I think it's highly dependent on the engine, and how smart (or dumb) the ECU is, and I wouldn't like to speculate too far beyond the ES9J4.xantia_v6 wrote:I have ordered a pair of these (one for the XM and one for the Xantia). At least that gives me free postage...
Would it be reasonable to assume that these (or equivalent higher pressure regulators) would be just as effective on other petrol engines of the era?
I think it works on this engine precisely because the ECU is pretty "dumb" by modern standards. Most ECU's these days have long term fuel trim and short term fuel trim. The short term fuel trim is from immediate feedback from the oxygen sensor and is only used in closed loop mode (oxygen sensor in active use) while the long term fuel trim is learnt over time based on the short term fuel trim being out for a long period of time, and applies in both open and closed loop modes.
For example say that the engine is running lean at idle due to Ethanol. When the oxygen sensor is warmed up it will notice this and the ECU will quickly make a correction to the short term fuel trim of say +5%. If you have long term fuel trim then this 5% short term trim being permanently required at idle would eventually (it takes a few minutes or more) cause the long term fuel trim to creep up to 5% while the short term creeps back down to 0%. The job of the long term fuel trim is to bring the short term to 0%. In open loop mode only long term fuel trim is used, in closed loop mode they are added together.
However what I've noticed on the ES9J4, is that it behaves as if it doesn't have a long term fuel trim memory! (Or if it does, it doesn't work very well) An example of this is in the above example it should learn 5% long term fuel trim for idle, then the next time you start the car, the long term fuel trim should cause the idle mixture to be correct before the oxygen sensor warms up and starts working (which takes 90 seconds) but it doesn't!
Before the fuel pressure change it consistently starts up and idles lean until the oxygen sensor warms up and starts making corrections. (Both my V6's have done this) You can see this by monitoring the oxygen sensor output on a scope. Generally in open loop mode the engine should err on the side of running slightly rich, but you can see from the oxygen sensor output as it starts to warm (but before it starts switching) that it starts off somewhat lean. After the pressure change it starts of slightly rich before it starts switching. I have noticed an improvement in initial running during a cold start as well.
It seems like it only has short term fuel trim. If this is true then it means any time you are in open loop mode (including idling for the first minute or so while the oxygen sensor warms up, WOT, or momentarily when you press the throttle) it is using pre-calculated fuelling calculations which don't include anything "learnt" by the oxygen sensor. So if there is a systemic error in mixture due to say Ethanol the open loop mixture will always be a bit leaner than it ought to be.
See above. In closed loop mode the ECU will compensate for the increased fuel pressure by reducing injector firing times. Closed loop mode is active when idling (but only after the initial 90 second warm up time of the heated oxygen sensor - idle will be richer now during that initial warm up) and when holding a steady rpm at a light to moderate throttle, eg cruise. So an emissions test is not affected by this change as the test is performed at idle and a steady 2500 rpm under no load.Deanxm wrote:Im wondering if the effect of higher pressure is greater fuel delivery (which i doubt) or better atomisation, the ecu must be getting the mixture right at the stock rail pressure or it would be running lean on the stock setup which would be compensated for.
Open loop mode occurs during initial startup, during WOT or high load, and momentarily every time you press the throttle, especially from idle.
Those that remember carburettors will remember the throttle pump that gives a squirt of fuel when you press the throttle, without it the sudden increase in air flow without an increase in petrol will cause a lean hesitation/stumble, especially if you press the throttle quickly from idle.
On a fuel injected car that same function is provided by the throttle position sensor. Every time the ECU sees the TPS open further, especially if it moves quickly it instantly drops into open loop mode and gives a large pre-calculated squirt of fuel to prevent a stumble. After about a second if you hold the throttle steady it will go back to closed loop again using the oxygen sensor, unless the throttle is open enough to constitute a heavy load - if its a heavy load it will stay in open loop.
If you increase the fuel pressure you increase the fuelling in open loop mode, basically. So WOT will be richer, initial startup and idle will be a little richer for about the first 90 seconds, and the "throttle pump" like squirt of fuel that you get every time you press the throttle will also be a little richer. Mixture during warmed up idle and cruise will be exactly the same as before.
Going from 3 bars to 3.5 in theory will make open loop mode about 8% richer, whilst E10 petrol should in theory make the car run about 5% too lean, so it is over compensating a little bit, but on the other hand Citroen probably tune the engine on the economy side anyway, so you're getting a little bit more performance than standard at the expense of slightly less economy - but because its only richer during acceleration it shouldn't affect cruise MPG significantly.
Hope that explains what's going on. Your point about atomisation is a good one - I believe that an increase in pressure will improve fuel atomisation as well (up to a point) and its probably no accident that a lot of cars use 3.5 bars or even 4.0 bars as their standard fuel pressure. They actually seem to be more common than 3.0 bars from what I have seen.