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

[Mandrake]

Finally a chance to write up Saturday's effort at Richards.

First of all a huge thanks to Richard for his help, I couldn't have done either of the jobs on my car successfully on my own, I'm glad we got his C4 Picasso sorted out too but also gutted that we ran into major problems with changing the rear spheres on his Xantia HDi. (I'll leave the gory details of that to Richard's own thread over here: http://www.frenchcarforum.co.uk/forum/v ... =3&t=44087" onclick="window.open(this.href);return false; )

We did both the exhaust and the hydraulic pump on my car, if I'd known what a pig of a job getting the pump out was I wouldn't have tackled it at the same time....(sorry Richard! ) on the other hand it was leaking quite a lot giving the alternator a good oil bath, and could have put me off the road in the near future so I'm glad it's done and out of the way.

I think the hydraulic pump in the V6 fits the definition of "shoe horned in" very well indeed. It had us both a bit baffled as to how it was actually removed but we eventually figured it out.

As I suspected from my last encounter with a Xantia hydraulic pump we weren't able to shift any of the torx bit caps for the steering pistons even with a small breaker bar, and no impact driver was available so we decided to leave those alone since it was unlikely they were leaking and the fact we might damage the pump if we tried too hard to get them out...

I think we found the leak anyway - the bottom suspension piston had a partially shredded o-ring that was pinched over the end of the thread on the plug, it wasn't sitting properly in its grove at all and must have been leaking. The top one was squashed into a v shape and had gone hard in that shape as well.

We took the end cap off the power steering section to replace the large o-ring and discovered that there seemed to be a very thin paper/fibre gasket on the faces between the cap and the centre body of the pump, that had broken up in to multiple pieces - this threw me for a loop because I wasn't expecting there to be any gasket there, there wasn't one there on the other Xantia pump I've done, just the o-ring. (although it was a different model of pump) No mention is made of it on service.citroen.com either. Since it was broken I had no choice but to remove it and just fit the new o-ring. The old o-ring was completely squashed flat with age allowing the end cap to turn easily with the bolts removed, so it can't have been sealing well and was probably relying on the gasket. The new o-ring was a good tight fit so I can't see that leaking any time soon even with the gasket missing. Does anyone know if there should be a gasket there or has someone added one to try to fix a leak ?

It's hard to be certain that all the leaks are stopped because despite wiping up as much oil as possible with the pump out there is still quite a bit inside the alternator which is probably spinning it out like a washing machine on spin I've done about 60 miles since refitting it and although there are a couple of drips of oil hanging from the alternator dripping onto the A/C compressor its a tiny fraction of what was there before. I'm pretty sure that its just the alternator throwing out a bit more oil from its armature. I'll keep an eye on it over the next few weeks and wipe up any oil I see to see if it has really stopped completely. I think this is a successful repair though, touch wood!

The pressure regulator seemed to be wet and dirty all over just as you predicted Jim, it's even wet on top of the bracket at the top The leakage seems quite minimal compared to the pump so I've wiped it up as much as possible and will observe over the next couple of weeks. I think my only hope to fix it if its the top o-ring is to remove it and send it to Pleadies to recondition it. Of course that means no car for a week maybe two and just now is not a good time for me to have the car off the road for that length of time...

Onto the exhaust in the next post.

[addo]

The super-thin fibre gasket is original; I reassemble with Threebond 1211 and no gasket after sluicing both faces with carby cleaning spray.

Get yourself an "Attack Driver" or copy thereof; you will come to appreciate the merits of this little beauty as the Koken one has a half inch drive so you can cobble up all manner of extensions and universals to get a whack in fiddly places. Many lookalikes have only the 5/16" female hex.

[Mandrake]

The super-thin fibre gasket is original; I reassemble with Threebond 1211 and no gasket after sluicing both faces with carby cleaning spray.

Hmm. The faces were clean and smooth after removing what was left of the gasket, but I didn't apply any sealant to it. Do you think it's likely to leak without any gasket but with a new o-ring ? As I say, the other pump I reconditioned definitely didn't have a gasket there, and that went for 4 years (until sold) without any leaks after I fitted the new o-rings.

At least now I know how the pump comes out if I have to take it out again.

[addo]

Not worth the grief to do it again, I say.

I'm assuming you didn't change front seal or o-ring (neither are PSA listed parts, they're bearing shop jobbies). You need a good puller or press to get the driven hub off the shaft nose.

[Mandrake]

Not worth the grief to do it again, I say.

I'm assuming you didn't change front seal or o-ring (neither are PSA listed parts, they're bearing shop jobbies). You need a good puller or press to get the driven hub off the shaft nose.

No I didn't touch those, I didn't withdraw the shaft from the middle section either.

I did change the front o-ring between the triangular plate and the centre housing on the previous pump I did - as you say PSA don't list a replacement seal but I used a sphere neck seal which did the job perfectly, however I decided not to replace it on this one as that seal is sealing against slight vacuum not pressure and won't be the cause of my leak unless it was really bad. (I replaced it on my previous car because I was trying to eliminate potential sources of air being sucked in to the pump)

[Mandrake]

First thing to do on the exhaust was get the oxygen sensor out. Predictably it didn't come willingly. With the oxygen sensor socket and 3 extensions we were able to get a bar on it through the wheel arch but it wouldn't budge, rounding off one of the flats...that was after a good soaking of LPS and also plus gas. We decided to try a giant ring spanner directly on the sensor which meant cutting the plug off the cable. It was still very tight but that did the trick.

The old oxygen sensor tip was completely and utterly coated with carbon to the point where even the slots were nearly clogged up. I wish I'd taken a picture of it but forgot, its no wonder its response time was so slow. In fact the entire exhaust system turned out to be completely coated with carbon on the inside so it must have been running rich for a long period of time - whether before or after I bought it I don't know. The oxygen sensor is definitely the original too - the manufacturing date on the plug was 1998.

Next we fitted the exhaust back pressure adapter, connected up the gauge and started it up expecting (hoping, in a way) to see horrible back pressure, and saw.....NONE. None at all. Not at high RPM, not with wide open throttle, not anywhere. I took a video of it as well. There is a slight flutter to the gauge when closing the throttle suddenly (turbulence in the pipe when the flow suddenly drops I guess, and its an un-damped gauge) but that's all:



On all the examples I've seen online of a blocked cat the gauge (which only goes up to 10psi) would be pegging itself off the end of the scale on a snap throttle test. A good exhaust would only reach about 1psi on a snap throttle and show nothing at steady high rpm, mine is showing even better than this.

The reading was so low that I was wondering whether the gauge wasn't working or wasn't connected properly - the vacuum side of the gauge definitely works and I could get about a 2 psi reading blowing into it. We also tried deliberately blocking the tail pipes with some rags and managed to get a 2 psi reading on the gauge (still not as high as I expected, but that was the old exhaust which probably wasn't leak free...) so the gauge was definitely connected into the exhaust stream. (The reading was pretty much identical with the new exhaust so I didn't bother to film it a second time)

So apparently I did not have a blocked exhaust... The only other possibility of a blocked exhaust would be if the blockage was further upstream, such as the flexible joint, but I can't see any easy way of testing that theory, and I don't think its likely anyway. I did look in the tail of the cat and it looked perfect - I could only see a fine wire mesh, which looked clean, carbon free and unblocked but that probably doesn't mean much if they usually break in the front I suppose.

So we went ahead and changed the exhaust over anyway. The old back box was full of rust that was rattling up and down inside when it was tipped on end, so it looks like the baffles were rusted out inside it. The new exhaust fitted pretty well, the only thing is the pipe going over the cat tail was a tighter fit than ideal.

The new exhaust doesn't have the loud, deep boom at 1500 rpm under load that the old one did, however it does sound a bit more "rawty" at around 2500 rpm, nothing unusually loud but you do notice it slightly when cruising at 60mph at 2500rpm.

We fitted the new oxygen sensor however silly me, I didn't check the connector fitted until after we'd already fitted the sensor, and it didn't fit... the round lugs on the side and the pin outs were all correct but the bump on the side of the plug was on the wrong side... We shaved that off and it went in. I tested it on the Lexia and can see it switching high/low much faster than the old one, something like 2x a second at 2000 rpm, or roughly 5x faster switching than the old one. I tried to connect my scope to the sensor by back probing the wire on the harness side plug but for some reason I couldn't get a connection this time so I eventually gave up. (I have done it successfully before with the old sensor)

We took it for a test drive and whilst I wouldn't say it was running as well as I'd expect it was running pretty nicely, the intermittent throttle lag seemed to be gone, and it seemed more responsive at low revs. On the 20 mile drive home it performed very well, again not as well as it did during the visit to David but better than it has been lately.

I was going to post yesterday but I thought no I better drive it again before posting a conclusion... Sunday I decided to go for a drive so we took a drive up to Stirling and back. Despite the fact that the performance the night before had been very good all the way home I could immediately tell the performance was sluggish and poor the moment we set off. All the way to Stirling and back, around 40 miles the performance was very mediocre and sluggish, right back to where it was before swapping the oxygen sensor and exhaust... No throttle lag, I think the oxygen sensor has fixed the intermittent throttle lag, but the performance was just very flat feeling. Lots of throttle opening for not much response.

So I'm still baffled. The exhaust is not blocked but the performance still intermittently sucks. The car didn't move anywhere between arriving home with good performance and setting off the next day with immediately poor performance, and it didn't get better or worse during the 40 mile trip.

When I got back I decided to put the Lexia on it and monitor some parameters during a drive with a co-pilot to read out the numbers to me... things of interest were injector opening times, oxygen sensor reading, and timing. On previous occasions I'd noted that under wide open throttle below 3000 rpm it was reading lean on the oxygen sensor suggesting a fuel delivery problem, I didn't see any trace of that yesterday. Every time I tested wide open throttle the oxygen sensor immediately went rich and stayed there, both low rpm and high rpm. So there doesn't seem to be a fuel starvation problem and there may never have been - it might have been the very sluggish oxygen sensor giving an incorrect reading.

Next was injector opening time, which varied from about 2ms at light cruise to about 8ms under wide open throttle at low rpm to about 11ms wide open throttle at higher rpm. I don't know if these are normal figures or not mind you...

The other thing of interest was ignition timing in particular the torque reduction retard which I've seen give suspicious readings before. What I was seeing is during driving the torque reduction retard figure jumping around between 0 degrees and -15 to -20 degrees under varying conditions of load, causing the final advance figure to vary between about 30 degrees down to about 10-15 degrees.

I might be misunderstanding it, and its not documented properly in any of the Citroen does I've poured through but my understanding is that the torque reduction retard figure is only used for two purposes - (a) to reduce engine torque while the gearbox is changing gear by retarding the timing (this is explicitly stated in the gearbox training manual) and to retard the timing in response to the knock sensor detecting pinging in the engine. (I'm assuming this is applied to the torque reduction figure, it may be applied to the base advance figure, it's not documented that I can see, but either way it retards the timing...)

When I got back from that drive I noticed its back to its old trick where if you open the throttle to about 1200 rpm the idle speed keeps hunting up and down.... I was looking at the sensor data for a correlation between the change in rpm and any other figures, the injector opening time is remaining absolutely constant, the throttle is being held steady, tps reading is steady, map reading is changing slightly but I think that's a reaction rather than a cause, but the one thing that correlates is the torque reduction retard figure which is jumping back and forth between 0 and -18 degrees, as demonstrated in this video: (the engine noise is very low in this video as it was taken inside the car)



As far as I can see there should be no torque reduction correction being applied under these conditions, which suggests that maybe the knock sensor is picking up too much metallic noise from the engine (noisy valve gear for example) which is making the ECU think that the engine is pinking and retarding the timing, thus causing the rpm to drop.

One of Scanner Danners videos shows a car that was intermittently very down on power which they traced to the timing being unusually and excessively retarded in response to knock sensor input - they were able to monitor both the knock sensor voltage and resultant timing retard in real time on an OBD-II scanner and demonstrate the issue, (unfortunately I can't monitor the knock sensor on the Lexia) the timing was being retarded by over 20 degrees from normal killing the performance. They proved it by temporarily disconnecting the knock sensor at which point the performance jumped back to normal:



Sadly in the case of that car the knock sensor was working fine and the cause of the problem was a mechanical fault in the engine that was genuinely making a lot of "noise" in the engine block thus falsely exciting the knock sensor...so no easy fix.

Although the knock sensor is impossible to get at I wonder if the connector is, I know its a black 3 pin plug and the knock sensor itself is mounted in the middle of the top of the V-block (underneath all the inlet stuff) but does the connector come out the left or right hand side to somewhere where I could get a scope on it or unplug it ?

I'm thinking this might be the next step. Could it be something as silly as noisy hydraulic valve adjusters making enough racket to trigger the knock sensor ? Perhaps an oil change and new filter is in order in that case ? (I haven't done one since I've had the car so its long overdue, I've been so busy chasing faults I've overlooked some of the basic maintenance items... )

Does anyone have any recommendations on a good choice of oil and filter for the V6 ?

[lexi]

Mmm. I thought that V6 of yours was just a little noiser on top end than the mk 1........just barely noticeable. Thought to change oil anyway.

AEP list the smaller diesel filter as a x over to V6 and that is what I put on. Same size was on it, however I think the proper one is a bit bigger. Fuchs 10w 40 semi I put in. Suppose you could try a full Synthetic which can help sticky Hyd tappets.
Lots of hassle/replacement but you still got a running car I suppose? Maybe you will get your finger on it in next move

EDIT: The oil filter that was on it looked as though it had been on there a a long time........well longer than mine are.

[411514]

Im afraid I cannot offer you any help in your struggles Simon as you seem to be far more knowledgable than me.
However I have just read your thread, start to finish and all I can say is I hope you manage to fix it and the end result justifies all your hard work and persistence.
Such a shame that such a minor fault has spoilt your experience so much. Best of luck, Sam

[Mandrake]

Im afraid I cannot offer you any help in your struggles Simon as you seem to be far more knowledgable than me.
However I have just read your thread, start to finish and all I can say is I hope you manage to fix it and the end result justifies all your hard work and persistence.
Such a shame that such a minor fault has spoilt your experience so much. Best of luck, Sam

Wow that must have been one heck of a long reading session Sam. Maybe you should have just jumped to the end to see that the problem is not yet resolved... Thanks for the words of encouragement, at least I've learnt a lot in the process even if I haven't got to the bottom of everything yet...

[Mandrake]

Mmm. I thought that V6 of yours was just a little noiser on top end than the mk 1........just barely noticeable. Thought to change oil anyway.
AEP list the smaller diesel filter as a x over to V6 and that is what I put on. Same size was on it, however I think the proper one is a bit bigger. Fuchs 10w 40 semi I put in. Suppose you could try a full Synthetic which can help sticky Hyd tappets.

Yes I've noticed the hydraulic tappets are sometimes a bit noisy when cold, maybe they are a bit sticky, although they didn't really seem any noisier than David's V6 Activa donor car... so you think fully synth oil might be beneficial for sticky tappets ? I'll give it some thought and maybe do an oil and filter change next month. I've never had to top up the oil once on this car, its still just slightly below half way which is where it was when I bought it, so oil burning is not one of its ailments

Lots of hassle/replacement but you still got a running car I suppose? Maybe you will get your finger on it in next move

Yeah I still have a running car, I know now I don't have a blocked exhaust and that the cat is probably ok, its just that when its misbehaving it drives like a 2 litre engine trying to push along a 1500Kg car

[Mandrake]

The ebay vendor has sold 2 of those, including 1 to me, and it is the best fitting V6 exhaust I have had, including genuine systems. The sound is fairly good, a little noisier than genuine, but much better than the other systems I have bought on ebay. The only issue with fitting was that the inner diameter of the sleeve going onto the cat tail was a bit tight, and I had to grind it out slightly with a dremel. I was happy paying £95, now the price is reduced to £65, it is a steal, and I suspect that there won't be any more after these have gone.

DOH! Looks like the exhaust repair wasn't quite successful. On a short drive tonight I thought the exhaust seemed a bit noisy, got home, sure enough the joint near the cat tail is leaking and chuffing away. The reason why is because we weren't able to get enough overlap in the joint because as mentioned above the sleeve that fits over the cat tail on this particular exhaust system is a bit too tight.

Because my cat tail is missing its outer sleeve we had put slots in the enlarged portion of the previous silencer pipe and use a normal exhaust clamp and paste - which worked absolutely perfectly with the old sleeve for 5 months as it had full overlap and was a snug but not over tight fit. We did the same thing with the new sleeve but couldn't get the cat tail to push fully in, so not enough overlap.

I guess what I'll need to do is pull the joint apart again and grind out the inside of the pipe a bit until it fits properly. What type of dremel attachment did you use to grind out the inside of your pipe ? I don't have a dremel but I do have a battery powered drill, am I likely to be able to use a small grinding bit attached to that to grind out the inside diameter slightly to get it to fit ?

Can anyone suggest a grinding bit for my drill that might do the job ? It would need to have a stalk a couple of inches long or so to be able to reach far enough inside to grind the full overlap region.

This maybe ?

http://www.diy.com/nav/fix/power-tools/ ... d=13269405" onclick="window.open(this.href);return false;

or these ?

http://www.diy.com/nav/fix/power-tools/ ... Id=9296209" onclick="window.open(this.href);return false;

[lexi]

You can normally pick those carborundum bits up in Bills Tool store for 2 quid each. You should be visiting there for your inevitable garage workshop. The proper Dremel bits are good, though quite pricey.
Your other choice is the flap type cylinder that goes on drill.......you can get these down to about 30mm dia, but they will compress into hole slightly.
Carbide Burs can be useful too in many shapes.
Then there are big round files You can borrow a proper Dremel tool here and buy your bit .

[CitroJim]

Simon, the knock sensor connector is very easy to get to. It's the green connector that's exposed when the airbox is removed...

The knock sensor itself is hard to get to and needs the whole inlet manifold removed and that means injectors and everything...

So, when did you last change the oil? If not recently then do. I change mine at 3 or 4K intervals and yes, the oil filter is a bigger one than the XU/D one even though the little XU/D one fits.. Personally I'd only use the correct big one...

[Mandrake]

Green ? The data from Citroen Service box says its a "3 way black connector", although it doesn't give any indication of where the connector sits... I'll have a look under the air box anyway, I have the pinouts of the ECU so I can confirm whether its the right plug from a continuity check.

As I said, I haven't changed the oil or filter in the year I've had it. I haven't done a huge mileage though, no more than 3,000 miles over a year as it was just under 100,000 when I got it and it hasn't reached 103,000 yet. Yes, I'd go for the correct filter.

[RichardW]

A thought crossed my mind the other day Simon - what petrol are you running it on? If not branded super U/L then maybe try that and see if it improves things? I didn't think the engine sounded particularly noisy when you were here - but than I am used to the dulcet tones of an HDi