P38A engine stalling at idle

Hi Glen, just a wild thought:

In an earlier post you implied that the problem occurs when the engine is below operating temp : so, given that the vapour point of LPG is lower than for petrol am wondering if you are getting ice buildup at idle due to the generally high relative humidity of SE Qld.

Also, the LPG system is not a sequential fuel injection type?

Got me thinking about problem icing which can occur with the carburetor on piston engined aircraft, e.g. throttle reduced on descent in high rel humidity conditions... solution is to pull on "carby heat" which directs warm air into the induction system thereby reducing the relative humidity of the air supply.

With the Thor setup on LPG, does it stall when the engine is at full operating temp?

Given it's not gas injected, perhaps after running rich there might be some other form of metering device which needs cleaning hence the gradual deterioration in performance?

cheers

Quote:

---

Originally Posted by bee utey

Does your gas system have an O2 sensor emulator when running on gas? It passes a slow square wave to the petrol ECU to prevent it from self learning incorrect mixtures when idling/driving on gas.

---

It certainly does seem to pass this emulated value to the engine ECU, but the OBDII shows a solid 0.45v as soon as I switch to LPG. That doesnt seem square to me :eek:

Is this a correctly emulated value? It doesnt stop the ECU from throwing a code, as 2x O2 sensor faults show up as soon as I switch to LPG and then it reverts to open loop fuelling (when I run next on petrol) until I clear the faults. It seems like the LPG ECU is not clever enough to fool the engine ECU. After many minutes on LPG it also throws 8 injector faults.

FYI: I have the SGA299 self learning feedback system LPG ECU (by AEB?).

Quote:

---

Originally Posted by bee utey

Approximately 0.1V lean, 0.9V rich. 0.45V is the happy medium but it will vary anywhere from 0.2 to 0.8 while the petrol ECU/gas computer adjusts fuelling in closed loop operation.

---

I do notice the O2 sensors flipping between these values nicely on petrol, 0.1 on throttle and 0.8-0.9 on coast. On idle it alternates. So I suspect the O2 sensors are fine.

Find out where the oxygen sensor feed is cut, reconnect the sensor straight to the vehicle ECU and disconnect the emulator output. This way you rely on the gas processor functioning in closed loop and cycling the mixture. Straight 0.45V is useless as an emulation value. It should mimic petrol operation.

Find the injector cut wiring, solder a 68 ohm 1 watt resistor between the 12V in and out wires. This is quite adequate in my experience to allow a small diagnostic current to flow without opening the injectors on sequential injection, including D2's. Don't ask me where exactly these wires are, I don't like the AEB system and won't use it. You should be able to identify the injector cut wires where they cut into the main loom's positive injector feed, brown with orange stripe wires.