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

[CitroJim]

Simon, I believe cylinder 1 is on the front bank nearest the bellhousing and cylinder 4 is in the same place on the rear bank...

As you look down on the engine from under the bonnet:

6 5 4

3 2 1

[Mandrake]

Ok folks, I've just spent the afternoon looking for this problem, and my conclusion is the coil pack is not faulty. On the one hand that means a £170 part doesn't need replacing, however on the other hand my troubleshooting has just raised more questions than it has answered.

Yes the coil pack has some cracks in the plastic around the perimeter, one on the far right hand side by the bolt holes, one on the side of the middle pack (see pictures below) however these appear to be cosmetic only, the potting is in perfect condition and there is no evidence of problems with the actual coils.

Using the Lexia to test fire the coils I tested each pair of outputs on each coil and they can all produce an identical 15mm long spark with alligator clips for spark gaps with no differences detectable between the different coils:

From this I have to conclude that despite appearances there is nothing wrong with the coil pack. It's a Sagem by the way, which I guess means it has already been replaced at some point in the past if Sagem is not the OEM brand...

(continued)

[CitroJim]

I don't like the look of the crack on the middle coil Simon...

Now, I know £170 is a lot to spend on a new one and ideally if you could try another as a test you'd know for sure.

I'm happy to lend you mine on a try and return ASAP basis... PM me if that will help you with this...

[Mandrake]

So with the coil pack seemingly eliminated what else could it be ? One possibility is the 2.2uF capacitor which I presume works much like a points capacitor, which seems to be literally crumbling away:

If that capacitor is faulty the coils will still throw a spark once every 2 seconds during the Lexia test, but would not fire properly at engine RPM, so I guess its cheap and worth trying to find a substitute capacitor. (Check the simple/easy things first...)

I took the three front plugs out to inspect them and while they look a bit worn I don't see anything horribly wrong with them:

I put them back in and swapped the plugs between cylinders 1 and 3 to see if the problem would move cylinders but it didn't. Of course the rear plugs or spark plug leads could still be faulty, but I'm assuming that means a new exhaust gasket and a couple of hours work to replace them.

It's probably due new spark plugs especially at the rear, but I think I'd probably just measure the plug leads when replacing the plugs rather than ordering leads in advance, as no doubt they're expensive.

On the theory that one of the rear cylinders might be misfiring I tried running the engine with each of the rear plug leads disconnected one at a time - not only does it run much worse as expected, the change in running is identical for all three leads - if one plug wasn't firing you'd expect little or no change when disconnecting that lead, but all three leads made a big and almost identical difference to the running, which tends to suggest that no particular cylinder is misfiring more than any other, also disproving the faulty coil pack theory.

Another thing I discovered is that leaving any one of the rear plug leads disconnected made NO difference to the dwell time for each coil reported by the Lexia - apparently a misfire is not indicated by the Lexia after all, meaning that the difference in dwell readings I'm getting is probably just a red herring. The difference is only 0.1ms after all.

Assuming I replace the plugs and the spark plug leads are ok, what is left ? Could the symptoms be due to an injector not working properly ? All the injectors passed the actuator "click test" of the Lexia, however I'm not sure that that proves that they're injecting fuel, only that the solenoid is being energised and is clicking.

So after all this I'm not really any further ahead, except that I don't think its the coil pack now.

Could I still be looking at a sensor problem which is causing incorrect advance/retard or incorrect injection ? One thing I did notice on the Lexia is that if I fast idle the engine at about 1500rpm not only does the RPM fluctuate up and down 200rpm about twice a second with a steady throttle setting, the ignition advance jumps back and forth between 15 and 30 degrees as shown by the Lexia as the speed fluctuates... (The question is whether the change in advance is cause or effect...)

Starting to feel a bit out of my depth on diagnosing this one now, especially with the lack of any solid clues from the Lexia.

[Mandrake]

I don't like the look of the crack on the middle coil Simon...

Now, I know £170 is a lot to spend on a new one and ideally if you could try another as a test you'd know for sure.

I'm happy to lend you mine on a try and return ASAP basis... PM me if that will help you with this...

Thanks Jim, I'll keep the offer in mind, but I really don't think the crack is a problem, its in the plastic around the base of the coil around the laminated iron (?) core, it won't be causing any problem there, its not at the insulated high voltage end.

I'm at the mercy of weather when working on the car out in the open, so I couldn't guarantee that I could test it and send it back quickly, so if its off one of your cars I couldn't do that to you. A different story if you had a spare one you could do without for a couple of weeks that wouldn't immobilise a vehicle.

I suppose I could try a breaker, as long as the price was low enough it might be worth a punt to see if it makes a difference.

[Mandrake]

For lack of any concrete diagnosis I'm going to replace the 2.2uF capacitor first - its cheap, quick and easy to fit, and has a reasonable chance of actually being the fault now that I've given the symptoms and test results so far some more consideration. I have had faulty points capacitors before (albeit on coil/points systems) that have caused misfiring and the fact that the cap is crumbling away in my hands as soon as I touched it means that whether or not its the cause of this particular symptom it needs to be replaced!

Assuming that doesn't fix it, the next step is either substitute coil pack (easy to change but expensive) or new spark plugs and/or leads. (a big job for the rear plugs, but probably much cheaper than a coil pack, and they're probably well overdue anyway)

Still open to suggestions if someone else has been down this road with this particular set of symptoms.

[CitroJim]

Hi Simon, sorry for the late reply...

The capacitor can be easily tested by leaving it disconnected. It'll only be a suppressor.

Rear plugs are a good thought and should be checked to eliminate them. As they're not easy, they may not have been done for ages. This means lifting the top half of the manifold and when you do this, replace the gasket (0348 K7). In fact you may find the old one is wrecked and replacing it might crack the problem.

Are you Ok with how to lift off the manifold?

Then, before any further major stuff, replace the coolant temperature sensor. This seems a problem in many V6s and is easily and cheaply eliminated. You never know...

Have you had any recurrence of the engine speed sensor fault? I wonder if this might just be faulty and suffering low output at the speed where the issue is apparent. However, i warn you now, it is a complete and utter pain to swap due to where it is so i hope it isn't...

After that, speak to me again about borrowing a coil. It's no problem as I can do without it for a few days - it's not like it's my only car

And if all that fails, I'd be very interested in the results of a compression test.

[Mandrake]

The capacitor can be easily tested by leaving it disconnected. It'll only be a suppressor.

Ah, so it will actually run without it then ? I might try disconnecting it and go for a drive around the block to see if there is any change. If the insulation is breaking down internally it may cause an intermittent misfire. I'll probably replace it in any case.

Rear plugs are a good thought and should be checked to eliminate them. As they're not easy, they may not have been done for ages. This means lifting the top half of the manifold and when you do this, replace the gasket (0348 K7). In fact you may find the old one is wrecked and replacing it might crack the problem.

Is it possible that when the cat fell down due to the exhaust break it could have given the gasket a good yank and done some damage to an already worn gasket ? Spark plugs look quite cheap (compared to the plug leads and coil pack certainly) so it could well be worth doing the rear ones, hopefully the gasket isn't too expensive.

Are you Ok with how to lift off the manifold?

Although I could probably figure it out, a little guidance of what I'd encounter and what needs to be done, special tools etc would certainly help. I'm the sort of person who likes to tackle a job well prepared, especially when I don't have a spare car or a shelter to do the work in. I don't like to start a big job like that without knowing I have what I need to finish the job and get the car mobile again. (I sure missed it the week it was off the road with a broken exhaust!)

Then, before any further major stuff, replace the coolant temperature sensor. This seems a problem in many V6s and is easily and cheaply eliminated. You never know...

Yes I was thinking that too. I presume its not too hard to replace - its down the right hand side (from the front) of the engine below the air intake system correct ? According to the Lexia it does seem to be working ok - when dead cold in the morning the engine coolant reading matches the air intake temperature reading exactly... but you never know.

Have you had any recurrence of the engine speed sensor fault? I wonder if this might just be faulty and suffering low output at the speed where the issue is apparent. However, i warn you now, it is a complete and utter pain to swap due to where it is so i hope it isn't...

You have me confused now, the engine speed sensor fault was in the gearbox not the engine, and yes it has reoccured once since resetting the fault codes a few days ago. Another odd thing is that the gearbox reports a permanent fault with the brake light switch - but why on earth does the gearbox need to sense whether the brake pedal has been pressed ?

If I press the brake pedal a few times while still on the fault code display the fault actually goes away....

After that, speak to me again about borrowing a coil. It's no problem as I can do without it for a few days - it's not like it's my only car

Ok will do if I'm still stumped after all the above.

And if all that fails, I'd be very interested in the results of a compression test.

That's a little bit out of my means at the moment, I don't have a compression gauge and to be honest I've never done a compression test before. I do know the basic idea of how to do it, but having to lift the exhaust manifold to check the rear cylinder compressions is certainly a pain in the **** on the V6. I guess it would have to be done at the same time as the rear plugs or not at all.

Mind you although the rear plug bank is difficult to access I've been pleasantly surprised at how accessible most things are, considering that its a large engine crammed into a small engine bay! I was expecting much worse...

Edit: I didn't realise how cheap a compression gauge is - do you think its worth me picking one up ? I could check the front cylinders right away, and check the rear ones if/when I do the rear plugs. What size thread fitting and maximum pressure do I need ? Would something like this do ?

http://www.screwfix.com/p/engine-compre ... tAod2DsAhA#" onclick="window.open(this.href);return false;

[RichardW]

Don't despair Simon - you are going in the right direction! A code read at a garage would probably have been met with a shrug and "Computer says it's OK mate - come back when it gets bad". If you've got a spark, and it appears you have as it will idle OK, then it's either leaning out or going rich when you try and speed up. I wonder if you might be able to see this on the oxygen sensor live data as lift the engine speed? This might point in the direction of not enough fuel or not enough air -although not sure which way it will go!! The timing may well be following the fault - as the knock sensor picks up the poor running. Wonder what happens if you disconnect the knock sensor.....?

[xantia_v6]

Some thoughts Simon:
There are 3 double-ended coils (so-called wasted spark system), so disconnecting a rear spark plug lead also stops the spark on the corresponding front cylinder (unless you short the coil output to ground).
The required spark voltage rises significantly with combustion pressure. Pressure is lowest with the engine not running, low with the engine idling, and highest with the throttle open at low/medium RPM.
It is the Inlet manifold that you remove to reach the rear spark plugs.
Low compression is unlikely to cause the sort of misfire that you describe without also disrupting the idle.
I would not spend too much money (other than on a spare gasket or two) before inspecting the rear plugs, your symptoms are quite similar to those of a rear plug drowning in engine oil.

[Mandrake]

Don't despair Simon - you are going in the right direction! A code read at a garage would probably have been met with a shrug and "Computer says it's OK mate - come back when it gets bad".

Thanks for the encouragement Richard, yes I think you're right, it would have most likely been met with a shrug or the offer of replacing lots of expensive parts by trial and error to find the problem...

If you've got a spark, and it appears you have as it will idle OK,

Yes the idle seems perfectly steady - apart from the apparently common V6 characteristic of the idle speed dropping near stalling and surging a few times when first starting in very cold weather (which it very rarely does) 9 times out of 10 it starts straight away and idles smoothly and steadily, cold idle is around 850rpm, hot idle 640rpm. It's only when I open throttle a bit that it starts to stagger and the speed hunts up and down a bit.

then it's either leaning out or going rich when you try and speed up. I wonder if you might be able to see this on the oxygen sensor live data as lift the engine speed? This might point in the direction of not enough fuel or not enough air -although not sure which way it will go!!

Yes I can watch live O2 data but like you I'm not sure whether higher voltage means leaner or richer mixture! Will have to look that up. The O2 sensor does seem to work, it cycles between about 80mv and 800mv every second or two once the engine is warmed up, as it should.

The timing may well be following the fault - as the knock sensor picks up the poor running. Wonder what happens if you disconnect the knock sensor.....?

As soon as you mentioned the knock sensor (previously overlooked) I had a sudden flash of inspiration that you might have just hit the nail on the head. Two points - one is I'm running on 95 octane instead of 98, so the knock sensor working properly is that much more important, secondly heavy load and wide throttle at low RPM is exactly where knocking would be an issue and the knock sensor would be doing the most work.

What happens if the knock sensor is faulty and I'm running on 95, could the shuddering and stuttering actually be engine knock/detonation under low RPM load ? With no data from the knock sensor would the engine be too far advanced in these circumstances ? Could it really be that simple ? If the sensor is not completely faulty could it explain oscillation in the engine rev and spark timing when holding the throttle near 1500rpm ?

When I was driving today I felt that the high RPM performance of the engine while slightly down wasn't bad at all, its only the low RPM performance where its really struggling, again consistent with timing issues and a possible faulty knock sensor allowing knocking to occur.

Does anyone know whether the knock sensor live data can be read by the Lexia and if so which sensor it is ? Also where on the engine is it ?

[Mandrake]

Some thoughts Simon:
There are 3 double-ended coils (so-called wasted spark system), so disconnecting a rear spark plug lead also stops the spark on the corresponding front cylinder (unless you short the coil output to ground).

DOH. You're right of course, I should have thought of that. I noticed when I was doing the bench test of the spark that I had to have both active at once to get a proper spark - hence the two alligator leads shown in the picture setting up two independent spark gaps. In that circumstance I was getting a nice fat 15mm long spark from each gap, and almost identical results from each coil. (hence concluding the coil pack is ok)

I can re-run the missing lead test with an alligator lead forming a dummy spark gap to allow the front cylinders to fire properly. This should allow me to identify if one particular rear cylinder is misfiring or not.

The required spark voltage rises significantly with combustion pressure. Pressure is lowest with the engine not running, low with the engine idling, and highest with the throttle open at low/medium RPM.

Yes exactly, hence my comments a long way back in the thread about the possibility of a weak spark. It takes about 3000 volts per millimetre to jump in air, so by my calculation a 15mm spark is 45,000 volts - is that normal for this type of ignition ?

It is the Inlet manifold that you remove to reach the rear spark plugs.

Ah, thanks for that clarification. I'm much happier removing the inlet manifold than the exhaust manifold.

Low compression is unlikely to cause the sort of misfire that you describe without also disrupting the idle.
I would not spend too much money (other than on a spare gasket or two) before inspecting the rear plugs, your symptoms are quite similar to those of a rear plug drowning in engine oil.

Ok I might leave the compression test for now. Do you still think it could be an oiled spark plug if the idle is perfect and the high RPM performance is only down slightly ? If it was oil, what would be causing the plug to drown in oil ?

By the way there is a small oil drip from the top rear of the engine that drips onto the exhaust and burns at high temperature - that's been there since I've had the car but I haven't done any investigation on it yet.

[addo]

What pulse frequency does the Lexia fire coils at? A dead ignition coil of this sort will normally give one decent spark per actuation, then either feeble or none.

Cosmetic appearance of the coils is no guide; I have some here that look awful but work perfectly - yet failed ones can appear fine.

[xantia_v6]

Low compression is unlikely to cause the sort of misfire that you describe without also disrupting the idle.
I would not spend too much money (other than on a spare gasket or two) before inspecting the rear plugs, your symptoms are quite similar to those of a rear plug drowning in engine oil.

Ok I might leave the compression test for now. Do you still think it could be an oiled spark plug if the idle is perfect and the high RPM performance is only down slightly ? If it was oil, what would be causing the plug to drown in oil ?

By the way there is a small oil drip from the top rear of the engine that drips onto the exhaust and burns at high temperature - that's been there since I've had the car but I haven't done any investigation on it yet.

If your cam-cover leak is on the inner edge of the cam cover (or cam block), then over a period of months it will fill the spark plug cavities in the head.
The engine oil is contaminated with carbon, water and other stuff that makes it a bit conductive, and under conditions when the ignition voltage is highest, the current tracks through the oil rather than the spark plug.

resealing the cam covers is an interesting days work (you might as well change the timing belt at the same time).

[Mandrake]

What pulse frequency does the Lexia fire coils at? A dead ignition coil of this sort will normally give one decent spark per actuation, then either feeble or none.

It only fires them once per second during testing.