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

[lexi]

Well done on the injector change mate..........not easy with no shelter or garage

[stevieb]

Thanks Alex, I needed a good laugh today

Agreed

Simon, you say the injectors haven't made a huge difference. Have you re-tried the ICV setting thingy? I have a feeling that'll change things if you try it

[Mandrake]

Well done on the injector change mate..........not easy with no shelter or garage

Thanks, all up it was nearly a 5 hour job including the coolant expansion bottle and nearly an hour wasted while I tried to figure out how to do a makeshift repair to the snapped off breather pipe.

My back and calves were killing me for the next two days I forgot to put the suspension into high before doing the work - the extra height really helps avoid back strain compared to bending over further while the car is at normal height.

[Mandrake]

Simon, you say the injectors haven't made a huge difference. Have you re-tried the ICV setting thingy? I have a feeling that'll change things if you try it

Erm, I'm pretty sure I did. It definitely had a couple of battery off resets. Performance was good but not stellar after the work on Saturday. It wasn't until I filled up the tank that I noticed a sudden improvement hence thinking along the lines of fuel delivery etc.

My thinking with the fuel tank is if the swirl pot chamber is leaky at the valve at the bottom then when the tank is anything but full the fuel level in the pot will be lower than it should be which may introduce air into the fuel rail or drop the pumping capacity of the pump when it's lacking sufficient head of fuel.

While testing the injectors in the shed I noticed that if the fuel level got down a bit in the swirl pot (it was basically running almost entirely off the fuel in the pot with very little in the bottom of the dish) that the petrol circulating in and out of the pot started to get whipped up into a frenzy like a food blender and I started getting air into the fuel rail even though the pump intake was still well submerged, also the fuel pressure gauge clearly showed the pumping ability dropped off with a lot longer "fill" time for the rail when first pressurising.

So I wonder if a combination of a weak pump, leaking swirl pot and an electrical fault causing intermittent low voltage to the pump are all conspiring together to cause a low fuel volume delivery that can't quite keep up under WOT. If its marginal and fluctuates it could explain a lot about why the problem comes and goes so intermittently.

I don't really understand how it can be worst at low RPM and not such a problem at high RPM, but that could come down to the canister purge solenoid programming (maybe it stays shut at WOT at high RPM for maximum power) or simply the programming of the open loop maps. (Maybe at high RPM its programmed to run even richer at WOT than low RPM so has more safety margin in injector times...I've read that a 12 to 1 ratio is normal for WOT compared to a normal 14.7 at cruise)

Whatever the root cause is there is either a misfire or a lean condition, and the O2 sensor reading corresponds EXACTLY with the periods of good and bad performance.

What's interesting as well is that recently I have noticed what I was beginning to think might be a gearbox problem - during acceleration in 2nd or 3rd where the TC is transitioning to lock up (and places a lot of steadily increasing load on the engine) there would sometimes be a momentary fluctuation in rpm (half a second) where the rpm would dip then rise again that felt like maybe a slipping clutch, however on watching the O2 reading the momentary dip in RPM corresponds EXACTLY with the O2 signal dropping lean, the instant it swings rich again the rpm recovers.

So it's actually a brief drop in power output from the engine as the load comes on due to the mixture going inexplicably lean for just a moment. (Purge solenoid activating maybe ?) So I think hunting down this problem of lean running in open loop mode under WOT is the key to solving this mystery, and now that I can easily and accurately monitor the O2 signal on the move I should be able to catch it red handed a lot more easily and see what symptoms are and aren't correlated. At the moment I have the scope blue tacked in the corner of the dash every time I'm out driving. (It only has a 3" screen so it fits nicely in the corner of the dash )

You probably didn't see it as its way back in my thread, but before I replaced my fuel pump I did a delivery test on it - Citroen quote 2 litres per minute at a full 3 bar load. I only measured 1.2 litres per minute which is only 60% of the rated delivery... the replacement pump (also second hand) has not been tested for delivery yet. (If it wasn't such a long day on Saturday I would have done it then) Could I have replaced one old weak pump with low delivery for another old weak pump ? Certainly possible.

For fuel pressure regulators - my original only produced 2.7 bars, the 2nd hand one I got from Stempy which was in the car recently while it was running very well produced 2.9 bars, unfortunately I decided to keep the fuel pressure regulator that came with the second hand rail together with it on Saturday - which only produces 2.8 bars. A mistake perhaps ? Could fuel pressure differences on that order be enough to make it lean (or less rich than it should be) under WOT ? Perhaps this ECU doesn't ever adjust fuel trim for WOT but only goes by preprogrammed settings, thus would be very sensitive to fuel pressure being low ?

I'm going to blank off the canister purge spigot again and see if I regain some of the lost performance from Saturday, that has been blanked off for quite a while now and it usually has run better in that state. Perhaps the extra air flow leakage when the solenoid opens pushes a barely rich WOT mixture to a somewhat lean mixture - the final straw so to speak. Or perhaps the hose between the solenoid and charcoal canister has come adrift so that every time it opens its introducing a large vacuum leak. At least with it out of the picture I'll remove one unknown variable. (I probably should have left it blanked on Saturday so as not to introduce too many changes at once)

[stevieb]

Erm, I'm pretty sure I did. It definitely had a couple of battery off resets. Performance was good but not stellar after the work on Saturday. It wasn't until I filled up the tank that I noticed a sudden improvement hence thinking along the lines of fuel delivery etc.

My thinking with the fuel tank is if the swirl pot chamber is leaky at the valve at the bottom then when the tank is anything but full the fuel level in the pot will be lower than it should be which may introduce air into the fuel rail or drop the pumping capacity of the pump when it's lacking sufficient head of fuel.

While testing the injectors in the shed I noticed that if the fuel level got down a bit in the swirl pot (it was basically running almost entirely off the fuel in the pot with very little in the bottom of the dish) that the petrol circulating in and out of the pot started to get whipped up into a frenzy like a food blender and I started getting air into the fuel rail even though the pump intake was still well submerged, also the fuel pressure gauge clearly showed the pumping ability dropped off with a lot longer "fill" time for the rail when first pressurising.

So I wonder if a combination of a weak pump, leaking swirl pot and an electrical fault causing intermittent low voltage to the pump are all conspiring together to cause a low fuel volume delivery that can't quite keep up under WOT. If its marginal and fluctuates it could explain a lot about why the problem comes and goes so intermittently.

I don't really understand how it can be worst at low RPM and not such a problem at high RPM, but that could come down to the canister purge solenoid programming (maybe it stays shut at WOT at high RPM for maximum power) or simply the programming of the open loop maps. (Maybe at high RPM its programmed to run even richer at WOT than low RPM so has more safety margin in injector times...I've read that a 12 to 1 ratio is normal for WOT compared to a normal 14.7 at cruise)

Whatever the root cause is there is either a misfire or a lean condition, and the O2 sensor reading corresponds EXACTLY with the periods of good and bad performance.

What's interesting as well is that recently I have noticed what I was beginning to think might be a gearbox problem - during acceleration in 2nd or 3rd where the TC is transitioning to lock up (and places a lot of steadily increasing load on the engine) there would sometimes be a momentary fluctuation in rpm (half a second) where the rpm would dip then rise again that felt like maybe a slipping clutch, however on watching the O2 reading the momentary dip in RPM corresponds EXACTLY with the O2 signal dropping lean, the instant it swings rich again the rpm recovers.

So it's actually a brief drop in power output from the engine as the load comes on due to the mixture going inexplicably lean for just a moment. (Purge solenoid activating maybe ?) So I think hunting down this problem of lean running in open loop mode under WOT is the key to solving this mystery, and now that I can easily and accurately monitor the O2 signal on the move I should be able to catch it red handed a lot more easily and see what symptoms are and aren't correlated. At the moment I have the scope blue tacked in the corner of the dash every time I'm out driving. (It only has a 3" screen so it fits nicely in the corner of the dash )

You probably didn't see it as its way back in my thread, but before I replaced my fuel pump I did a delivery test on it - Citroen quote 2 litres per minute at a full 3 bar load. I only measured 1.2 litres per minute which is only 60% of the rated delivery... the replacement pump (also second hand) has not been tested for delivery yet. (If it wasn't such a long day on Saturday I would have done it then) Could I have replaced one old weak pump with low delivery for another old weak pump ? Certainly possible.

For fuel pressure regulators - my original only produced 2.7 bars, the 2nd hand one I got from Stempy which was in the car recently while it was running very well produced 2.9 bars, unfortunately I decided to keep the fuel pressure regulator that came with the second hand rail together with it on Saturday - which only produces 2.8 bars. A mistake perhaps ? Could fuel pressure differences on that order be enough to make it lean (or less rich than it should be) under WOT ? Perhaps this ECU doesn't ever adjust fuel trim for WOT but only goes by preprogrammed settings, thus would be very sensitive to fuel pressure being low ?

I'm going to blank off the canister purge spigot again and see if I regain some of the lost performance from Saturday, that has been blanked off for quite a while now and it usually has run better in that state. Perhaps the extra air flow leakage when the solenoid opens pushes a barely rich WOT mixture to a somewhat lean mixture - the final straw so to speak. Or perhaps the hose between the solenoid and charcoal canister has come adrift so that every time it opens its introducing a large vacuum leak. At least with it out of the picture I'll remove one unknown variable. (I probably should have left it blanked on Saturday so as not to introduce too many changes at once)

That all sounds very familiar. Mine has become noisy again this last week or so. There's a slight drop in responsiveness, though the mpg remains good. Initially the noisiness came and went and I noticed that when it was noisy there was that tell-tale wisp of vapour from the exhaust, even on a milder day. When it was quiet there was no vapour, even at startup. The extra noise is now there most of the time.

Other than the fuel dropping low last week (reported above when it caused a drop in power and a change in exhaust note), the only thing that has changed is that I've stopped adding any Redex (I ran through two bottles in two consecutive weeks). The throttle blip on approaching lights/junctions is also back with a vengeance. I'm convinced this is fuel pressure related, as all these symptoms were there before I cleaned up the fuel pump terminals but eased a lot afterwards. I stuck a big lump of juice in the tank the other day, hoping it might purge out any bubbles in the swirl pot, but it's made no difference yet - maybe it needs brimming to have the desired effect, but as it hasn't been brimmed for well over a year I can't imagine so.

I think I'll try your test Simon, and see what the voltage is at the pump. I think I've asked this, but am I right in thinking that a worn, dragging, or overloaded motor may draw lower volts corresponding with its lower rpm's (albeit matched with an increase in amps instead)? I'm sure I remember from my childhood that if you stall a model train or Scalextric car the motor will draw less volts, but its amperage draw increases (until the power supply overload protection kicks-in and cuts the power). I certainly still have the sticky fuel gauge problem - though the onboard computer's range changes as expected, and the low-fuel light comes on when it should - so maybe it's just a sticky needle and nothing to do with the pump/electrics, but has me feeling very suspicious...

Subject to finances I may invest in a fuel pressure gauge this weekend. I'm curious what mine is, as I've never tested it. This job is secondary to my wishbone bush replacement though - do I have Delphi bushes that have only lasted around 9 months!?!?! Oh yeah! Thankfully I have a spare pair of wishbones that I'm running on right now.

[addo]

Bear in mind that fuel pressure is moderated by the vacuum signal, so to read maximum delivery pressure you need to plug the vacuum line to the regulator.

A gauge isolator is something else that can help; with this you are able to place your gauge in the cabin and monitor it under real load conditions. Otherwise you can tuck it down near the wipers.

[Mandrake]

Bear in mind that fuel pressure is moderated by the vacuum signal, so to read maximum delivery pressure you need to plug the vacuum line to the regulator.

Yep I'm aware of that, those 2.7, 2.8 and 2.9 bar readings were all taken with the vacuum hose unplugged.

A gauge isolator is something else that can help; with this you are able to place your gauge in the cabin and monitor it under real load conditions. Otherwise you can tuck it down near the wipers.

Interesting, do tell about this gauge isolator.

Any sort of adaptor would also need a short right angle joint at the schrader valve otherwise it wouldn't clear the air filter box, and you can't judge engine performance driving the car with the air filter box removed...

Being able to monitor fuel rail pressure while driving under real load conditions may be the only way to conclusively prove whether low fuel pressure under heavy demand is the cause of the lean condition.

[addo]

This one's a bit pricey.

http://www.summitracing.com/int/parts/a ... /overview/" onclick="window.open(this.href);return false;

I'm sure there are Chinese copies available at the right price.

[Mandrake]

This one's a bit pricey.

http://www.summitracing.com/int/parts/a ... /overview/" onclick="window.open(this.href);return false;

I'm sure there are Chinese copies available at the right price.

Yeah, nice idea but WAY too expensive.

[addo]

For you... There will be Chinese copies at one third the price, as dead cert that one itself is Chinese.

My theory is to spend well on the cars; when women want me to cash up for something I legitimately plead poverty.

The PRV is presently decatted, when on song it goes like a scalded moggie - feels suitably raw and vintage. It suffers greatly from alternator fluctuation; I can see the headlights flickering at idle when stopped, and they dim with each sweep of the wipers or a press on the window lift button!

[stevieb]

Time for an apparently dis-jointed update.

Simon, you and I have both noticed variations in our power/mpg etc, right?

My girlfriend is now suffering with it on her 13-plate Toyota Aygo. Her car went into the main stealer for it's 10,000 mile service on Thursday morning, and apart from the obligatory oil and filter change that were due they've apparently done a "fuel injection clean", which we presume is a bottle of injector cleaner chucked down the fuel filler neck. At £12.50 it's a lot dearer than the Redex I've been buying from Asda at £3 a bottle.

Her car is now fluctuating a lot with the available power, changing every time it's been started. Sometimes it struggles to climb from 50 - 70mph on the motorway, yet other times it sails along.

The only other common factor with mine is that she filled up with Shell fuel - mine might be worse on Shell, but I haven't really been paying attention to the fuel I've been using (I try to duck into cheaper garages if I see them on my travels, so I mix brands a lot). I know you use a lot of Shell fuel too Simon.

I wonder if Shell fuel is creating the problem, or maybe the mix of Shell fuel and injector cleaner. I know mine is running VERY noisy at the mo and my mpg has dropped a little bit, but I put in £50 of Shell fuel the other day. Maybe I'll run it low and refill with a different brand next time. It might be worth us considering...

[Northern_Mike]

Well I can honestly say mine gets Asda cheapest 95ron and runs like a dream. It's on 138k, I imagine the 8 previous owners didn't all use Shell V-power.

Sent from my GT-I9300 using Tapatalk

[addo]

I'd be looking for something really obvious. Did they wash the engine bay?

[stevieb]

I'd be looking for something really obvious. Did they wash the engine bay?

Not that we know of as it was immaculate beforehand, but we'll be investigating all the obvious things in the morning along with a check of the air filter (not due for a change 'til 40,000 miles!!! ).

Like I say, I'm wondering if there's a clash between the additives used in Shell fuels and the injector cleaner chemicals. When misbehaving, my own car often feels like I've used low-grade fuel, and the girlfriend's is feeling a lot like that when it's bad. My concern is that it's so much like mine and Simon's symptoms, with similar levels of variability, yet in a barely run-in car. So far Shell fuel is the only common factor....

[Mandrake]

Mike: mine ran mostly on asda 95 for the first few months I had it and ran absolutely fine, however it wasn't until later when I was already having problems that I started trying other fuel. For a while I ran BP unleaded then BP Ultimate and now Shell V-power for the last few months.

If I try to run on 95 octane now the MPG drops by 2-3 MPG and it looses a lot of low rpm performance, but that may be a side effect of it seemingly running lean at low rpm. (Lean mixtures knock more easily so a higher octane fuel would disguise some of the symptoms of running lean)

Steve: funny you say that about V-power, I've been running it exclusively for nearly 6 months now (all from the same service station too) except for one month where the local Shell station was out of v-power for over a week...

Struck me as a bit odd at the time, that they would be out of stock so long, unless there was a problem with their tank....funny thing is it seemed to run better on the BP ultimate I substituted that month, although I didn't think much about it because as you know the symptoms are so variable. Perhaps I'll try going back to BP ultimate for a while, and buying from a different service station.

Knowing my luck my usual service station could have contamination problems in their tank...