Citroen XM Forum

Bad fuel would not make it spit back and limit revs to this degree, by all means try it but its only been off the road a few months.

The lexia will show live data, you can watch what th ignition timing is doing, engine speed, throttle position, coolant temp, manifold pressure, air temp etc etc, in short you can see what the sensors are telling the ECU and judge if they are believable figures or not, the the lexia shows believable information being sent to the ecu then you can rule out sensor issues full stop.

You can also test components like canister purge valves and test fire injectors etc which can make diagnostics easier and quicker.

If the lexia shows all sensors working fine but you have proven the engine runs as it should by adding more fuel than the engine is getting you start to narrow the culprit down, if the ecu is getting all the right information to be able to meter the fuel correctly but there is not enough fuel going in it will be a dead cert your issue is with the pump, fuel lines, filter, damper, pressure regulator injectors.

Dont forget with stuff like map and temp sensors, if they go open circuit the ecu will see and flag up an EML, if they are telling fibs, like 10% throttle when you are WOT or similar then the ecu will not flag this as a fault it will just adjust mixture and timing according to those incorrect inputs and run like hell.

I would really like to start with a pressure guage Tee'd into the fuel supply and a timed open fuel supply discharge into a container to confirm fuel flow and pressure to the rail is correct, you would be suprised how little pressure it takes to squirt fuel everywhere and the pulsing in the feed gives me a niggling feeling of a duff supply, the problem is after all confirmed as fuel related. Having said that not being able to pull 4-5 or 6'000RPM with no load would suggest almost no fuel is getting through.

D

OK, was just a thought. Wasn't sure how long it had been off road.

That is further helpful advice concerning the Lexia use Dean and I have sent a plea for help to Dave(theRave)! I can now confirm that the fuel pump is definitely not at fault as I have tried another known good pump with no change. A direct 12v supply was also connected direct to the pump in case there was an issue with wiring. The problem is certainly proving evasive and I am now thinking could it be a wiring/sensing issue that does not show up through basic checks? Hopefully the application of a Lexia can be arranged and will reveal all. Regards, Matthew T.

I wonder if the throttle pot isn't being sensed at the ECU, so fuelling is not increased from idle condition.

Lexia will reveal that easily if it is, sensor could be working well but if there is a broken wire somewhere..........

Matthew is the main ecu earth attached and well grounded?, its the blue wire bolted on the inside of the O/S wing.

D

- and all the others while you're at it, if not already swatted. One more thing ticked off.

A lot of the engine ecu is earthed (yellow wires) to the intake manifold which is odd because the manifold is sealed on rubber 'O' rings, it would be better to earth them to the heads, block or even better run them back to the chassis or battery for the sake of a couple of feet of extra wire but thats not really important.
The ECU itself is earthed direct to the chassis via this blue wire, i think (not checked wiring diagrams) that one reason the ECU itself would need to be earthed is to ground out the injectors to trigger them, if this earth was loose or not connected the ECU would have problems triggering the injectors especially at higher revs and under load.

Wills suggestion of poor or no data from the TPS is a good one and could well be the issue, the lexia will show that very simply.

D

Thanks for your further comments chaps and why would the tps not be sensed at the ecu Will? As mentioned earlier the tps sensor was also unplugged and a fault code stored. I also think it is a more general issue as the engine even sounds a bit starved at tickover!

I have checked and cleaned all the earths Dean, including to the o/s wing, while on this car the original plugs have been removed and replaced with bolted tabs. I think the ecu earth wire is yellow as I have just taken a further look and there is an additional thin grey wire that has been t'ed in adjacent (not factory).

It is also possible to confidently further rule out the fuel pump, filter, and associated pipework and electrics, since earlier this evening I temporarily rigged up a spare known good pump above the engine. I do not think there is any problem with rail pressure or injectors, there just has to be something electrical/electronic holding back the supply!

Regards,

Matthew T.

Sorry Matthew, i just assumed as both my looms have a blue earth wire.

The ECU detects open or shorted circuits very easily and logs them as a major fault, this brings on the EML light.
If there was a high resistance plug or bad earth/power supply though you could get an inacurate signal, im sure the TPS does nothing more than give a variable voltage depending on the degree of rotation of the thottle plate, say 1v closed varying infinately to 5v fully open, if your TPS only ever puts out say 2v then the ecu thinks there is a 20% throttle opening when in fact there could be a 100% opening so when it also calculates fueling by doing some maths using the manifold pressure and temp for air density and crank for engine speed it estimates there is X amount of air entering the engine and injects the correct amount of fuel for that volume of air, the trouble is if your throttle is 100% open there is far more air entering the engine and you end up with a very lean mix and the engine wont run.

The ecu is still happy, it see's input from all sensors, its doing its maths right and all is well in the world, lack of engine speed will not upset the ECU as it has no way of knowing what load you are placing on it other than maybe the MAP sensor at Part throttle, you could be towing your house up mount everest for instance.


D

I would meter each ECU input and output in turn, which I feel, pinpoint the problem.