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

[Hell Razor5543]

What was it Sherlock Holmes said?

"How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?"

Mind you, that was before cars (and, ESPECIALLY, Citroens) were invented!

[RichardW]

2) I've just had a sudden realisation of another possible cause of the valve timing being out WITHOUT the belt actually jumping a tooth. What if the cam sprocket bolts have been under-torqued when the belt was last changed, and they've worked loose enough to allow the cam sprocket adjustment to slip when subject to enough torque, for example abrupt acceleration and deceleration events ? (Or rough gear changes due to a cranky gearbox, which has certainly happened plenty of times)

When you mentioned cam timing, my first thought was "Does it have vernier sprockets"......

Best hope the weather holds and you get round to mine early then!! I've got M8 bolts somewhere I've used for timing engines before (only 3 though )- if you just want to check, it should be possible to rotate the engine till the first cam comes to its timing position, and then check that all the others are in time as well, and the cam sprocket bolts are tight. There is always the danger that when you take off the covers you will see something you rather wish you hadn't though....

[Mandrake]

When you mentioned cam timing, my first thought was "Does it have vernier sprockets"......

Yep.

Here's a good picture stolen from David's blog:



I've never owned or worked on the timing of a car with adjustable cam sprockets before, only single piece sprockets, so it took me a little while to grasp how the timing was set - as soon as I realised the peg goes through the part attached to the camshaft not the outer sprocket it all suddenly made sense... Now that I understand it its a beautifully elegant system allowing you to easily get the timing perfect despite differences in belt stretch between old and new belts and even adjust it without having to remove or re-tension the belt...

So much better than counting teeth and/or trying to compare hard to see timing marks on sprockets and block visually!

Best hope the weather holds and you get round to mine early then!! I've got M8 bolts somewhere I've used for timing engines before (only 3 though )- if you just want to check, it should be possible to rotate the engine till the first cam comes to its timing position, and then check that all the others are in time as well, and the cam sprocket bolts are tight. There is always the danger that when you take off the covers you will see something you rather wish you hadn't though....

If the weather is good and the brake job goes well (famous last words) then it would be nice to check the timing - if only to ease my mind and rule something else out. As you say we might see something horrifying under the covers - but I think I'd rather know what I'm up against than keep my head in the sand!

If there's even a chance that the cam timing could be the long sought after problem that has be dogging me for so long (especially if its loose bolts on the cam sprocket) then I need to know. And if there is something horrible going on under the covers and belt replacement is urgent then I'll find the money for the parts and find a way to get it done.

[CitroJim]

I have yet to fine a V6 engine accurately timed after a few thousand miles, also you have to be jolly careful how you set up the timing at belt change time and it can take several goes to get it spot-on. I wonder how many ignore a small error as being too hard to get right as the belt would have to come off again to make it good.

Definitely worth a check at least at this stage of the game...

[lexi]

There is over 100k and 14yrs on belt? Time it was done anyway or it will be John Lee Hooker time

[Mandrake]

Are we sure it's on the original belt though ? Surely that would have failed long ago ? Could someone have changed the belt before you bought it and perhaps done a crap job of the timing and installation ?

All going well the cover will be off next weekend for a close inspection of the belt and timing so we'll have an idea as to its condition then. I agree that it needs to be done and long overdue if its original, but its £300 in parts and a lot of work that may lead nowhere if I can't find/fix its running/emissions problem and it can't be made to pass the next MOT just over two months from now...

If the timing turns out to be the problem I'm searching for or I find it elsewhere and it passes the MOT then it will get done.

I've been doing a lot more reading about valve timing and its a very interesting but very complicated topic... it's not something I've paid any attention to before so didn't really have idea what effects timing being out might have, apart from bent valves... If I had, and had I known that the engine used vernier cams I might have checked the timing long ago.

On engines with a single cam per head you can't have an error in lobe separation angle (and therefore overlap) because its ground into the cam, but with separate inlet and exhaust cams an error in timing between inlet and exhaust will change separation angle and overlap, also this engine has separate cams per head so there is a lot of potential for complicated effects if the four camshafts are out by different amounts!

Any imbalance in valve timing between front and rear banks is going to cause rough raspy running for example due to uneven power pulses between banks...it's quite conceivable that there could be a timing problem only on one bank.

I was previously thinking that retarded valve timing (or retarded inlet valve timing at least) fitted the symptoms more (lack of low rpm power) but after rethinking it I'm now wondering if it could in fact be the reverse - an advanced inlet cam. Why ? Because retarded doesn't match the lean running and knocking symptoms.

Advancing the inlet cam should give more power at low rpm, but the way it does it is by increasing VE at low rpm, eg more total air is inducted because less is pushed back out before the valve closes. That's all very well for a carburettor or mass air flow injection system - because both will automatically detect increased air flow and increase the fuel to match that extra air flow to keep the mixture correct.

A speed density injection system (like this) will not! In fact it will do the opposite - more air flow with the same throttle angle due to a VE increase will increase vacuum which is a signal to a speed density system to decrease injection pulse width not increase it!

So any unexpected increase in VE from advancing the valve timing will cause the mixture to go doubly lean (at low rpm) - partly because the air flow has increased over what it should be, partly because the injection pulse has been reduced as a result of the MAP reading dropping. The computer is programmed with air flow assumptions at all RPM conditions - change the valve timing and you break those assumptions. In open loop mode the ECU has no idea that it's now fuelling lean at low rpm.

But wait, it gets better. The engine is clearly showing signs of knocking under load below 2500 rpm, and lean mixture can cause that, right ? Well yes, but it may not be the only contributor.

I've bumped the fuel pressure up to 3.4 bars now, it isn't running lean at low RPM any more, it does run much better, but I can STILL feel that inconsistent power delivery below 2500 rpm where the power fluctuates under a constant load and feels like it is being "held back", and it only happens on a warmed up engine. I don't think its a misfire, it feels like the ECU is still making adjustments to the ignition timing under load in response to knocking. (I could check for sure with the scope but its a lot of work to set that back up again)

Why would it still be knocking if the mixture is rich enough ? Possible answer - Dynamic Compression Ratio. Again, something I've never heard of until I started researching valve timing... turns out that static compression ratio (10.5/1 on this engine) is only part of the picture. Static compression ratio is the ratio of combustion chamber space at BDC versus TDC. Everyone knows that.

Dynamic Compression Ratio is the actual ratio that the gasses get compressed at on a running engine and while affected by static ratio its also critically affected by other factors, most significantly inlet valve timing. An inlet valve that closes later than normal reduces Dynamic Compression Ratio and peak cylinder pressures at low rpm, while an inlet valve that closers earlier than normal increases Dynamic Compression Ratio and peak cylinder pressures at low rpm.

So if an inlet cam is advanced one of its effects is that at low rpm it gives an apparent increase in compression ratio and an increase in peak cylinder pressures - which will make the engine more prone to knocking and detonation! That's even if the mixture was correct (if it had been a carb or MAF system) so if you then add a lean mixture on top of that (due to the speed density system) the tendency to knock is even greater still, possibly to the point where it is knocking all the time under load at low rpm with anything other than maximum ignition timing retard.

This effect of leaning the mixture and increasing dynamic compression and peak cylinder pressures will only happen at LOW rpm, hence the knocking is only occurring below a certain threshold - about 2500 rpm. The scope did indeed say that timing was normal above 2500 and very suddenly dropped back to a full 12 degrees of retard just below 2500 rpm.

Seems to fit the symptoms huh ? Lean running and knocking, knocking which is reduced but not eliminated with a richer mixture...

What else does it fit ? I've done a cranking compression check on this engine before and they were unusually high - about 212 psi from memory. In all my research I've rarely seen any reference to any cranking compressions on a passenger car engine being higher than 190 with 140-190 being more typical. True it is a 10.5/1 engine but it still seems a bit high, without having an official figure to check against. An advanced inlet cam would cause the cranking compression to be higher than normal.

An advanced inlet cam with normal exhaust cam would reduce the LSA and increase the valve overlap since the inlet valve starts opening sooner before the exhaust has finished closing, one of the symptoms of that would apparently be a reduction in idle vacuum - I only measure 15" HG at idle, when typical apparently should be 17-21" for the majority of street tuned engines. Again perhaps a clue, although I don't have an official or comparative figure of what it should be for this engine.

Excessive valve overlap can also increase emissions at idle apparently - because some of the incoming fuel will just go right through to the exhaust without ever being burnt, eg higher HC's.

Because there are two inlet and two exhaust cams the possibilities for weird symptoms depending on the specific combination of timing errors is huge, the reality is any timing error in any direction of any cam is likely to cause problems of some sort or another so I'm really looking forward to checking the timing to see if it could be the eureka moment I've been looking for for so long...

[lexi]

Mmm... I am doubting that belt had been changed Simon. Looking at other things, like a blinkin oil filter with a date of years gone by etc etc. just left me with that impression of not cared for . I changed the aux belt as that looked knackered and ready to go.
There is the possibility that another £200 and hassle may not fix this. You will go for it ! Needs done anyway?

[furriegurrie]

The timing belt for these engines is huge. It's not very likely to fail. In the later versions with the same belt but a different spanner the changing interval is stretched considerably to something like 220k kilometers. When I replaced mine it looked used but actually not very worn at all.

[Mandrake]

Yes from the pictures I've seen its a pretty massive, hefty belt, so it could be that roller guide bearings or failed tensioner are more common causes of belt failure on this engine than the belt breaking of its own accord.

There is a dry somewhat noisy roller in there somewhere that has got slightly noisier in the last two years, although it might be the black roller on the auxiliary belt as it seems to be coming from that vicinity.

I won't know for sure until I get a chance to run the engine with the auxiliary belt removed and the cam covers off. In any case I wouldn't be fitting just a belt unless it was an emergency, when it gets a belt it will get a tensioner, rollers and water pump.

By the way here is one of many articles on the effects of cam timing, just look at that list for small LSA:

http://www.compcams.com/Pages/413/cam-t ... angle.aspx" onclick="window.open(this.href);return false;

That's pretty much a list of what I'm seeing:

Narrow power band - check.
Increase chance of engine knock - check.
Increased cranking compression - check. (At least it seems unusually high)
Idle vacuum reduced - check. (At least it seems unusually low)
Idle quality suffers - check. (At least when hot)

To get a lower LSA would mean an exhaust cam is retarded or an intake valve is advanced. I wonder which is more likely....

[Hell Razor5543]

To get a lower LSA would mean an exhaust cam is retarded or an intake valve is advanced. I wonder which is more likely....

Knowing your luck, and the way your problem child works, BOTH!

[xantia_v6]

You can of course check the valve timing without changing the belt, the covers need to be removed and the crank damper, but it should only be a half day job (I did it in under 2 hours after over-tensioning a belt).

If you shop around and wait for discounts, you should be able to pick up a complete timing belt kit for about £200, particularly if you are prepared to buy an early style water pump and cut off the extra lug.

I cant help thinking that you should run the car with a 4 gas analyser, to corroborate your diagnosis that it is running lean. You would be welcome to borrow mine, but it is currently in NZ, and I don't expect you would appreciate paying the postage ( about 5kg I think).

[RichardW]

If the cams are out, they can only be retarded I think.... The outside part of the sprocket that is driven by the belt is 'dragging' the cam round. If the bolts slip, then the sprocket will keep turning with the belt whilst the cam lags behind, until the bolts get to the end of their slots, when it will then be pulled round - thus it must be behind where it should be.

[Mandrake]

You can of course check the valve timing without changing the belt, the covers need to be removed and the crank damper, but it should only be a half day job (I did it in under 2 hours after over-tensioning a belt).

Was that 2 hours just to strip it down, or 2 hours to strip it down, adjust the belt timing and put it together again ? I hope the latter, as I can't imagine it being 2 hours just to strip it down or we'll be there all day. (Doing rear brake discs/pads too)

If you shop around and wait for discounts, you should be able to pick up a complete timing belt kit for about £200, particularly if you are prepared to buy an early style water pump and cut off the extra lug.

I've got the details of the kit David H bought for his Activa conversion, that was about £250.

I cant help thinking that you should run the car with a 4 gas analyser, to corroborate your diagnosis that it is running lean. You would be welcome to borrow mine, but it is currently in NZ, and I don't expect you would appreciate paying the postage ( about 5kg I think).

Thanks for the offer, a gas analyser certainly would have been useful a few months ago but I think I'm now at the point where I'm sure its running lean - it has a known good O2 sensor that is clearly showing it swing hard lean under wide throttle at low RPM, and pretty consistently too.

The scope is on the dashboard whenever I'm driving so it's not like I checked it once or twice, its been continuously monitored during the last couple of months driving at least, and I've even used my variable resistor in series with the IAT to tweak the mixture on the fly to see the reading under a constant load condition go from lean to rich as I adjust it...

If we assume for a moment that a valve timing error is the cause of it running lean (due to altering the breathing of the engine at low rpm) then its quite possible that only one bank is running lean, depending on which cam(s) are out.

In closed loop mode this would cause the ECU to increase the injection times until the LEAN bank of the two was running at a stoichiometric mixture (since it only measures excess oxygen and adjusts the mixture until it just disappears) which would cause the "good" bank to be running rich. This rich running bank could easily explain the higher HC and CO readings at the last emissions test, especially the idle one. Only if the mixture is well balanced across all 6 cylinders will the emissions be low...

So in closed loop mode the "good" bank runs rich and the "bad" bank runs at the correct mixture. Meanwhile in open loop mode during acceleration the "good" bank runs at the correct mixture while the "bad" bank runs lean. (Also the "bad" bank would have abnormally high peak cylinder pressures and a strong tendency to knock, causing the timing of both banks to get heavily retarded with the consequent big loss of power)

[Mandrake]

If the cams are out, they can only be retarded I think.... The outside part of the sprocket that is driven by the belt is 'dragging' the cam round. If the bolts slip, then the sprocket will keep turning with the belt whilst the cam lags behind, until the bolts get to the end of their slots, when it will then be pulled round - thus it must be behind where it should be.

If the bolts just came loose on a cam wheel then I agree, it could only reasonably be a retarded cam as the end of the slot would just pull the cam around a few degrees late, in which case it would have to be an exhaust cam retarded to fit the symptoms, I think.

But if the belt has jumped a tooth or the cam bolts are tight enough to grip under normal load but slip during any sudden kickback then it could potentially jump either way. I've read that belts jumping a tooth due to poor tensioner loading often happens when the engine is stopped and it kicks back slightly as it comes to a stop.

There have been times in the past when the gearbox has changed gear very poorly and given a hell of a jolt that is enough to pass right through the car as if you hit a pothole and shake the engine harshly on its probably worn engine mounts - that kind of sudden jolt causing a rapid acceleration or deceleration of the crankshaft might be enough to cause the belt to jump a tooth if the tensioner isn't working properly (possibly even jump forwards during a sudden deceleration) or a cam wheel adjustment to slip if its not properly tight. Maybe...

(Perhaps having gearbox problems and engine problems at the same time was not an unlucky co-incidence after all...)

I never thought I would see the day that I'd be hoping to find a valve timing problem on an engine when removing the timing cover! Previously it would have filled me with dread and apprehension but I've got to the point now where if it is out it's been running that way for over a year, so if it was going to bend a valve or break it probably would have done so by now...

[RichardW]

Fair enough

Hopefully we'll see next weekend, but given the number of blind alleys we've been down with it, I'm not getting my hopes up!!