The frontmost is just a switch for the 7th injector (for real cold starts). The single wire is for the gauge, and the smaller sender with the square plug is a resistive sender that tells the ECU how much fuel to add depending on coolant temperature. Measure it:
Thanks to all the forum members for there advice. Finally found the problem.
I used brake cleaner to look for vacuum leaks. There were a few minor leaks that were fixed by tighten hose clamps. The main problem was a major leak between the AFM and the LPG mixer (Impco 225). When I sprayed around the connecting hose at the LPG bell cover input the motor stalled. I removed the hose and found that it had a small hose fitted into the main hose to reduce the diameter. So i bought a silicon turbo hose which had the right step down diameter. Fitted it and the motor still stalled with the brake cleaner test. Then I realised that the only way this leak could still be there was if the LPG mixer had a crack. Removed the LPG mixer and found the bell cover (which was plastic) had cracks in it and also had blown a hole. I’ve since found out that this is a cheap copy of the Impco mixer. Impco mixers were only made with metal bells. So I’ve temporarily duct taped the cracks and hole. Have order an aluminum bell cover. Test drive with the duct tape and problem solved.
So it all makes sense now, having the leak downstream of the AFM causing petrol stalling and leaning out the LPG mixer causing the gas to backfire.
The whole chain of air filter, AFM and lpg mixer mounting is not great, and has probably created the cracking over time. I need to get in there and redesign it. I believe that the original petrol system has the AFM mounted on rubber mounts? In this setup, it is mounted solid via a bracket to the body and the lpg mixer floats (held in by the hoses).
1 Like
Yes, the AFM used to be on 3 rubber mounts.
Thanks David. Think I’ll refit those mounts.
How do you set the air bypass screw in the AFM? I think they are set from factory but I’d like to double check it. Is it just a matter of adjusting it and listen for the sweet spot in the motor idle?
Yes. I think 2.5 turns from bottom is an accepted starting point. Mine‘s almost bottoming out. Usually, when it runs perfectly it’s a little rich for the MOT if they test that. They are set from the factory and sealed in emissions markets… thankfully, that’s not us.
Indeed, Andrew - it’s a whole lot of shaking going on. Reputedly, the same mounts were used on Harley Davidson bikes…
The AFM mixture screw is really to be set using a CO meter to somewhere between 1 - 2% CO, which is slightly rich - as the xk likes it. The adjustment has no perceptible effect on engine idling - it cannot really be set by ear. However, a bit of trial and error with test driving may reveal something - disregarding emission control tests of course…
A damned fine job done!
Frank
xj6 85 Sov Europe (UK/NZ)
**
Frank, is that 1-2% co measured at idle?
Yes, at 750 rpm, no load, hot. I would adjust it a little rich - 2-3 percent CO, wherever it runs the sweetest and a bit leaner for the MOT people.
What about those colortune plugs, are they accurate?
I think so but I never bothered. Just leave the allen key in place so if needed you can adjust it 
It’s what I did and when they checked the CO as mandatory I had to back it out quite a bit. I counted the turns so I know the next time (the idiot with the tester aimed for .5% CO as I later found out).
Zenit, when I used a colortune plug it gave me a mixture that was too lean, I tried many times to use it correctly but I ended up test driving the car making small adjustments until it ran well and the spark plugs looked ‘right’. Perhaps I was doing something wrong with the colortune set up but it really didn’t work so well for me.
Thanks Bob, sounds like a good way to do.
**
Too much so, Andrew - they really give the Lambda=1 setting, which is too lean for the xk engine’s liking. As David says; add a bit fat won’t hurt - and a bit of trial and error. But don’t automatically blame all engine misbehavings on mixture…
Frank
xj6 85 Sov Europe (UK/NZ)
**
Got the new aluminum bell cover for the LPG mixer today and fitted it. Then adjusted the wiper arm (loosen of the centre screw, which had it’s head rounded/mangled - more challenges). Then set the air bypass screw where it seemed happy. Best it’s ever idled on petrol.
Took it for a test drive, started on petrol and everything seemed ok, swapped to LPG I experienced the same back fire problem under load up a hill. Then switched to petrol and it did not want to run, felt like it might stall. Got home, just before it got dark and found the AFM looked very weird. The arm that opens and closes the electrical switch, was on the other side of the switch. Looks like it’s bent, not sure how it got and if bent how that happened. Could there be back pressure from the LPG backfire?
Attached is a picture.
Just when I thought I had it sorted, at least I’m fixing other problems along the way. Getting to the point where I’m dreaming about this stuff.
Sure looks very out of whack. I suggest you either try your best to adjust it (but it is very sensitive) or get a used one. If I was near you I would send you my spare which seems to be alright. I don’t think they should be that expensive.
Another possibility of course is that someone before you did his own modifications to make it do whatever he wants.
Backfires usually blow off the hose and air filter before wrecking the flap mechanism, but who knows… looking at the screw it was probably not right in the first place?
**
The ‘electric switch’ is a reed switch controlling the fuel relay, Andrew - it’s a safety circuit.
With the flap closed, no air flow, the switch is ‘open’ - fuel pump won’t run. As air is drawn in, the flap moves and closes the switch - and fuel pump runs. Simple test; turn ign ‘on’ and move the flap - the fuel pump should start with minimal (1/4"?) flap motion…
As an aside; the fuel pump is activated with the ign key to ‘crank’ - bypassing the reed switch, which activates the fuel pump relay with the engine running. However; this dual function uses a diode pack (red ‘relay’) - which sometimes fails, with bewildering results…
Frank
xj6 85 Sov Europe (UK/NZ)
**
You should also move the flap through the whole range; there should be even resistance to movent - no hang-ups in either direction. As David says; the ducting between the AFM and throttle body is usually the first to blow of, but a backfire may indeed be powerful enough to cause flap distortion…
The feedback to the ECU is from the ‘U’ shaped wiper - pointed at '2 o’clock in the picture. As the flap moves it wipes along the resistor connection - wear or bad connection may distort the signal to the ECU, causing fueling anomalies. Sometimes a slight displacement of the resistor connection effects a cure…
Generally, mixture is adjusted by the AFM bypass screw, but if the AFM setting is ‘out of range’; the relative position of the wiper and flap can be reset…
Frank
xj6 85 Sov Europe (UK/NZ)
**
Thanks for the offer David. I can get hold of another meter if needed.
I bent to AFM reed switch arm back into shape, replaced the mangled screw and got it running pretty good on petrol. I think the backfire caused the damage. I noticed the AFM flap does scrap ever so slightly against the body at a couple points, maybe a result of the backfire as suggested.
I’ve found someone locally who is checking out the gas setup and mixture on Monday, I think it is faulty and leaning out under load. They are also familiar with the AFM used, so should be able to double check that as well as the petrol mixture.
**
Also bear in mind that PG is less energetic than petrol, Andrew - so both fuels must be set up to to be compatible. If they are not; the changeover will be perceptible - and it should be imperceptible. No doubt your professionals will sort it…
The rubbing of the flap against the AFM body is of concern; any (tiniest) friction will interfere with the signal to the ECU - and will cause related mis-fueling and engine reaction. The cause of rubbing may indeed bent hinge pins, or whatever - and if not repairable; you need an AFM replacement…
You should also, if occasional starting difficulties are encountered, verify that the AAV is working correctly…?
As an aside; with the xk petrol injection set-up, fuel is delivered at 6 points, and use air flow for computing amount of fuel required - with rpm inputs required for triggering. With single point injection; manifold vacuum and rpms is a crude starting point. Both of course refined in various ways for temperature and dynamic transitions…
And I don’t know what components are used for LPG control…
Frank
xj6 85 Sov Europe (UK/NZ)