle-body/plumber's nightmare...

The sensor is functionally inaccessible. You can barely see it, let alone get to it.

[Mandrake]

That's a good thought. As you say, its very hard to tell if one of the hydraulic tappets is misbehaving as they're all fairly noisy in the first place...there really is a cacophony of clicking and tapping...

However, it would manifest itself as low compression wouldn't it ? On the day I measured the compressions I went for a test drive both before (to warm up the engine to get a valid reading) and after doing the spark plug leads and checking the compressions and the performance was poor in both the before and after drive with the compressions measured in between and found to be perfecty ok. So I don't think hydraulic tappets can be the cause of the general loss of power that I'm seeing, although I see no practical way for me to prove it conclusively one way or the other either.

It could be the cause of what sounds like a misfire immediately after a cold start though - if one valve is not closing properly until the oil pressure builds up it could cause a misfire due to low compression that goes away after a short time. Again, I don't see an easy way to prove it.

I need to double check the basics, perhaps it would be worth my while getting 3 cheap in-line spark testers (they can be had for a few quid each) and watching the three of them carefully during the first start of the day on a cold engine to see if there is actually any loss of spark or not. If there is a loss of spark on one or more cylinder pairs its unlikely to be the tappets, if anything a leaky valve would make life easier for the spark as the pressure would be lower than normal.

Although getting my scope onto the crank sensor signal might be very difficult I should be able to get a vampire connection onto the three control wires that enter the coil pack on the left hand side (or make a breakout connector) to look at the drive waveforms from the ECU to the coils. That will give me an idea whether there is any correlation between any missed spark and the drive to the coils.

There's a few avenues of investigation for me to follow anyway.

[Mandrake]

Simon, the Crank sensor is down the back flank of the bellhousing and not a million miles from the diff. You should be able to see a flat brown three way connector somewhere under the throttle-body/plumber's nightmare...

The sensor is functionally inaccessible. You can barely see it, let alone get to it.

Does having long skinny arms and long skinny/bendy fingers help ? I've managed to get at bits on cars in the past that looked damned near impossible to reach on first inspection...

Pity I didn't look for it when I had all the throttle body and inlet manifold off. There is much better access down the back of the engine from the top with that off.

[DickieG]

That's a good thought. As you say, its very hard to tell if one of the hydraulic tappets is misbehaving as they're all fairly noisy in the first place...there really is a cacophony of clicking and tapping...

However, it would manifest itself as low compression wouldn't it ? On the day I measured the compressions I went for a test drive both before (to warm up the engine to get a valid reading) and after doing the spark plug leads and checking the compressions and the performance was poor in both the before and after drive with the compressions measured in between and found to be perfecty ok. So I don't think hydraulic tappets can be the cause of the general loss of power that I'm seeing, although I see no practical way for me to prove it conclusively one way or the other either.

It could be the cause of what sounds like a misfire immediately after a cold start though - if one valve is not closing properly until the oil pressure builds up it could cause a misfire due to low compression that goes away after a short time.

If a hydraulic tappet is malfunctioning then from my experience they don't fully open the valve so you won't really see low compression as the valve will still close normally, of course whether the relevant faulty tappet is on an exhaust or inlet valve may have an effect on overall compression but as the engine isn't firing you won't really be able to replicate or observe low compression on an engine with this fault because when running the combustion chamber has a mass of expanding gases that must be moved in/out whereas on a simple compression test the air within the chamber can remain still with no ill effects to the reading you get on the compression gauge.

When I had this fault on a CVH engine the fault was intermittent and often much worse at low revs or when starting after the engine had been left for a while such as overnight which presumably allowed time for the tappet to drain.

The only sure way to test for this fault will be using feeler gauges on the rockers and look for one or more with a larger gap than specifications.

[addo]

I'd be surprised if they have rockers on a quad cam setup like this.

[CitroJim]

Nope, you're right Adam no rockers in the ES9. The cams work directly on inverted buckets with the hydraulic tappets acting as shims effectively between the valve stem and the bucket.

Simon, as Richard says, a compression test would not show it up...

I'm worried that you say there's a real cacophony of sounds. All you should hear is a bit of injector tick. The top end should otherwise be as quiet as a mouse. If it's noisy then there's something amiss...

All three V6s I've had have been whisper quiet at the top end and two were very high mileage... Only occasionally can I ever hear a noisy tappet and that's invariably just after a hot start with the idle very slow as the oil pressure is re-established in the tappets.

[addo]

A faulty lifter often manifests on a compression test, as noticeably slower pressure climb than "good" lifter cylinders. It's the sort of additional detail that's taught - you watch the needle's behaviour as it climbs on each pot.

[Mandrake]

addo:

I did watch how quickly the pressures climbed and did not notice any obvious difference between the cylinders, although doing it by myself without remote starter operation meant I wasn't able to time it or anything. From memory it took about 3 pulses to reach 200 psi, and maybe another 3 to reach the final value but I wasn't counting or being ultra critical about it. A slow rate of climb would only finger an inlet valve though correct ? An exhaust valve wouldn't affect the reading much as long as it sealed when closed ?

Jim:

I'm starting to get swamped with a lot of "could be this, could be that" speculation to the point where its a bit overwhelming...personally I find the tappets a bit noisy when hovering directly over them but I don't know what they should sound like... and I have not noticed any correlation between their noise and how well the engine runs, or in fact whether there is any change in the noise at all. I also do not notice any difference between the different tappets, they all seem noisy to me.

I will listen carefully next time but I fear that judging anything by the noise is just going to be speculation due to my unfamiliarity with the characteristics of the model of engine. I really need to pursue tests that give concrete measurable answers...

It is overdue for an oil change, but again its only speculation that it might affect the tappets and changing the oil/filter in search of this problem would just amount to more "swap and pray", even though it is due to be done.

I think I have enough tests to carry out first before worrying about valves, such as checking the air intake is clear and getting a scope onto the ignition system.

If I "waste" too many more weekends tinkering with the car without results I'm in serious danger of being shot so hopefully I can be a bit smarter in my testing and home in on the general area of the fault a bit more quickly than in the past. More measuring and less swapping based on hunches...

[addo]

The other giveaway with a sticky or collapsed lifter is its noise! Dak-dak-dak-dak.... This can on some motors, be confused with a variator - but the Xantia's V6 doesn't have VVT. The compression pressure jumping similarly on all pots says to me that (as expected) there is no valvetrain issue.

An interesting and useful find for you is Autometer accessory 3224. Unlike most such kits offered, it doesn't use solder but brass or copper compression olives to seal the tubing - making it ideal for exhaust pressure checks.