high pressure during cut off

[mgrover]

I guess lets assume the worst fuel trim I have on petrol in terms of LTFT is 10%, maybe 15%. Then I would of assumed that the lpg injectors would simply copy the petrol ones. But when I do run LPG it shoots off past 20. For idle anyways. Currently am interested in just making idle work well and I assumed auto calibration would do that but no dice.

My injector timings have a clear distinction between 1-4 and 4-8 since each bank has its own trims.

[Gilbertd]

If petrol long term trims are that far out, it's running lean on petrol. I would suspect an air leak in the intake somewhere.

[mgrover]

the intake has been tested and retested several times with a smoke machine, no leaks at all.

[Brian_H]

Has anything been done with the MAF? And is it a genuine unit, or something else?

[mgrover]

It did used to be a crappy Chinese unit which caused lean codes on petrol. I swapped it with a 2nd hand genuine one.

Now at idle it reads 5g/s and 2500rpm 20g/s so within range. I also ordered a separate one made by Toyota but supposedly fit. That was fine to but I reverted to the 2nd hand one.

I don't really fancy chucking money at a maf that may or may not fix issue.

I do wonder if the plastic intake boot is knackered. Will have to figure a way to smoke test it

[LPGC]

I guess lets assume the worst fuel trim I have on petrol in terms of LTFT is 10%, maybe 15%. Then I would of assumed that the lpg injectors would simply copy the petrol ones. But when I do run LPG it shoots off past 20. For idle anyways. Currently am interested in just making idle work well and I assumed auto calibration would do that but no dice.

My injector timings have a clear distinction between 1-4 and 4-8 since each bank has its own trims.

Autocal will never work well on an engine that has a petrol pressure sensor and needs a physical petrol return or electrical pressure emulator fitting because autocal works by comparing pinj on petrol to pinj on gas, while the petrol ECU on a vehicle with a petrol pressure sensor will compensate pinj for petrol pressure and petrol pressure won't be exactly the same between actually running on petrol and emulated petrol pressure (even if emulation is a physical return plumbed in). Autocal also tends not to work well on engines that idle with pinj much below 3 and tends to give a base map that is too rich under boost conditions. With the exception of a couple of systems using specific kit you would be very lucky if autocal by itself gave decent results across the load range on a 4.2SC rangerover.

An engine that uses wide band sensors may have no problem with mixture with fuel trims at 15% (it may be unusual and may point to an under-reading maf or something but still shouldn't be a problem). But like I said in a post above, widely different trims on different banks could point to a problem. And trims on these engines don't work in the normal way during initial learning of trims after a hard fault or trims reset, they have at least got to pass all I/M readiness checks before fuel trims start to work like you would usually expect fuel trims to work. Plenty other vehicles have the same aspects regards fuel trims, e.g. some Beemers and Jags will at first allow a fuel trim at idle up to around 30%, which becomes the baseline trim for the rest of the map after I/M readiness, in a sense a bit like simple autocal on an LPG ECU.

If an engine uses a map sensor a vac leak won't usually cause high trims because the map sensor means if the manifold pressure is increased by anything including a leak the engine will get the correct fuelling for the manifold pressure, a leak just serves to increase manifold pressure to higher than it would normally be for the given throttle position but the map sensor sees real map pressure regardless of any leak, there is nothing to measure airflow on the throttle and there is nothing to measure airflow through the leak but the map sensor doesn't care where the air pressure in the manifold came from and reads correctly regardless... but although trims may be fine other problems may present related to TPS and too high idle speed etc. If the engine uses a maf sensor a leak is false (unmetered) air entering the engine, the leak will flow most when the difference in pressure between manifold and atmosphere is greatest which is at idle (vac leak) and under high boost (boost leak). A small vac leak may show higher trims at idle and very low loads but decreasingly affects trims with increasing manifold pressure and rpm. A boost leak will make trims more negative but because the engine is likely to be under a lot of load and shifting a lot of air anyway a boost leak has to be bigger than a vac leak to cause the same extent of movement of fuel trims (though in opposite direction) compared to a vac leak. Or to put the difference between a vac and boost leak another way a 5mm hole is comparatively a large fraction of a nearly closed throttle but is a comparatively small fraction of a nearly open throttle. If both a map sensor and maf sensor are fitted if there is a manifold leak the map sensor will still read true but the maf will read less airflow for any given map under vac conditions and more airflow for any given map under boost conditions.

If trims are high but quite consistent across the engine load range it usually points to a maf issue but as said may not be a problem. If I read petrol trims of 15% and the driver reported no mil issues I might aim for a 7% trim on LPG, half way between 0 and 15%, not too far from 15% but erring on the side of caution and not causing a mil issue because it avoids trims drifting higher than 15%.

If I see trims of 15% on one cylinder bank and close to 0% on another cylinder bank on petrol on an engine that has only 1 maf or map sensor I expect some sort of issue with the engine itself or with a lambda sensor or one side of the exhaust. If trims are close to equal on petrol but as far as that apart on LPG it points to a problem with the LPG system. If trims change on both banks even though pinj remains the same switching between petrol and LPG at a given engine rpm/load you should calibrate on trims rather than using the pinj (dot to dot) method because if fuel trims remain the same when you switch between petrol and LPG it will be the fuel pressure reading that causes the difference in pinj.

At some time or other I wouldn't mind betting I have fixed fuelling issues on most of the 4.2SC rr's ever converted to LPG.

[mgrover]

So am currently allowing it to relearn the adaptations. I'll see where it settles, when its passed the IM readiness I'll check the Adaptive Trims. Then then testing again. It feels like an air leak but I can't find the sucker.

"If trims are high but quite consistent across the engine load range it usually points to a maf issue but as said may not be a problem. If I read petrol trims of 15% and the driver reported no mil issues I might aim for a 7% trim on LPG, half way between 0 and 15%, not too far from 15% but erring on the side of caution and not causing a mil issue because it avoids trims drifting higher than 15%."

This is what I want tbh. I've tried several times to bring down the trims to a respectable standard ie 7%. But it just seems to bring the trims down to an acceptable level, then after a bit of driving goes back to leaning out.

But I am a complete n00b and most my efforts are fairly hamfisted. Ie changing the flow of injectors to compensate for across the timing ranges.

At this point I am tempted to buy another RR, same engine and spec, check its fuel trims are on point with the tools before I buy it. Then swap whatever I want over and sell this one

[hitman]

A bad point for pulling in air on these is the pcv hose connection on rhs of engine if you are sitting in the car. I have had to resort to wrapping in self amalgamating tape to get it to seal, @!## design....

[mgrover]

A bad point for pulling in air on these is the pcv hose connection on rhs of engine if you are sitting in the car. I have had to resort to wrapping in self amalgamating tape to get it to seal, @!## design....

did it show up in a smoke test?

also how does the pcv work? since for me when vacuum pressure is high then it'll suck the diaphragm and allow it to the inlet manifold and when vacuum is low it won't allow it

i did replace the pcv about a year ago so cant see it being issue.

[hitman]

There's a massive pull on that system when your sitting at idle and mine was absolutely pissing through the hose connection at the pcv, it didn't need a smoke test to find it. Just a poor design where the hoses clip together.

[LPGC]

There is some easily broken brittle plastic air ducting on these engines and plenty scope for seals on vent ports on cam covers etc to leak (drivers side front cam cover vent - the seal between cam cover and fitting as well as the seal between fitting and pipe). If the evap purge system is at fault it can also cause false air / fuel vapour (mostly false air in these cold conditions) to enter the manifold at low manifold pressure, the ECU is so sensitive to changes in airflow and mixture that it can detect and throw an error code if increase in airflow is not what it expects when it pulses the epv.

I still haven't read if fuel trims are around equal on both cylinder banks or are only at around +15% on one cylinder bank...(?).

Another thing that can make even a good maf under-read airflow is an air filter that doses the maf with yellow dust. It's a good idea to only use decent quality air filters (on any vehicle that uses a maf but especially on a vehicle with a fussy ECU / fuel system like this model).

[mgrover]

Am waiting to do enough miles to learn the fuel trims before posting the new picture of them. Normally bank 1 is 15% and bank two is like 10%

[mgrover]

so i changed the thermostat today, checked the fuel trims and now at idle its rocking like 20% instantly, even after clearing the adaptations. i had to remove the plastic intake which meant unplugging the maf, i wonder if my maf is either knackered OR the cable to it is dodgy, but who knows.

[Brian_H]

Is it possible to clean the maf? Debris from filter etc might account for it underreading, particually if its a used item off another vehicle?

[mgrover]

i tried to clean it with some contact cleaner. in the meantime i've ordered another maf, a denso one, but not RR branded so might regret that

[hitman]

Maf readings to compare

All in neutral not fully warmed up: All readings approx.

Idle 6

2000 rpm 19

3000 rpm 27

4000 rpm 38

5000 rpm 54

[LPGC]

Is that 20% on both banks?

Again if there's been an error code and/or IM readiness hasn't passed fuel trim readings particularly around idle may not work like you'd expect.

Until the engine has warmed up the engine oil will be thick, bearings will be tight and lambda sensors will be a bit less sensitive than when hot, they use more air and fuel when cool than when warmed up.

[mgrover]

I'll check the maf thanks hitman

I upped the pressure of lpg to 1.6. The map seems better their.

20 on both banks.

I smoke tested so much that the smoke started to come out the smoke machines full port. I assume it's air tight then.

These tests are all with a warm engine.

Currently am trying to compensate for the leanness in the lpg map. It works, great. I go for test drive. Boom its lean again, no idea why, I mean if it was an air leak I would of already compensated with the extra lpg.

So am at a loss

edit: i do have a theory, if I unplug the maf the car obviously freaks out but runs roughly, it evens out sooner or later. does that mean its taking the readings from the o2 sensors for that? i need to dig up the workshop manual to check.

edit: also just incase the lpg ecu is doing something it shouldn't, would simply unplugging it effectively disable it?

[mgrover]

bit more confusion, can a maf die from getting too hot somehow? when i first fire her up the sensors are showing perfect 1.0 o2 ratio. also about 5.8 g/s from maf. then when it gets past a certain heat it seems to go down to like 4.8 g/s. but the revs haven't changed at all for this to add to the confusion

am tempted to just buy a new genuine maf, but dont fancy spaffing money against the wall for no good reason.

[Brian_H]

edit: i do have a theory, if I unplug the maf the car obviously freaks out but runs roughly, it evens out sooner or later. does that mean its taking the readings from the o2 sensors for that? i need to dig up the workshop manual to check.

edit: also just incase the lpg ecu is doing something it shouldn't, would simply unplugging it effectively disable it?

On the first point, it should have a default map that gets applied in most cases, But I can't confirm on your particular vehicle.

The lpg ecu won't be doing anything to the maf (As it doesn't connect to it). You should be able to disable the lpg ecu by removing the fuse from the battery feed to the lpg ecu (if there is one, if not disconnect the lead from the battery terminal). Disconnecting the ecu from the loom will result in failing to start or run as you will have broken the connection to the petrol injectors (the lpg ecu breaks the connection between the injector and original car loom, which gets looped through the ecu to control if the car is on gas or petrol. Without the ecu in place you have a gap in the car loom effectively so no petrol injectors can fire).