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

[Mandrake]

Hopefully it physically fits the tank! If not you may be able to swap the actual pump over between the assemblies.

Is it new or second hand ? I'd be very wary of second hand after this one...

[stevieb]

Hopefully it physically fits the tank! If not you may be able to swap the actual pump over between the assemblies.

Is it new or second hand ? I'd be very wary of second hand after this one...

A brand spanker

As I say, the parts catalogue lists it as fitted to the same tank as mine so I can't see any reason why it might not fit. I'm curious whether there was a spec change between the D8 variant and the one fitted to mine (a very early D9 with a lot of D8 parts installed). I can't imagine what though...

[Mandrake]

Nice. Fingers crossed it does indeed fit, and of course, let me know the results.

[stevieb]

Nice. Fingers crossed it does indeed fit, and of course, let me know the results.

I certainly shall, though it may be a while before I can be certain, depending in how my working days pan out - ideally I need to repeat some of my recent jaunts out for a fair comparison. If heat is the issue then the warming weather will certainly exaggerate the problem if it turns out it isn't the pump that's at fault.

I'm car-dependent for work, so at least I'll have the peace of mind of knowing the pump is new, even if it doesn't fix anything.

Now, I must remember to pop to the local main stealer and order a new seal/clamp tomorrow...

EDIT - Trip to Derby booked for Thursday morning. Derby seems to have become my regular test-run as the car always runs great on the outward journey, but after a 30-40 minute stand is often thirsty afterwards. The ever-lengthening stretch of 50mph roadworks on the M1 are a good indicator too. Fingers crossed the pump lands tomorrow now. Apparently it's been despatched....

[Mandrake]

I've talked a lot about the quoted delivery specifications for the fuel pump but I'm not sure I've ever posted the actual file so here it is:

https://dl.dropboxusercontent.com/u/782 ... System.pdf" onclick="window.open(this.href);return false;

As can be seen on page two, the specifications are very unambiguous - 120 litres per hour at 3 bars, which is 2 litres per minute, so 1.2 and 1.3 litres per minute measured is well short of the spec.

Another thing I've just realised is they quote a wattage for the pump, from which current can be derived. It's not clear whether this 50 watt figure is at 12 volts, or more likely a nominal 13.8 volts, in which case the current is 4.2 amps or 3.6 amps respectively.

When I was testing my old pump in the shed with the injectors I was running it off a 13.8 volt 5 amp regulated supply I built many years ago - which has current fold-over at 5 amps. Well guess what - that power supply can't run the pump without the current fold-over activating and shutting down. The power supply has a button I can press to temporarily disable the shut-down that occurs beyond fold-over to allow loads with a high "surge" current like a motor to start running and let the current fall below the fold over current. So if that pump was drawing 3.6-4.2 amps it should have kept running after I released the button, but it would not. (The power supply would shut down, indicating the pump was still drawing more than 5 amps)

For some reason I didn't actually measure the current the pump drew as I assumed at the time it would normally be more than 5 amps. Interesting - I really need to measure the current drawn by the pump on the car. If its a lot higher than normal that too is a sign of a problem...

[stevieb]

No new pump yet. I'll bet it arrives tomorrow whilst I'm outta town with a lot of miles to do

Some more info on the existing pump though, as my trip to Derby this morning may have been revealing.

Firstly, the cold start was great - smooth, quiet, efficient. About halfway to the motorway from here (all free-running roads) is a fairly steep hill where you can get your boot down. I did that this morning and the car flew. However, I noticed that where the hill levels off and approaches the motorway, the mpg didn't rocket up to 999.9 as I braked, and there was some power loss and extra exhaust noise. If during that blast the engine was drawing more flow than the pump could deliver (below the 2.0 rating, but higher than the real-world figure), then it stands to reason that the pressure in the fuel rails will drop and performance will suffer until it recovers... I should remind everybody that this change is marginal - I only notice it because I know the car so well and notice when things are different.

Once on the motorway and in the 50mph roadworks the noise faded, power returned, and the mpg rose again. It was a smooth but quite noticeable transition.

The 50mph in the roadworks didn't last long and I hit traffic - mostly crawling. By the time I'd got off the motorway the engine bay temps would've been up, and unsurprisingly the power/noise/mpg issue came back, albeit a lot less obvious. The free-flowing A38 showed a slight improvement before I hit traffic again. The slower pootle into Derby city centre was the same - again, only very slight, but difficult to judge due to the nature of the roads in that area.

Departure from Derby about 35-40 minutes later and it was usual warm-start territory - reduced power, increase in noise, loss of mpg, more right boot required - 50mph in the M1 roadworks was 25-35mpg as opposed to the 45-65mpg I'd had after the earlier cold start. Overall power loss in these situations is minimal - rapid acceleration is still there - but it takes a LOT more right boot with matching fuel consumption and a slight turbo-like lag before it kicks-in.

Fast forward to this evening and a quick check of something else in the RH rear wheel arch with the engine running and I notice the fuel pump is very noisy. At least I don't remember it being that loud before. Fuel was low though. I stuck £20 in to tide me over (I want to keep it low ready for the new pump) and the fuel gauge didn't rise - it took a good ten minutes before it crept slowly up to where it should've been - not one big step, but slow creeping.

I'm wondering if there's an electrical fault - whether in the pump or elsewhere - and whether affected by heat or not - but I've certainly had gremlins with the fuel gauge not rising or falling in the past, not to mention the corroded harness/fuel pump connector (which is clean but now showing signs of corrosion/arcing). Something is definitely amiss there. Now to see if this new pump makes any difference when it arrives. I can see all this turning out to be a bad earth...

[Mandrake]

No new pump yet. I'll bet it arrives tomorrow whilst I'm outta town with a lot of miles to do

Typical that it didn't arrive in time. I know that feeling of having a decent length trip coming up and wanting to test/observe something on that trip (rather than needlessly wasting petrol on a test drive) but not having it ready in time... quite infuriating...

If during that blast the engine was drawing more flow than the pump could deliver (below the 2.0 rating, but higher than the real-world figure), then it stands to reason that the pressure in the fuel rails will drop and performance will suffer until it recovers...

An interesting fact - I did a bit of digging to try to find out just how much fuel volume is really required for WOT on a 3 litre v6 - I found a couple of resources with calculators that will calculate this for a NA engine based on maximum RPM, volume displacement and/or maximum horsepower.

And it turns out that for a 194HP 3 litre V6 running at WOT at 6000 rpm it takes about 0.95 litres per minute just to reach a stoichiometric mixture. Factor in that at WOT it will (should be) running at about 12 to 1 not 14.7 to 1, and the demand rises to 1.16 litres per minute. Then you have to factor in a healthy margin for the pressure regulator to remain open so that fuel pressure is still accurately regulated otherwise pressure will drop and the mixture will lean out.

Also when you put your foot down quickly the vacuum signal to the regulator demands that the fuel pressure rise very quickly, but you need a lot of excess delivery to allow the pressure to rise quickly (near instantly) when the demand from the injectors is simultaneously high. (All 6 injectors opening together for a fraction of a second) If delivery is marginal the pressure rise when manifold vacuum drops suddenly will be slowed down and cause a momentary lack of power / hesitation.

So if my original pump was only managing 1.2 litres per minute (and current one 1.3) when they're COLD, that's very marginal indeed when WOT at high RPM is going to demand up to 1.16 litres per minute...

This 0.95 to 1.16 estimate also fits in well with what I've read that a fuel pumps delivery normally has a 2 to 1 headroom - if it requires up to 1 litre/minute the pump will be rated at 2 litres/minute. Plenty of excess delivery is key to allowing the regulator to effectively regulate the pressure.

1.2 litres/minute is also below all the "rules of thumb" I could find for minimum pump delivery when testing fuel pumps. They all suggested a minimum of at least 1.5 litres / minute with anything less than that indicating a fault.

Departure from Derby about 35-40 minutes later and it was usual warm-start territory - reduced power, increase in noise, loss of mpg, more right boot required - 50mph in the M1 roadworks was 25-35mpg as opposed to the 45-65mpg I'd had after the earlier cold start. Overall power loss in these situations is minimal - rapid acceleration is still there - but it takes a LOT more right boot with matching fuel consumption and a slight turbo-like lag before it kicks-in.

That "turbo lag" symptom is interesting - last year I was getting that really badly, very frequently - you could literally snap the throttle down half way pulling out of a roundabout and have a half a second delay before any power appeared - and you could clearly hear it in the engine note, where the increase in engine noise that you would expect when producing power was delayed by the same amount, and kind of ramped up slowly over half a second. I haven't seen that happen recently, at least not in that extreme fashion, there may be a small amount of lag to throttle response sometimes but nothing too major.

I don't know which of the many changes I made cured (or minimised) that symptom or when it changed, except that it was approximately 6 months ago. I have a hunch what the problem might have been but I'll know more soon...

Fast forward to this evening and a quick check of something else in the RH rear wheel arch with the engine running and I notice the fuel pump is very noisy. At least I don't remember it being that loud before. Fuel was low though. I stuck £20 in to tide me over (I want to keep it low ready for the new pump) and the fuel gauge didn't rise - it took a good ten minutes before it crept slowly up to where it should've been - not one big step, but slow creeping.

With my original fuel pump, during the precious few very hot days we had last summer it started making a noise that was loud enough for me to hear it from the drivers seat, windows all closed, over and above road noise driving at 40mph.

At first I panicked thinking it was the gearbox failing because of the history of gearbox symptoms I've had with this car - it actually sounded like it was coming from somewhere in the engine bay, and it was kind of a whining/whirring noise mixed in with the sound of a dry bearing. While stationary I couldn't hear it but once I exceeded about 30mph it started to get louder and louder and reached a peak at around 50mph. Then when I'd slow down below 30mph it would slowly fade away until I couldn't hear it again. It didn't seem to change pitch with engine or road speed (which is why I finally dismissed the gearbox as the source) but seemed to get louder with fuel demands. It's very likely that the tank level was fairly low at the time too.

It wasn't until the following day when I started the car with the drivers window open and clearly heard the same noise at idle that I realised it was the fuel pump making the noise and that (while stationary) I was hearing it reflecting off the pavement and in through the window. This noise was one of the reasons I decided to replace my original pump as that kind of noise can be a sign of the pump getting near the end of its life - either brushes/commutator or bearings failing.

It only did it on the really hot days though, as soon as we had a few cooler days the noise went away, or at least dropped back to a more normal level where you couldn't hear it while driving the car, only if you placed your head near the ground near the pump. So the fact that a noisy pump can get noisier when hot (and under load) is quite telling I think.

I'm wondering if there's an electrical fault - whether in the pump or elsewhere - and whether affected by heat or not - but I've certainly had gremlins with the fuel gauge not rising or falling in the past, not to mention the corroded harness/fuel pump connector (which is clean but now showing signs of corrosion/arcing). Something is definitely amiss there. Now to see if this new pump makes any difference when it arrives. I can see all this turning out to be a bad earth...

The pump motor and level sensor don't share the same earth return, so its unlikely that the same fault is affecting both. On a RHD Xantia the earth return for the pump goes through the harness along the base of the drivers door and earths somewhere up behind the dashboard of all places!

It sounds like you need a new level sensor too (which will come with the pump assembly) so changing the assembly over should fix your fuel gauge reading as well. Have you not checked the voltage/earth to the pump before ? I've checked mine and the earth is fine, although there is a little bit more voltage drop in total (mainly on the positive side) than I'd like to see. On the other hand if the pump is drawing excessive current (which it seems to be) that will cause additional voltage drop as the wiring in PSA cars is fairly marginally rated!

[Mandrake]

An interesting little fuel pump related anecdote for you Steve.

Tonight I drove about 20 minutes on a mix of motorway and urban with the car running reasonably well, although still showing clear lean running symptoms on the O2 monitoring scope.

The car was then stopped for almost exactly half an hour before I attempted to start it. Upon first attempting to start it I could tell it was going to stall. It struggled for a couple of seconds after starting then stalled. I waited a few seconds and tried again, it came to life, stumbled a couple of times then stalled again. A third time and it did the same, finally the fourth time it managed to keep going and idled roughly for about 10-15 seconds before it finally started idling smoothly.

It seems likely it was suffering from vapour lock after sitting for half an hour, and it will be the faulty leak back valve that allowed the vapour lock to occur in the first place. Also very interesting is that each time the car stalled I could clearly hear the fuel pump running on for about 5 seconds, slowly spinning down in speed with a dry noisy bearing type sound.

After the engine finally did keep running the "dry" noise of the pump gradually faded away as the car started running properly again. Doesn't seem good for the fuel pump does it ?? At the time the tank was a bit over half full so starvation of the pump shouldn't have been possible.

I recall observing a similar dry bearing type sound on previous failed starts, and I wonder if the vapour locked mixture of fuel and vapour being pumped back to the tank caused the pump to start ingesting a lot of air bubbles until the vapour lock cleared ?

[stevieb]

Very interesting reading Simon. And no, I never have tested the volts at my pump - I'm glad to hear the sender and pump share different earths though - it eases my mind a little that this might all just be an earth fault.

The stalling after warm start sounds familiar, except that in my case it sometimes hunts a lot at start-up. It's never stalled, but I get the impression the anti-stall is desperately trying to stop it happening. It doesn't happen often, but often enough to make me wonder what's caused it. It's not consistent with problems though - sometimes it warns of bad-running, sometimes not.

I should say the turbo-lag symptoms are very minimal on mine - not horrific, but noticeable for a normally aspirated engine. I always know when it's going to happen because the onboard computer reports a massive drop in mpg as the revs begin to climb. On a healthy run the mpg doesn't plummet anywhere near as much. The lowest mpg the computer can display is 9.4mpg - you know it's bad when booting it in fifth shows 9.4mpg on the display, because the reality might be even worse...

My experiences today might add some weight to things. A cold start run to Hull was okay, but not great. (I should point out that my journey starts with gates open, car start and pull out of gates, close gates, drive off. I think some mornings the time it takes me to close the gates makes a difference, but maybe it's just me...). Anyway, my journey out of Hull was interesting - a warm start with the usual slight loss of power and mpg via a petrol stop. I got to the 70mph A63 and it was very thirsty - 25-30mpg at 70mph. Then I tried something. Assuming fuel supply or pressure is the issue I dropped it into neutral for five seconds, the logic being that I was taking some load off the pump and giving it chance to add some pressure back to the fuel rails. Back into fifth gear and suddenly I've got 35mpg at 70mph. This lasted around ten minutes before it started to fade again. During this ten minutes any attempt at increasing speed was met with a sudden drop of mpg. Easing off the throttle didn't give a matching increase in mpg.

A later run out of Bradford was similar. Usual symptoms on a warm start, but then a short steep downhill stretch where I left it in second and forced some strong engine braking. Suddenly things improve for ten or fifteen minutes of my journey up to Harrogate...

Pressure/supply-related? I think so.

[stevieb]

New pump fitted after a 70-odd mile journey back from North Yorkshire. And the conclusion is...

Nobody really wants to know, do they? Surely not....

[Mandrake]

Let me guess, no change ?

[Northern_Mike]

The latest version of Tapatalk on the Samsung S4 shortens the title of this thread to "Xantia V6 Broken"

Sent from my GT-I9505 using Tapatalk

[Northern_Mike]

The latest version of Tapatalk on the Samsung S4 shortens the title of this thread to "Xantia V6 Broken"

Sent from my GT-I9505 using Tapatalk

Thought your posts here had been blacklisted NM

Are people really that small - minded these days that they can't take a bit of light ribbing?

Sent from my GT-I9505 using Tapatalk

[Mandrake]

New pump fitted after a 70-odd mile journey back from North Yorkshire. And the conclusion is...

Nobody really wants to know, do they? Surely not....

C'mon Steve, spill the beans. Did it help or not, or is it too early to tell given the intermittent nature of the symptoms ?

[stevieb]

Hehehe, sorry about the teasing - it just had to be done.

Sadly today's planned trip out has been cancelled due to the clamp/joint between the cat and exhaust beginning to let go (the mild steel of the cat is rotten, so it looks like a new cat is required if I can't get a section welded in - barely worth it though, is it?) so I can't confirm any more than what I learned on a 10-mile round trip yesterday. Sadly there's no exhaust hanger in this area, so if this joint breaks the front end of the exhaust will drop and I don't want to think about the consequences of that happening...

With the new pump installed, right after a 70-odd mile journey (so everything except the pump would've still been warm), after a good bit of spinning to purge out the air, it started up perfectly. It then ran perfectly. It accelerated great. Engine braking was perfect. Mpg was fab. Basically the car felt like I think it's supposed to. The return journey was a couple of hours later, so it'd had time to cool - so I didn't have chance to confirm things were good with a full-on warm start with the new pump, but it ran great again.

Problem solved methinks...