pcmhacking.net

I drive the car 99.9% of the time on gas and I think the injectors might be getting a bit dirty on the tips because it doesn't idle that well on petrol. It idles just fine on gas though. Then I thought maybe the one-wire oxygen sensor might be worn out (unknown age) but in these engines is it actually used at idle? Mine doesn't use it on gas at all so the fact that it idles okay on gas doesn't mean anything as far as that is concerned.

P.S. Will stale 6 month old petrol make it idle bad? Runs okay on it otherwise.

stale petrol can be a problem. at this point you should get your laptop hooked up and log some data from the car and see if that sheds light on anything. from your other thread, i'd think checking voltage might be a good start. did you say you had a cam, but are running a stock tune? if the gas is carby gas then efi fuelling might be out for the cam (as the manifold pressure to airflow relationship will be changed), and the gas carbie fueling might be better as its not related to a map sensor tune.

Not sure how you have it setup but the ECM might be trying to adjust the fuel (which it can't obviously) when on gas and as such has learned an incorrect correction value. Could try resetting the ECM

antus, it's a bog standard cam, in fact the entire motor is such. The ecu only had a different spark table; the fuel table is untouched.

charlay86 I see what you are saying. I would have thought though that after 5-10 km of driving on petrol it would start to dial itself back to where it should be. I'll reset the thing and see what happens.

Gas always causes issues with petrol, iv got mine setup on dual maps and still have the same sorts of issues, last time i tried petrol it ran at 10.0 on my wideband and stalled when i floored it. Iv never had any of my dual fuels run well on petrol. So if someone knows how to fix these issues id also like to know

Cause a fault that disables blm? Relay on the lpg circuit to disconnect o2 sensor signal from the ecm perhaps

the most dual fuel i've run is petrol and methanol ......

my only suggestion is if its right on one map (map a for example) but really different on map b then change the injector rate .... that said, if you do this you want to make sure the change effects the whole rev range by the same amount.

eg. you might have a petrol map a and methanol map b, so map a will have .044 injector rate, but then you find map b is really rich, change the injector rate at idle to get the same AFR on methanol, but then give it a few revs and see if the change stays the same at different load points. if it does its all good.

in my case above i only really use petrol to flush methanol from the fuel system so it really doesn't have to run that good, but with enough work you will be able to dial it in better. You might also have to tweak AFR tables too.

Last night I drove about 25km down the freeway at 100kph on petrol and it ran just fine but still idled a little rough when I got to the end. Like it had a VERY mild cam in it. Not a show-stopper but it should be better like it is on gas. Have not yet reset the ECU. One thing I should mention though is that when you start it up cold on petrol it idles just fine. Whether the IAC being open further than normal affects this I don't yet know. The investigation continues.

On a stock bin you will be in closed loop as you enter idle mode when coming to a stop, but after a few seconds the target AFR will drop from 14.7:1 to somewhere in the low 14s to high 13s (idle AFR table starts being used).

The BLMs are stored in a RAM table so you have different load points with different correction factors. Really need to get it on 12P and start logging it Will show straight away what is happening with the BLMs after running on gas, might just need to do what charlay86 suggested and fault out the O2 sensor on gas so it leaves the BLMs alone.

Yeah, at the moment the too-high-for-too-long and too-low-for-too-long bizzo for the oxygen sensor is disabled so that it doesn't bring up the CEL because gas is not closed loop, but that's all.