Xantia V6 fuel pressure regulator upgrade.

[xantia_v6]

There is a difference in ECU for the ES9J4 between Automatic and Manual Gearboxes though, apparently the mapping around the idle region is a bit different as a few people have reported that using an Automatic ECU in a manual Activa V6 conversion (such as David's) can result in hesitation when accelerating from idle. Makes sense when you consider launching from stationary is a very different load for the engine on a manual gearbox with a clutch than an auto with a torque converter!

Also the Automatic ECU communicates with the gearbox ECU and has extra programming for that, whilst the manual ECU has no gearbox ECU to communicate with...

The ES9 ECU for manual transmission also has no fuel cut-off on overrun. The trip computer always shows some consumption when coasting downhill, whereas on an automatic car, the consumption drops to 0.

[Mandrake]

Ive been all over the management system, TPS is fine, ICV new and at great exense for the big Bosch units (did make some differnce to idle stability) plugs new NKG, leads are old but good, coil pack fairly new as is the Ignition amp, throttle body doesnt get dirty at all but all the tiny air passages from the crank breather etc have been cleaned because after 24 years there was some dry crusty bits in them. Its just a bit of a stumble if you dab the throttle too hard when pulling away and not all the time these days, most people wouldnt notice.

Have you checked the fuel pump delivery ? Somewhere there should be a specification for the fuel pump that it can deliver X litres per hour at the specified pressure. For example the pump on the Xantia V6 is specified as 120 litres per hour at 3 bars.

When I was chasing engine performance problems on my first V6 I measured the delivery of the pump - to do so I carefully disconnected the return line before it goes back into the tank and fed this to a 1 litre measuring jug, and rigged a way to run the fuel pump continuously while the engine is not running that I could switch on and off easily. By using the return line you ensure that the pump is running into a 3 bar load because the pressure regulator is still in the flow path - if you just pumped directly from the pump into a jug the test is not valid because it is not under load...

I found that the original pump could only produce 60% of the rated delivery under load! I ended up getting a pattern part pump assembly made by Bosch for a reasonable price and that measured about 90% of the claimed volume delivery so was an improvement.

Another test you could do is assuming that the pressure regulator is a vacuum type, disconnect the vacuum hose to the regulator, put a fuel pressure gauge on the rail (hopefully it has a schraeder valve otherwise you'll have to Tee it in...) and then with the engine idling floor the throttle up to about 3000rpm or so whilst watching the pressure gauge.

It should not drop at all, although it might flutter either side of the reading slightly due to pressure pulses. If it sags significantly under this quick full throttle load the fuel pump is probably sick. It may have enough delivery for normal running but let the fuel pressure sag momentarily when you snap the throttle, thus a stumble.

I will be fitting a tct rail with integral reg for other reasons, it wont hurt to up the pressure and suck it and see, it either helps, makes no difference or makes it worse and there is only one way to find out, after 24 years a strip down and clean of the injectors would probably help more though, maybe.

If you've checked all those other things and tried the pump tests above and still cant find anything, then I agree - why not give it a go. A modest increase in pressure is not going to hurt anything.

I didnt think about the Ethanol content, are we sure they dont use ethanol in high octane fuel? i would guess they cant as its got a lower calorific value and would compromise the claimed benifits of premium unleaded but are they not forced to use a certain % of renewable ethanol in the fuel by some directive?.

Some high octane fuel in the UK in some parts of the UK has no Ethanol. Some high Octane fuel in other parts of the UK has at least 5% Ethanol in it - that's all we really know for sure.

The south west of England and all of Scotland seem to have more Ethanol in high octane fuel than elsewhere, apparently due to the output of the nearby refineries. Being in Scotland I assume I get the higher Ethanol quantities.

[Mandrake]

The ES9 ECU for manual transmission also has no fuel cut-off on overrun. The trip computer always shows some consumption when coasting downhill, whereas on an automatic car, the consumption drops to 0.

Interesting - are you sure about that ?

I don't doubt what you've seen on your trip computers, but the overrun cutoff is RPM dependent, and the different gearboxes may lead to a different RPM pattern that doesn't trigger the cutoff as easily.

Specifically, on the ES9J4 (at least the Auto!) it will only cut off the injectors if you are doing over 1600 RPM when you close the throttle. If you close the throttle at 1500 rpm or less the injectors will continue to fire a small amount to maintain a stoichiometric ratio. If you are above 1600 RPM when you close the throttle and the car slows down, the injectors will remain cut off until you drop to 1100 RPM at which point they start firing (a small amount) again. Once they do, they will not cut off again until you exceed 1600 RPM then close the throttle again.

I've seen this documented somewhere and I have confirmed it on an Auto while monitoring the oxygen sensor output on a scope - its very easy to see when the injectors are in overrun cutout because the O2 reading will go hard lean and stay there, then when they kick back in it will start switching back and forth rich and lean again.

Seems odd that they would disable this on a manual...

Interesting point here is that you can actually save a little bit of fuel by coasting down hills in 3rd instead of 4th, at least at certain speeds. In 4th the gearbox tends to keep the torque converter unlocked at lower speeds (say 40mph) which means as soon as you take your foot off the throttle the engine drops to more or less idle speed while you coast, which causes the injectors to fire as you have dropped below 1100 RPM. However leave it in 3rd at the same speed and the torque converter will be locked up (applying some engine braking) and will keep the RPM above 1600 due to that engine braking - the injectors will completely cut off when you remove the throttle and coast down the hill as it stays above 1100 RPM!)

[xantia_v6]

The reason that ECUs for manual transmission usually don't have overrun cut-off is that it can cause jerky-ness under circumstances when the engine operating point is near the throttle cutoff point (e.g. driving down a very slight incline). I have seen this written by Roger Bywater (who worked for Jaguar development before setting up his own re-mapping business), and have seen the difference between my Xantia automatic (which had a trip computer fitted) and my XM manual, which I owned concurrently for about a year.

[Mandrake]

Ok that makes sense. I did wonder if it could be something like that but glad to see it confirmed by someone who works in the field.

I can see now why those doing RHD Activa V6 manual conversions are adamant on getting the manual ECU, even though the Automatic ECU will "work" it would lead to a more jerky driving experience.

[DHallworth]

I've just had an email from Amazon telling me it's been dispatched and will be here tomorrow

Shame it wasn't in time for the weekend but now I've got the exhaust to do anyway it'll get fitted next time it's on the ramp.

David.

[Mandrake]

Did the trip down south blow a hole in the exhaust ?

[DHallworth]

There was the smallest of pin holes in it before I set off... It sounds now like you could put your fingers through it... It's gotten a bit loud.

David.

[DHallworth]

The first pressure regulator I ordered has arrived now. I've just ordered another two, one for the XM and one for the 406 Coupè.

At that price for genuine Bosch ones I might as well.

Will have a shot at fitting it at some point.

David.

[xantia_v6]

I have now fitted a 3.5 bar regulator to the XM.

Before fitting the regulator I went for a drive of about 4 hours, mostly on the cruise control at 130 km/h. The trip computer indicated an average of 9.6 l/100 km, which is completely typical of the consumption of this car. Note that I had filled the tank with 98 RON, which in France is supposedly ethanol-free.

After fitting the 3.5 bar regulator, I took the car along the same route (not quite as far, so only 3 hours driving, but with a bit of additional town driving in the middle), and the trip computer indicates 9.3 l/100 km. This initially looks like a 3% improvement in economy. But...

The trip computer calculates consumption just from the injector pulse width, not taking into account the change in fuel pressure. I believe that the pressure increase should be increasing flow rate by about 8%, so adjusting the reading for that maybe economy is reduced by 5%. But...

Given that the system would have been running in closed-loop mode almost all of the time, I would not really be expecting the economy to be changed at all. Maybe the bit of around-town driving had the effect, or maybe I have not taken something into account?

[Deanxm]

Economy, pfft. What about drive ability, throttle response, low/top end power? Any improvements or no noticeable difference?

D

[xantia_v6]

I did not really notice much change in the way that the car drives. Maybe the idle is a little smoother, but as the car is currently running on 98 RON (with supposedly no ethanol), maybe there is no improvement to be expected.

[Deanxm]

Fair point, did you notice a change between high octane/premium and 95ron fuel before the increased fuel pressure? could be the gains are more to do with a failing pressure regulator deliving below spec rather than just inreasing fuel pressure above spec if you see what i mean.

D

[Mandrake]

I have now fitted a 3.5 bar regulator to the XM.

Before fitting the regulator I went for a drive of about 4 hours, mostly on the cruise control at 130 km/h. The trip computer indicated an average of 9.6 l/100 km, which is completely typical of the consumption of this car. Note that I had filled the tank with 98 RON, which in France is supposedly ethanol-free.

After fitting the 3.5 bar regulator, I took the car along the same route (not quite as far, so only 3 hours driving, but with a bit of additional town driving in the middle), and the trip computer indicates 9.3 l/100 km. This initially looks like a 3% improvement in economy. But...

The trip computer calculates consumption just from the injector pulse width, not taking into account the change in fuel pressure. I believe that the pressure increase should be increasing flow rate by about 8%, so adjusting the reading for that maybe economy is reduced by 5%.

Yes that's correct. Changing the fuel pressure will affect the accuracy of the mpg reading on the trip computer for the reason you state - in theory it would now claim that your MPG was 8% better than it really was. No trip computer on a Xantia to worry about of course but something to keep in mind on an XM. My MPG figures are taken by the old fashioned brim the tank method so wouldn't be affected.

But...

Given that the system would have been running in closed-loop mode almost all of the time, I would not really be expecting the economy to be changed at all. Maybe the bit of around-town driving had the effect, or maybe I have not taken something into account?

I would expect economy to get a little bit worse in general - it should be exactly the same when cruising with a constant throttle, (in closed loop) but how often do we do that ? Every time you overtake someone you are accelerating and momentarily going into open loop. (Depending on how much you depress the throttle it will either come back to closed loop again in about a second or stay permanently in open loop if the acceleration is hard)

It goes without saying that city driving with frequent speeding up and slowing down would put it into open loop quite frequently so would suffer more in terms of economy.

Interesting that your trip computer is (after allowing for the fuel pressure change) reporting an average economy reduction of 5% despite an 8% increase in open loop mixture - the discrepancy is presumably the proportion of time that the car is in open loop versus closed loop.

In my own case I'm finding that my tank to tank MPG is varying far more than the few percent that I might expect from the mixture change, just due to different driving conditions and styles. The first fill up after I changed the pressure I got 20 mpg on my normal work day grind, which is a very fuel inefficient 5 minute trip in the morning and evening through residential areas. In the past I have only ever got 18-20MPG in these circumstances.

The next fill half the tank was used with an approx 150 mile road trip and the other half back to the work day grind, and I managed 24mpg there. The latest fill I only got 18mpg on my work day grind, but I have to admit that I have been driving in a rather sporty fashion lately... the car is running so well that I can't help myself and find myself frequently pulling away from junctions unnecessarily quickly just for the hell of it. If I was to drive a bit more conservatively I'm sure MPG would go up a bit.

Anyway, I didn't increase fuel pressure with the thought of improving fuel economy - I did it to try to squeeze a little bit more performance out of it and for me it seems to have worked. The V6 is so heavy on fuel to begin with that a few percent doesn't really make much difference...

I did not really notice much change in the way that the car drives. Maybe the idle is a little smoother, but as the car is currently running on 98 RON (with supposedly no ethanol), maybe there is no improvement to be expected.

Hard to say - if there's no Ethanol in your fuel you may not see much benefit. I don't think it's possible to get Ethanol free 98/99RON fuel in parts of the UK like Scotland so maybe it helps more here. I'm happy with the way the car is driving so I'll be leaving it in, but maybe it's not for everyone.

Have you tried a battery off reset of the ECU ? Just disconnect the battery for more than 10 minutes then reconnect. On the Xantia I can just remove fuse F14 in the drivers fuse box to remove the permanent live to the engine ECU to achieve the same goal without disconnecting the battery, but I don't know what fuse it would be on an XM.

I still find that after an ECU reset like this that the engine is much more nippy to begin with then settles down after a few miles - had the same experience with the previous V6 as well, I'd love to get to the bottom of the real cause for this. The increase in performance after a reset (and the subsequent fall off when the effect "wears off") is less with the higher fuel pressure but there is still a quite significant change. Something that the ECU learns over time through feedback causes performance to drop eventually. (I think it's knock sensor related, and possibly "false knock" detected by the sensor but don't know for sure and don't know how to prove it - it could be a quirk in the engine or ECU design)

[Stickyfinger]

Reports say the South West is all Ethanol added as well.

I have to say the car is running exceptionally well at the moment. My recent trip to Bovington (despite the coolant leak) was most entertaining with the car responding very well. The "pull" at low Revs does seem improved and I have found myself staying in the higher gear more at low speed as it feels smooth pulling away.