3.5 EFi wont start.

First thing to check is the fuel pump. On the flapper EFIs the fuel pump gets its power from a relay next to the ECU. The relay is operated by a number of possible inputs - one of these is a feed from the starter circuit so the fuel pump runs while you are cranking it over. My guess would be that you are getting power to the pump while cranking but the pump is stopping when you stop cranking so the engine will run for a few seconds and then die through lack of fuel pressure. Once you stop cranking and the engine is running the air flow meter drives the fuel pump relay (blue/purple wire) for as long as there is air flowing through the meter. The idea is that if the engine dies for some reason the air flow will stop so the fuel pump will be switched off.

The first thing I would do is provide a temporary ignition switched feed to the fuel pump and see if that resolves the problem. The easy way to do this is to remove the fuel pump relay and put a jumper between the ignition feed and the fuel pump wire. The ignition feed to the bottom of the relay is the white wire (probably 2 white wires) and the fuel pump feed is the purple and white wire. With the two connected together you should hear the pump running as soon as you turn the ignition on.

If that works then replace the fuel pump relay and bypass the air flow meter switch by fitting a jumper wire between the brown/orange wire and the blue/purple wire at the air flow meter connector. If that cures the problem then the air flow meter is suspect.

The relay next to your ignition coil is the overrun fuel shut off relay which disrupts the ignition pulses to the ECU when the vacuum switch on the plenum chamber detects an overrun situation. When this operates the ECU stops getting ignition pulses so it stops opening the injectors. When you disconnect it the ECU no longer gets it's ignition pulses so it doesn't inject any fuel.

cheers

Dave W.

In addition to the excellent info above, the blue/white wire should be connected to the coil -ve, but through a resistor. (It should also connect to the relay via the bottom of the base, not jammed in between the relay and it's base). The purpose of the resistor is to prevent a voltage spike from damaging the ECU. Can't remember the resistors value off the top of my head, but it's quoted in the RRC factory manual. Also, with this lead disconnected it's still possible to inject a small quantity of fuel, by sharply pressing the throttle pedal. The 4CU injection system triggers acceleration fuelling by sensing a change of throttle position voltage above a certain value within a certain time, and alters the fuelling by simultaneously firing all

8 injectors either once or twice, regardless of rpm or current firing sequence. Under normal circumstances it "bank-fires", but the extra fuelling is in addition to the normal bank-firing.

-- Badger. B.H.Engineering, Rover V8 engine specialists.

now live but still under construction,

Thank you very much for the excellent explanations fron both of you, i will follow your advice and investigate further, will post results here for future reference for others in a similar position.

I didn't realise that the blue/white wire was supposed to be there, ill remove the block it is pushed into, clean it up and see if there is a better way of fitting it, ill also have a friend fit a resistor.

Thanks for all the help, it really is very much appreciated.

On or around Mon, 6 Sep 2004 11:30:02 +0000 (UTC), Dave White enlightened us thusly:

does the hotwire have the same device, and if so, where? I'm looking for to disconnect the gas on over-run, managed it on the Ford, with it's K-jet. Not seen how it works on the hotwire on the disco. On the ford, it has a mechanical "throttle-shut" switch and a speed-sensitive relay.

If I can get over-run shut-off to work on the gas system on the V8, I'll be highly delighted. gives much better engine braking, apart from the relatively minor fuel saving.

No, there's nothing like that on the hotwire but then the hotwire is a much cleverer piece of kit so probably uses the throttle pot to judge when to do it. During the mapping I've been doing on mine I've not noticed that characteristic but then I haven't been looking for it and it's not easy to reproduce when you're balancing an oscilloscope on the wing, holding it steady with your left hand while you operate the throttle with your right :-)

For gas you could fit the vacuum valve off a flapper system. The valve screws into the plenum chamber and, if you're lucky, it'll be the same thread as the spare port in the same place on the hotwire system ! The valve operates a simple switch when overrun occurs and, on the flapper, this in turn activates the relay and cuts off the ignition signal to the ECU. You could maybe rig it so that (via a relay) it kills the power to the solenoid after the vapouriser unit, thus shutting off the gas supply to the engine. There's not really a better way of stopping the gas unless you have a multi-point LPG injection system.

There might be a couple of problems with doing it with the LPG solenoid though -

1) The delay when you step on the gas again, as it will take a while for the gas to come through, and there may be enough pressure in the pipe to the regulator to fuel the engine anyway unless the overrun condition continues for a while. 2) Weakening the mixture off and then re-enabling the supply again might well cause a blowback which would risk damaging the air flow meter even if you have an arrestor fitted.

For petrol you could mimic the flapper technique and put the relay between the coil -ve and the ECU pick up wire and it would work just like it does in the flapper system. Whether you'd notice any difference or not...

cheers

Dave W.

Thank's very much both contributors, i took your first bit of advice regarding the relay , i put a multimeter across the first lead, and there was power when cranking, across the second lead there was no power when cranking, taking out the relay, and putting a wire across as you mentioned, the engine fired and ran as sweet as a nut, I am extremely grateful.

Regarding the resistor, is it possible to purchase a wire with the resistor inline, the blue/white wire had come out of the connection underneath, i should be able to solder it back into place.

Once again thank you very much.

On or around Tue, 7 Sep 2004 16:28:08 +0000 (UTC), Dave White enlightened us thusly:

this is exactly how I did it on the ford. The system however only interrupts the solenoid supply to the vapouriser, leaving the rest of the gas system live, so the gas supply comes back almost immediately. The ford system was cunning in that it has the speed-sensitive relay, which only allows the fuel cut-off at more than 1700 rpm or somesuch figure, however I assume that the vacuum switch has similar effect on the RR system. Will investigate.

neither a problem using the shut-off on the vapouriser itself.

never noticed it. and backfires have happened and not apparently hurt the hotwire AFM, which I think doesn't have a flap to break like the flapper one.

Looks like i spoke too soon, replacing the relay had no effect, i tried again with a wire across the relay holder told to earlier, and the engine runs beautiful.

Can anyone suggest what to look for now ?.

On or around 8 Sep 2004 01:55:06 -0700, snipped-for-privacy@englishrosetrading.com (RR) enlightened us thusly:

the relay signal?

should be on terminal 85 or 86 if it's a normal relay, the other one should be earthed. Normal switching relays have terminals as follows:

30/51 supply 85 trigger 86 earth 87 NO load - connects to supply when triggered 87a NC load - disconnects from supply when triggered

in your case, I'd expect the load on 87a.

these numbers, I assume are part of someone's Grand Scheme to number all the terminal connections on vehicle components. Ignition coils tend to be 1 and

15, ISTR. Anyone know whose grand scheme, and/or have a copy of the complete list?

Thanks Dave, as i posted earlier, i have now tried this, and the engine ran fine on petrol, so i bought a new relay this morning fitted it, and the problem still exists.

This was my next step, done this, but the fix you suggested has no effect at all, with the new relay fitted the engine does as before, startes then peters out, do you have any other ideas.

BTW the wire going into the pump relay is not blue/purple, it is green / white, is there a way to trace this wire back anywhere ?. I can post new iamges of the pump relay if you wish.

Thanks for your help.

It is indeed on 85, when cranking the engine with a multimeter across

85 & 86, the reading goes up to 9v, the engine fires almost immediatley (no time to climb to 12v ?), once i release the key, there is no voltage across the two poles.

I am going to check all fuses, but do you have any other ideas on what is possibly causing this, any idea where the wire on 85 goes to so i can check for breaks etc.

As a temporary solution would any damage be caused if i rigged up a live feed from a switch to post 85, could i drive the car safely like this?.

I only need a solution to get me to an LPG station should i run out of LPG.

OK, this at least tells us where the problem lies... the fuel pump relay is only activating when you are cranking.

The blue/purple wire is at the air flow meter, not the pump relay.

OK, it seems that the fuel pump relay is not getting a signal from the air flow meter to keep it activated so we either have a broken air flow meter, a wiring problem or a problem in the diode pack. The signal to the fuel pump relay goes from the main relay to the air flow meter on the brown and orange wire. When the flap inside the air flow meter moves ( this is why it's called a flapper EFi BTW) it operates a switch which connects the brown/orange wire to the blue/purple wire. This blue/purple wire then takes the feed to the diode pack which then sends it out on the green/white wire to the fuel pump relay. The flapper AFM is a fairly simple device, it uses a spring loaded aluminium flap to close off the air intake. As the engine sucks in more air the flap is opened by the air pressure/vacuum and moves a variable resistor so the ECU can know how much air is going into the engine. The switch for the fuel pump relay is there as a safety device so that, in the event of an accident for example, as soon as the engine stops the flap shuts and the fuel pump relay turns the fuel off. You should consider this before jury rigging a bypass... Without this switch the fuel pump will run as long as there is power to the ECU and, in the event of an accident will continue to feed fuel at high pressure into the fuel lines.

The diode pack is located in a relay package next to the fuel pump relay. It looks like a relay but is, in fact, a relay box with 3 diodes inside it. IIRC it normally has a red plastic housing. You need to see if there is a +ve signal on the blue/purple wire at the base of this pack when the engine is running on petrol (loop out the fuel pump relay to test it as before). You would normally expect a signal there while cranking but sometimes the engine isn't cranking fast enough to move the flap in the air flow meter sufficiently.

If you are getting a +ve reading there but not on the output (the white/ green wire) then you have a fault in the diode pack.

If you are not getting a signal there then you need to test at the air flow meter. Pull back the rubber cover on the AFM connector and take all measurements with teh plug connected. With the ignition on but engine not running, check for +ve at the AFM on the brown/orange wire. If you have a +ve there then you can push the flap (take the air filter off and push the metal flap with your finger) and you should get a signal at the blue/purple wire. If you don't get a signal then the switch in the air flow meter is knackered. To confirm this use a jumper wire between the brown/orange and the blue/purple and the fuel pump should operate whenever the ignition is on. If you do get a signal on the blue/purple wire at the AFM but not at the diode pack then you have a break in the wiring somewhere between the two or a bad contact in the AFM plug.

Given that the vehicle is running on LPG, if you want to save some time you might want to do the AFM checks first. LPG backfires can and do damage air flow meters, especially the flapper type.

cheers

Dave W.

It might be worth fitting the valve and just using it to light up a warning light on the dash. That way you can try it in a number of situations and see when it operates, prior to connecting it to the LPG solenoid.

My Disco has a LONG pipe between the vapouriser and the regulator which I'd have thought would hold enough gas to run it for a while. OTOH, given that the solenoid was shut preventing any fresh fuel in the gas probably can't go into the engine without forming a vacuum. You'll have to let me know how you get on :-)

The hotwire AFM is a lot less likely to suffer damage from a backfire although I've heard it can be damaged, never seen it myself though so could be urban legend :-) The only bit on my Disco that suffers is the plastic airbox - the first backfire I got visibly lifted the bonnet as the airbox lid hit it with such force !

cheers

Dave W.

Definitely not urban legend Dave, seen a knackered one on a vectra (sorry for swearing) after a couple of bangs. The air temp sensor part of it didn't like the shock waves of the backfiring and ended up embedded in the airfilter, some 18" away! I've seen the protective gauze blown out of many a hotwire landy V8 after a bang or two.

-- Badger. B.H.Engineering, Rover V8 engine specialists.

now live but still under construction,

On or around Thu, 9 Sep 2004 17:08:10 +0100, "Badger" enlightened us thusly:

p'raps I should check mine, as it sometimes bangs on changeover for no reason I can determine. 's never banged enough to blow the airbox apart though, but mind it has the cylindrical type airbox...

On or around 9 Sep 2004 01:09:49 -0700, snipped-for-privacy@englishrosetrading.com (RR) enlightened us thusly:

aha. well that's yer culprit then. there undoubtedly should be.

reading back a bit, the relay should be supplied by a feed from the AFM, which may not be happening.

don't see why not. You'd lose the saftey feature that shuts off the fuel pump if the engine dies, and the over-run fuel shut off as well, if I read this rightly.

does your LPG system cut the pump supply?

mind, on my system it cuts the injector pulses and leaves the petrol pump running - this makes for more-or-less instant switch back to petrol, which is handy.

Thanks.

That is i thought, thanks again.

I am not sure, i suspect so, i will investigate this later today, what i can tell you is that when you switch from petrol to gas you hear a relay switching, so im going to see if i can find this relay.

I dont think it does that, ill check it out today though.

On or around 10 Sep 2004 02:57:07 -0700, snipped-for-privacy@englishrosetrading.com (RR) enlightened us thusly:

It's quite obvious normally, the loom that does it on mine has 2 little black boxes each with a cable trunking coming from them, at the other end are 4 pairs of connectors, on e of each pair plugs into the injector and the other is the opposite gender and attaches to the normal injector supply. The black boxes produce a timed shutdown of the injectors which in theory reduces the lean-fuel period during changeover to gas - the gas system goes live when the changover unit switches it and the black boxes cut the injectors about 1 second afterwards. In practice, it doesn't seem to work - if you set the changeover to low revs, the engine dies before the gas gets through, while if (as I have it now) you set it for about 2000 it's still prone to the occasional backfire. Actually, I've found the most reliable way is to get it to change on the 2nd-3rd upchange - rev to about 2300 in

2nd, change to 3rd, let the clutch in without applying throttle and then apply throttle again.

once it's been running on any given day, I force it to start on gas anyway.

Mine doesnt have any of that, or any of those problems either, it will also start from cold on gas without any trouble.

I suppose that my next move should be to see where the loom from the changeover switch goes to, this i think could be where the problems lie, though i am loathe to start messing around in this area as replacement parts could be hard to find.

Will keep you posted.