I refer to my earlier post below, re. Danny's LPG.
There is now one main issue that is causing us a problem and I wondered if anyone can shed any light.
The issue is this - the LPG ECU is not interpreting the correct RPM signal. The software parameters are definately correct. As reported the wiring was shoddy so I re-made the connections thinking this would solve it.
Unfortunately this didn't fix it. At idle, the LPG ECU thinks it's about 2000rpm when in reality it's just over 500.
It's getting the signal from the bright green wire and it's definately the right one. I've also checked earths and ECU 12v feed is ok.
Any ideas on this one? It's now the only cause of any evils, we've pretty much fixed the rest, apart from a final tidy up!
A couple of times it behaved normally and I suspected very stronly it was a bad connection, but I've double and triple checked and the wiring for this seems consistent :-/
Any ideas gratefully received..
As you all know by now Danny has had nothing but trouble with his "Pro" Gas conversion, and has never been able to so much as use his LPG conversion since it was installed.
After months of the installers not getting anywhere I had a look at it for him today.
The findings as to why it wasn't running on gas are as follows:
1) Coolant feed to the reducer was non existant, only just a trickle. What they had done, is take the tiny coolant pipe from the Nearside of the throttle body, bodge a rusty old union on there to connect it to a larger pipe, and run it to the reducer. There was then a pipe from the reducer, going into another cobbled up home-made adaptor, and then joined back onto the top of the header tank. Completely inadeqete and the reason why it wasn't happy under load and at high revs. To overcome this, I've ripped out all of the bodge pipes and re-connected the throttle body coolant feeds. Popped the scuttle off and have taken a feed directly from the coolant bridge for the evaporator, which flows through it, and returns back to the HBV. This way the evaporator will get consistent coolant regardless of heater controls. The result was an instant difference, a red hot evaporator! It's also a million times more tidy than the installers job, with coolant hose wrapped around the air inlet trunking.
2) The LPG filter was fitted the wrong way around. You may not think this would make a difference but, once put right, the idle was a million times smoother and is no longer hunting around and lurching like it was before.
3) Appalling wiring throughout.
- The lambda wires had just been cut, and not terminated, so were free to short.
- The wiring for the gas temp sensor was dodgey, (reporting -70 degrees at one point!)
- The wiring to RPM signal was appalling
- Wiring to the battry and 0v - again really, really shoddy, looked like schoolboy work. Resulting in an intermittent supply to the LPG ECU.
- The injector splice cables are again a huge mess and all over the place
- The gas injectors themselves are just hanging loose
4) The wrong type of evaporator and injectors were configured on the software (Thankfully my Stag lead worked).
Having done some quick repairs on the above, apart from a minor switching issue I'm looking at, the car is now running really nicely on LPG 
Other faults found, were:
5) Manifold nozzles have been drilled in situ - you can see the swarf everywhere!
6) The evaporator is very insecure and has been mounted via a bracket connected to the BRAKE MASTER CYLINDER!
7) Gas switch inside the car very poorly fitted.
Rusty old gas unions used to join coolant pipes
9) Bare wires all over the shop.
10) components routed near EGR valve, won't last long there!
11) Pollen filter had been removed to make way for the gas ECU
12) Gas ECU insecure and hanging off it's cables..
13) Only 6mm feed pipe used for vaporisor, it really needs 8mm on an engine with this capacity.
In short, it looks like an ameteur DIY job at best, it's nothing like professional standards.
In a few hours it's gone from being a shambles to something that works, so that's good progress really.
Unfortunately couldn't finish it all off due to the sheer volume of work, but we are confident that another days tinkering and it'll be 100%.
It needs a day to properly solder and heatwrap the connections, do some continuety testing on suspect parts of the LPG loom, connect up the lambda's to see what they're doing and get the mapping spot on, and generally tidy up the install - it should be good as new, then 
Must say, I much prefer the Stag 300 kit I've used compared to the OMVL21 fitted in this application. The Software is very very limited compared to the stag one, and components are not as good, IMO.
Hopefully all's well that ends well, pleased you're happy Danny
Author
Topic: Danny's LPG - final issues (Read 2062 times)