XJ6 S3 4.2L 1982 Stalling problem

The XJ6 S3 I have is on dual fuel. Recently, I’ve been chasing an intermittent backfiring which was occurring on LPG when driving up long hills. The problem did not show up on petrol. Today, I was still trying to figure out the problem, when driving on petrol going up a hill the power surged on and off and the car then finally stopped. The car cranked but would not start on petrol or LPG.

I had a look and found there was no power at the coil. The problem was a broken wire going to the coil. It was a brown coloured wire in a small three wire loom. One wire goes to a temp sender and the other two to the coil + and -. The insulation on all three were dry, brittle and cracked as they run close engine and get cooked. I cut off all the suspect wire sections and added a new piece of wire to all three original wires.

The car now starts on petrol and LPG. Runs fine on LPG but stalls on petrol unless I keep the revs up.

Any ideas what could be causing this petrol stalling? Maybe I’ve disturbed something else whilst making these wiring changes? Vacuum leak, airflow connection, fuel injection signal wiring???


Did the wire go to the rear of the car with a bullet connector in between, that’s the signal for the ECU! If it’s disconnected it will stop the engine on petrol despite sparks and all still working for gas. The ECU injects on every second pulse from the coil. Bad connection? Otherwise no idea.

It seems to me its a fuel tank issue, since you mention uphill.
Clean the tank and other parts along the fuel items.

Hi David,
I didn’t track the wire all the way but it goes into another wire that runs around the front of the motor. I’ll check where it goes today. If i keep it a bit above idle speed it keeps running but maybe its a bad connction. There is splattering when i take the foot off the accelerator, it just stops like the key has been turned off.

I did not see any connector on the cable, is it a single connector or does it go into a multiple wire plug/connector?
Where abouts is this connector located?

Is there a way to check that inject every second coil pulse? Guess you’d need an oscilloscope? Most of my tools/cro are in storage while i look for a new rental house. Only got a multimeter at the moment.

If it goes around the front it has to be another wire.

If you turn the ignition on and disconnect and tap the bullet connector wire together you will hear the injectors pulse. It’s a very sensitive process so don’t worry it it seems to be every, or every third time and so on. Don’t leave the ignition on too long of fire them more than a few times (so it doesn’t flood).

It’s a neat trick to check whether the ECU, the signalling and the injectors all still work. It’s also a neat trick to add some fuel if it’s having a hard time starting and it’s not flooded (yet).

It can’t tell you if the amount that’s injected is correct or if some wire is not hooked up right.
In your situation I would look at air leaks first. Especially if it starts on petrol but doesn’t leak. Also go through your fuel pressure eventually, and the temperature sender with the square Bosch plug…
Good luck!

There are two different routes for each fuel, Andrew - and it looks as if the petrol is not being adequately delivered, as Walter indeed suggests. Have you tried changing to the other (petrol) tank when the engine falter…?

I’m not well versed in Jaguar dual fuel set-up - but I assume that ECU is involved in metering both fuels according to pedal movement/air ingress. Basically; rpms/power are dependent on the amount of air going into the engine - which is controlled by the throttle, and measured by the AFM. And the amount of fuel adjusted accordingly by the ECU - unless LPG delivery is controlled by a separate agency…? Ie, delivered directly to the manifold rather than through the injectors?

So the amount of LPG is obviously delivered to spec, but petrol is not. Which may be restriction in petrol lines (including clogged main petrol filter), failing fuel pump or failing fuel pressure regulator…?

On the ignition side; the LPG may (arguably) be easier to ignite - and a weak spark may fail to ignite petrol? In either case; there are two white wires on coil ‘+’ - one from ign key, the other to the ign amplifier. And three wires on coil ‘-’; one white/black to ign amp, one white/black to the ECU (via the bullet connector mentioned by David).

And a white/slate blue going to the tacho - which at any time should slavishly show actual engine rpms. loss of ignition; the tach drops to ‘0’ and erratic tacho readings may imply an amplifier fault. The ignition system (coil, amplifier, distributor and plugs) is entirely self contained - independent on other components…

The ECU opens the injector once every engine revolution - on every third(!) ignition pulse. The best way to test this is to just listen to the injector (all simultaneously) ‘clacking’ during cranking. They are noisy, but you can use a ‘stethoscope’, a long scewdriver between your ear and an injector to amplify.

Not clacking; a (injector) test lamp (noid) can be connected between the two wires on a disconnected injector plug. With ign ‘on’, there is no light - while cranking; the lamp should flicker. Or with a digital voltmeter; showing '0’V with ign ‘on’ and ‘something’ while cranking. An analogue voltmeter will show '0’V and somewhere between 0 and 12V while cranking.

The cause of this is that one injector wire has constant power (ign ‘on’) - and the ECU briefly grounds the other wire to open the injector. Don’t bother about counting pulses - the ECU either operates the injectors - or it does not…:slight_smile:

xj6 85 Sov Europe (UK/NZ)

Thanks Frank. I’ll read through that and digest it.
The latest test i did was to do with the AFM. Before it stalled i manually moved the contact wiper up about 3 or 4mm and it idled without stalling. What does that mean?
I’m wondering if the contacts in the afm are opening too early and disconnecting the fuel pump circuit?
Or is this pointing to a vacuum leak?

Pointing to a vacuum leak or other lean condition. You can run it with the AFM cap off and look at the contacts but they should be okay. I wonder though how it runs on lpg with the leak?

The contact wiper is slaved to the AFM flap, Andrew - and flap deflection reflects the amount of air to the engine. So adjusting the wiper arm relative to the flap effectively alters the mixture - it ‘tricks’ the ECU into assuming that more (or less) air passing the AFM. The arm/flap is set by factory to give the approximate correct mixture - and fine tuning is done by the AFM bypass.

Two things may occur over time; the contact between the resistor block and the wiper may deteriorate/fail and mislead the ECU. Or the spring gets weaker, requiring resetting the arm to bring the mixture adjustment within the range of the AFM bypass. Your adjustment either restored/improved contact (the fiddle factor) - or simply fattened the mixture. Which may indeed counter an air leak; a common test is, with the engine idling, gently pushing on the AFM flap. If even the gentlest push cause engine to falter, mixture is correct - if rpms increases; it is a sign of an air leak…

In short; adjusting the wiper, fattening the mixture, counters lean mixture - whether caused by air leaks, restricted fuel delivery, too low fuel pressure, clogging injectors, faulty CTS data, misreading AFM…take your pick…

As to the fuel pump control; there is a reed switch in the AFM that is ‘break’ with the flap closed (engine not turning, no air opening the flap) - as the flap opens, the switch activates the fuel pump. To with, the pump should run with ign ‘on’ and the flap ajar. This is a separate (safety) function, not related to the flap/wiper adjustment…

As I (mis?)understand it; the fuel injection is disabled with LPG selected, and vice versa. The ignition system works the same for both fuel alternatives…

xj6 85 Sov Europe (UK/NZ)

Must be like that, it probably breaks/makes signal to the ECU or lpg System. The lpg working well makes me wonder if it isn’t an air Problem but rather:

ECU thinks engine is hotter that it is

Fuel pressure too low

One, two or even more injectors not receiving signal or blocked (you can check with the bullet wire disconnected and only one injector plugged in, if it clicks, then the next, etc.)

Any tips for checking for leaks, other than visual inspection?

On lpg it starts ok but i do seem to have to prime it a bit more than normal. Once started it idles well and drives well except when driving up a long hill where it eventually loses a little speed/power before backfiring. Maybe the 2 problems are caused by a leak, making the lpg lean out and backfire as well as causing the stalling on petrol?

The lpg system is a simple impco venturi setup. So, air filter then AFM then lpg mixer. So the lpg mixture is independent of afm and ecu.

What does the allen key screw, that is located in afm housing corner do? The one that is located down a 10mm hole on the housing.

That screw regulates the air that bypasses the flap, in other words, turning it in enriches the idle, out adds unmeasured air, like a controlled vacuum leak. This would cure the symptoms to a degree but not the problem.
The common way to check for vacuum leaks is spraying carb cleaner on all parts and listening for changes in idle.

I would check and temporarily block all vacuum lines too, that’s four ports on the intake manifold (one on top for the pressure regulator, one somewhere behind the runner for the rearmost cylinder and one drain in the rear, plus the brakes). The two aluminium devices and the crankcase breather also tend to blow off (especially the one you can’t see, behind the AAV), and the large duct to the AFM of course.
Lean backfires will blow off hoses sometimes.

I’m guessing you can’t check the AAV without partially removing it?

You can pull the hose off and shine a light through. You don’t seem to have a faulty AAV, just an air problem. Brake cleaner probably your simplest aid.

Ok, get some brake cleaner today and see what happens. One other possible clue to what’s going on, starting car from cold it idles fine for about 20 sec or so then it slowly drops in revs and stalls.

The 20 seconds of fine idle by revs suddenly dropping and stalling could be a problem with the coolant temperature sensor wire not being properly connected. I had the same problem last week on a 1987 series 3 car. I remembered that about five years ago I had the same problem traced to a faulty coolant sensor which sends incorrect signal to the ECU. I got out of the car, grabbed a pair of pliers, and squeezed hard the crimp connectors where the wires from the new ‘pigtail’ connectors met with the wiring from the car’s wiring loom. I got back in the car, turned the key, and all was well once again…

Is that the square connector with 2 leads going to it. Is it a switch or is resistive sender?

There is another single wire sender nearby, i guess that is for the temp gauge?

It is a 5 mm hex (Allen key) at the bottom, Andrew…

…to be turned as David explains clockwise to reduce bypass, fatter mixture - CCV to lean it. Reducing bypass increases the flap deflection - more fuel…

Years ago I drove friend’s van, dual fuel like yours, but unfortunately I did not pay much attention to the technical set-up - only that the transition between fuel types were absolutely seamless. And that he said to start it on petrol - otherwise I was none the wiser…

Irrespective of fuel selection; air is still going through the AFM, but I do not know if that, or something else, participate in controlling LPG delivery - and indeed if the ECU is relevant in this respect. I need a lesson on LPG feed controls for better advise…:slight_smile:

The AAV is ‘active’ with both fuels. As David says; remove the top air hose to it and look at the slide - it should be half open when cold, and closed when hot. The slide is spring loaded (bimetallic spring) - and the slide can be exercised with a small screw driver if sticking. Or indeed sprayed with a cleaner as advised.

Air through it is measured by the AFM - but as said; AFM’s influence on LPG fueling is uncertain. However, there is also an air hose underneath the AAV - if detached/leaking it is an air leak…

xj6 85 Sov Europe 9UK/NZ)

That the engine dies (on petrol?) after some 20 seconds when cold, may be that it starts with the extra petrol injected by the CSI - then dies as that petrol is used up, with insufficient feed. Or an air leak requires more fuel than is provided - applicable for both fuels.