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

[Mandrake]

I only put branded 95 octane in it - unless I'm away from home I fill up at a nearby Shell. Otherwise I try to stick to Shell or BP and avoid supermarket fuel... I did try it on Shell 97 super U/L for a couple of months (back when it was running well) but it made no noticeable difference, except to my wallet. I can't really afford 97 so 95 is what it gets...

It's not overly noisy as a whole, no, but the tappets are sometimes a bit on the ticky side...Although I'm going to check the knock sensor out (a quick test drive with it unplugged, and also to see if I can get my scope onto it to see how much sensor activity there is) I'm circling back around to the ignition again, believe it or not.

I'm now happy that the exhaust is not blocked, compressions are ok, fuel pressure is ok, injectors are balanced and working, the map sensor is ok, TPS is ok, oxygen sensor is new, etc however I was never quite happy with what I found last time I had the spark plugs out, (5 of the 6 plugs somewhat carbon fouled) the fact that one plug was in a lot different condition to the rest, the fact that that same plug had a much "quieter" sounding spark when doing spark testing, etc.

But what's really got me having a second look at the ignition (and intake) is that I've noticed recently what seems like an intermittent misfire again It's particularly noticeable when started cold, but only SOMETIMES, there is a very uneven beat at idle that distinctly sounds like a single cylinder misfire that goes away after about 30 seconds or so, but even after that this uneven beat is there when the engine is held at 1200 rpm and I distinctly remember that it was NOT like this when it was running well. I can see the engine rocking a bit when the engine note is uneven so I'm reasonably sure it is missing sometimes. Yet another symptom is a major stumble if I snap the throttle open quickly from idle, usually when hot.

So I'm going to go back over the ignition and inlet system again with a fine tooth comb. One possibility is a small vacuum leak on the intake runner to the number 1 cylinder where the top manifold bolts down, that would cause that cylinder to run lean and sometimes lean misfire, meanwhile the oxygen sensor would cause the ECU to run the other 5 cylinders rich to compensate the oxygen reading in the exhaust...(cylinder 1 showed signs of running lean while the other 5 cylinders showed signs of running too rich, although the latter might have been due to the faulty oxygen sensor)

I'm also wondering whether the plugs I've fitted are not a suitable temperature range for the predominately short trips I make causing them to foul... It has Bosch FR7DC+ at the moment which are single electrode, yttrim, non-resister plugs, while the original specified type are Bosch FR8KDC which are dual electrode, copper core resistor plugs... (I can't seem to find the temperature range specs of either plug though)

Another possibility is that the spark is jumping to ground on the cylinder 1 plug at higher combustion pressures due to an insulation breakdown on the plug or rubber boot... I've just picked up an adjustable spark gap style inline spark tester that will allow me to actually get an estimate of the spark voltage under wide throttle conditions: (the neon testers I have now won't give any indication of the voltage)

http://www.pedparts.co.uk/product/950/s ... ion-tester" onclick="window.open(this.href);return false;

If I can find an obvious difference between the gap the spark will jump in series with the 1,5 cylinder pair to the other two cylinder pairs with this under WOT, I'll know I have an insulation break down on one of the plugs that is causing the spark to jump somewhere else other than at the electrode under engine load conditions...

[CitroJim]

Yes, green. The Knock sensor itself has a black connector but it plugs into a green connector!

Unless I'm gone colour-blind of course

[Mandrake]

Yes, green. The Knock sensor itself has a black connector but it plugs into a green connector!

Unless I'm gone colour-blind of course

Ah well that explains things, you didn't say it was a black connector plugged into a green one! should be easy to identify then...

[CitroJim]

It's really easy Simon, Pop off the airbox and look in the vicinity of the thermostat housing..

The connector should be clipped to the side of the engine but it often breaks and ends up floating..

Here in this picture you can see it clearly. The two-tone green affair. Unusually both parts of the connector are green albeit different shades...

[RichardW]

Ha ha - Simon and I were joking only at the weekend that every job on the V6 starts with "remove the air box"

[CitroJim]

Ha ha - Simon and I were joking only at the weekend that every job on the V6 starts with "remove the air box"

It does Richard, it really does...

It's no joke, it's a bit of a PITA really, having two pipes and an electrical connector...

At least Citroen did it right in the XM as there's no need to remove the airbox for anything hardly...

[lexi]

At the mileage you do Simon, using using the super unleaded may equate to another quid a week.
I wouldn't put anything else in a V6.........Especially with 10 per cent Ethanol in our semi skimmed fuel!

[Ben82]

At the mileage you do Simon, using using the super unleaded may equate to another quid a week.
I wouldn't put anything else in a V6.........Especially with 10 per cent Ethanol in our semi skimmed fuel!

Not really sure I see the relevance in Simon's case... The ES9J4S runs on at least 45% Ethanol without any performance hit, I can't see the ES9J4 being too different (if at all).

[Mandrake]

I wasn't aware that there was any Ethanol in our petrol actually.... nor am I up on exactly what the effects of ethanol in a petrol engine are...

All I know is I couldn't notice any increase in MPG whatsoever with 97 octane, (I keep logs of every fill up) nor did I notice any power increase, although to be fair I can't remember whether I tried it back when the performance was normal or after I was already having issues...

I already spend about £70-£100 a month on petrol so its quite a bump in price to go for 97 octane!

[lexi]

I think it is going up to 20 per cent here shortly. Plastic fuel tanks on motor cycles are losing their shape

FWIW guys I even put the Super fuel in the 61 Landy with a mere 76 bhp and 8;1 Cr. It runs better and I can set it all up on it.
Better day for spannering today Simon?

[Mandrake]

Yes indeed it was, I spent a good 4-5 hours on the car today, most of it was sorting out the leak on the exhaust. Last weeks repair didn't go quite as planned so I dropped the centre section out again, ground out the sleeve that fits the cat tail for a better fit, (which took ages... ) refitted and sealed both the cat joint and the back box front joint which was the one that was actually leaking. So far so good, and much quieter!

I also tried disconnecting the knock sensor and went for a test drive - it didn't quite have the effect I expected, the opposite effect in fact! The ECU obviously noticed the knock sensor wasn't working, the power dropped dramatically and the engine became unusually smooth and quiet, almost like a kitten. It didn't have the normal throaty noise that it has under acceleration - I didn't check with the Lexia but I'm pretty sure the ECU went into a failsafe mode where it retarded the timing, it certainly felt like retarded timing.

It was also making a subtle but unusual noise when opening the throttle wide at lower RPM - which I assume was knocking, although it didn't sound like any knocking I've heard before on older cars, more like a slight groan. I reconnected the knock sensor and everything went back to how it was - more power, throaty engine noise back, strange noise under low rpm wide throttle gone... so the knock sensor is definitely working, although whether it's working properly (and not intermittently) is another question I guess!

I won't be pursuing that line of investigation any further for now though, it was worth checking but I don't think its the issue. I will be doing more testing on the ignition system soon though and double checking the spark plug condition. My gut feeling is I've got an insulation breakdown somewhere in the cylinder 1,5 circuit, cylinder 1 being the one that was producing a quiet spark compared to the others last time I checked, and also the one whose spark plug was unusually clean compared to all the rest, as if there wasn't much combustion going on in that cylinder.... If I'm right cylinder 1 is misfiring, but only under load due to insulation breakdown at higher under-load voltages. I have a few new ideas on how to put it to the test and catch it red handed...

[Mandrake]

I've personally encountered on other cars:

Infiniti V6 - blocked precats
C5 I - blocked cat
Xantia CT - blocked cat; loose pieces then blocked mid-unit
406 - blocked cat

I'm curious to see what your thoughts are now that the exhaust has been found not to be blocked or restricted, when you were so sure it was ? Assuming you're satisfied that my back pressure testing did indeed show no problem...

If it was your car, where would you look next ?

We now know the exhaust isn't blocked. We know the fuel pressure and regulator is ok. We know the injectors pass a balance test.

We know the compressions are good, (at least when the compression tester is connected, there is still a small possibility a plug isn't sealing air tight on the head) the coil pack and plug leads have been replaced, map sensor has been replaced, (but wasn't faulty) faulty oxygen sensor has been replaced, TPS tests ok, knock sensor seems to be working, all other input sensors also seem to be working correctly. (air/water temp etc)

Current clues:

Recent symptoms include intermittent misfiring. When cold the first minute or two during the elevated idle phase the car is clearly misfiring on at least one cylinder based on the uneven engine beat, which gradually goes away after a minute or so. Hot symptoms (other than general lethargic performance and being down on power) include an ugly stumble when snapping the accelerator open from a normal slow idle.

Last time I had the spark plugs out all plugs except for cylinder one had a slight yellow ring around the base of the insulator (pictures previously posted) - apparently this is normal and called corona stain, caused by prolonged exposure to very high voltage electrostatic fields. The fact that plug number one had no corona stain suggests that it has not been subject to as high voltage as the other plugs.

In line with this observation at the same time (a few months ago now) I also did an audible spark test - engine off, using the Lexia to fire individual coils, listening for the sound of the spark. To test the front bank individually I disconnected the plug leads to the rear cylinders and shorted their outputs to ground so there is only a spark jumping to the front cylinders. Cylinders 2 and 3 gave a nice loud spark, cylinder 1 was very muted and barely audible as if there was leakage occurring preventing the build up of as high a voltage. I tried swapping plugs 1 and 2 - no change, cylinder 1 remained the quiet one. I also tried swapping the boots between plug one and two - also no change, cylinder one was still the quiet spark.

I then lifted out the coil pack and measured how far the spark could jump from the bottom outputs of the coil pack to an earth wire - all three outputs jumped about the same distance (I don't recall what exactly, somewhere around 20mm) so as far as I could see the voltage output from each coil is the same.

And that was the point that I left it at the time - I had an unexplained discrepancy with the spark intensity for cylinder one which I'm going to have another go at testing shortly.

The plugs were replaced a few months ago but not with either the original factory type (FR8KDC) nor the type that were previously fitted. (NGK something) I'm a little bit suspicious that the plugs I have are not suited to lots of short trips as they do seem to be fouling. I have a set of the factory specified FR8KDC to go in.

Unknown factors at this point:

1) Although I tested the injectors themselves and the wiring I didn't measure the drive waveforms from the ECU actually firing the injectors to ensure they are all being fired reliably and with the same pulse width. I could do this with my scope relatively easily. A bad crank sensor signal could cause erratic firing of the injectors for example, or there might be a dodgy output driver in the ECU. (hope not...)

2) I also haven't measured the ignition coil drive waveform from the ECU to the coil packs - again a dodgy crank sensor signal could be causing intermittently missing current ramps or erratic timing.

3) The first time I fitted the new spark plugs I under-tightened them and they were all loose when I checked them next time. The second time I think I might have overtightened them...so there is a small possibility of a cracked insulator.

4) There are significant differences between the factory specified FR8KDC and the FR7DC+ that are in there now. The former is a resistor plug with a copper core and twin electrodes while the latter is non-resistive with a single yttrium electrode. I don't know if both plugs are the same heat rating, if the FR7DC+ is a colder plug it could explain the fouling as I do mainly short trips...(EDIT: looking on the Bosch website it looks like the number is the heat range with higher being a hotter plug, so the FR7DC+ is indeed one heat range colder than the correct plug so the FR8KDC is probably a better choice for the short trips I do and could be part of my problem right there)

5) I was never entirely happy with the cleanliness of the surface of the head where the spark plugs seal, but trying to clean at the bottom of the spark plug tower is very difficult. When I change the plugs I'm going to put a lot more effort into actually scrubbing that face clean, somehow...(maybe a toothbrush with a bit of petrol ?) Perhaps one plug is not quite seated properly causing a loss of compression.

Before I take anything out I plan to measure the spark voltage to each cylinder both at idle and under load using an adjustable gap type spark tester - the idea being to poke a small wire under the boot of each of the leads where they come out of the top of the coil pack - then connect the spark gap from ground to each lead in turn adjusting it until it doesn't quite "steal" the spark from the cylinder, this way the voltage can be estimated and compared between cylinders. This would need to be done under load where the spark voltages are highest.

For the front cylinders I'll have to lift the coil pack out but leave the boots pushed into the spark plug towers (in case the problem is a boot breaking down arcing to the tower) and then run a jumper wire to the coil pack, then do the same test with the spark gap.

In theory I should be able to compare the spark voltage of each cylinder under load. (throttle snap, or maybe loading the engine against the torque converter with brakes on) I'm really hoping that this test will show a discrepancy between the voltage to one of the plugs, most likely cylinder one. I have three spare spark plug boots so if I do find one that is breaking down I should be able to replace it.

Regardless of what the spark tests show I'm going to scrub the top of the cylinder heads in the spark plug towers and swap these new plugs in. Fingers crossed that I can make some headway on this issue...

[addo]

I have had a "rough low RPM, excessive fuel consumption" issue in recent months. I could sense it was dropping a spark here and there, also cold starts in the morning took several attempts.

It turned out that the Beru Titan-X plug tips were completely consumed (gaps averaged 2,5mm) and the wells rather full of oil in three instances. That'd do it. My mate at the specialist factors was quite surprised at the short plug life (60K km) but I ascribe that and the oil pooling to when it was badly blocked and blowing madly into the crankcase. Hosed it all into the cylinders with brake cleaner and then cranked it through with the coil plug off.

You have several options in your case. You could possibly have a problem with the ECU, so borrowing a complete set is one possibility. Accessing the Texton box in-situ is always fun. Also, you could try some loaded tests, I am not sure why you discredit these but its no different to an athlete being put through their paces by a coach or selector. You don't say "nice legs; you're hired" except possibly for the women's beach volleyball team. From a scientific point of view, it may be good to focus first on dependably creating the conditions under which a problem occurs.

[Mandrake]

Well the plugs that are in there at the moment have only done maybe 1000 miles at most, so the electrodes won't be eaten away but there is definitely an intermittent misfire, its become more and more obvious in the last month.

Possibly just fouling plugs due to a colder plug in a car that is primarily doing only short trips (ironically the reason for the lack of long trips is because the car hasn't been running well, a bit of a self fulfilling prophecy! )

I agree loaded tests are important, for testing the spark I do need to load the engine to allow the spark voltage to build up - its probably only about 3000 volts at idle but will reach as high as 20,000 volts or more under full combustion chamber pressures... if there is a breakdown at say 10,000 volts on one or more plug/lead it will only misfire under load... that's the type of problem I suspect I'm having.

I was experimenting with using the brake and torque converter to load the engine somewhat - if I put the gearbox in snow mode it starts in 3rd gear in which case the brakes, even the handbrake can easily hold the car back while the torque converter provides a load for the engine - it rises to about 2500rpm and holds there when opening the throttle, further throttle just loads the engine more but holds the same rpm as that's the stall speed of the torque converter I guess...

Probably not a good idea to hold it like that for too long, but I think it should be fine for 5-10 seconds or so at a time which is all that's needed while testing the spark. In theory its no more strain on the gearbox than accelerating from a stop in snow mode where it is trying to accelerate the car away in 3rd gear...

The only other way to do it would be wide open throttle snaps from idle which would give about 2 seconds of high spark voltage to try to measure...

[addo]

Stalling up isn't much of a load on the motor really. It does get the fluid rather hot, which ages it slightly if you do this a lot.

It's not until you really have physical restraint on the car - a heavy load, steep hill or on a dyno (which simulates these conditions - thought about buying a couple of flying runs?) that you get anything like the conditions that really test a spark or qualify gas throughput. The V6 has a lot of poke when running well, and you should be able to bounce off the limiter on a mild slope with four up and a boot of luggage. Don't be afraid to push it; the engine's not made of toothpicks and plasticene.