Citroen XM 2.0i XU10 idle problem help required

[Deanxm]

Im going to throw this out for the collective to pour their ample minds into.

There is a 2.0i XM that drives fine, the guy working on it is very switched on, a professional and nice chap but he has hit a wall in terms of diagnosis.

The issue is when the engine is fully up to temperature only the idle speed will climb to about 12-1300rpm and then drop, nearly cut out and recover. When the car is in gear it will do the same but because of the extra load of the auto box being in gear the engine will cut out.
He has checked the ICV is working fine, removed it and fixed a mechanical metering device in place so the engine cannot (in theory) regulate its idle speed but it still does the same thing.

It drives fine, starts on the button, plenty of go, no issues at all but just the idle.

what say ye?

D

[CitroJim]

Dean, what injection system has this XM got on it? This was a common problem on the 205GTi with the Bosch LU2 JetTronic with the 'flapper' AFM...

I know some very early XMs had a 'flapper' AFM and if the one in question has this system I may be able to offer a few thoughts...

Too mentally wrecked to do it tonight; such things I can only do when I'm relatively brighter in the mornings!

[Deanxm]

Jim! It is a Bosch Motronic system that uses a MAP sensor and TPS to judge air intake volume (no MAF of any kind).
when you are suitably rested and feeling fighting fit i would be interested to hear what the issue is caused by on the Jetronic system as it may help us look in the right direction in this case.
You make sure you look after yourself though sir!

For the purpose of further comments Motronic uses the following.
TPS
Air inlet temp
Coolant temp
MAP sensor
Crank angle sensor
Wasted spark ignition with ignition amplifier (which as far as i can tell is just two solid state relays that are triggered by the ECU to ground the coil pack/s.

Any ideas would be great at this stage.

D

[CitroJim]

Dean, on such a system virtually all issues of this nature are likely to be caused by air leaks somewhere in the system given that all the usual suspects have been checked, cleared and eliminated...

Air leaks can both affect idle speed - as that's effectively what the ICV does anyway - bleeding in more or less air to regulate the idle - and mixture ratio by confusing the MAP sesnsor...

Those air leaks may be tiny and they can be anywhere south of the throttle body...

Basically, all joints on the inlet manifold need to be checked and all pipes leading off from it - especially around the crankcase breather system. It could even be caused by a duff PCV in the breather system...

Also, any hoses leading to/from the ICV and MAP sensor too...

Any hoses connected to the inlet manifold may be perished and leaking even though they look good.

My thoughts on the varying idle speed is an air leak that is varying in size with temperature...

One odd one I once had was a leaking blanking cap on an inlet manifold... A cap blanking off an unused vacuum take-off spigot...

Hope that might be of some help and give some further ideas...

Zel, of this parish, fought lots of odd injection problems on his Lada and all came down to air leaks and perished pipes in the end...

I'll be interested to hear the outcome of all of this...

[Deanxm]

I do agree that it sounds like unmetered air, the ICV has been eliminated by removing it and partially blocking its port, the engine still revs the same though.
There are not many places for there to be an air leak, the crank case vent cannot be one as the port is metered by way of a tiny pin hole disc in the port, if you remove the breather pipe it makes no difference from experience, the full list of culprits would be-
Injector seals
Map sensor pipe
Evap pipe up to the second solenoid valve
Throttle body gasket
Manifold gasket
ICV
Also the throttle plate spindles could be worn but it's not something I have seen before.

I was half thinking of going to see the car myself on Monday if I had time but what puzzled me is why is the rpm jumping when the ICV is removed, i thought the ICV was not working and the jump in engine speed was due to ancillary load, so if the hydro pump cuts in, it drops the revs, if it were in gear too this would be too much and stall the engine but i have been told this has been checked and works fine.

I agree it sounds like un-metered air getting in though. The lambda signal has been confirmed as fine (swinging from .1-.9v) so i dont think it is fuel related or a mechanical cylinder related issue either.

D

[xantia_v6]

I think that the drop in rpm may be due to the over-run fuel cutoff deciding to save some fuel, so is a red herring.

Spraying ether or other combustible gas around the injectors and manifold joints may locate a leak.

Is there a vacuum port on the manifold that has lost its plug?

[Deanxm]

Here is the engine in question, there is one blank port on the right next to the ICV, on the back of the intake is the crankcase breather connection but this union has a metering disc in it (part of the union) so removing the pipe, or opening the oil cap will not cause any vacuum leak symptoms.
On the left at the front is the MAP sensor pipe that dives off into the ECU box and connects to the ECU via a small pot, behind that is the line from the charcoal canister, this can be disconnected from the solenoid valves on the canister side and not cause an issue either as the valves themselves limit the air leak.

The overrun fuel cutoff is activated only on a trailing (closed) throttle and only when engine revs are above 1800rpm i believe, so the engine should not be cutting fuel as it is only revving up to about 1300rpm.

As much as I argue against you I do so with the information i have been given and the gentleman insists there are no vac leaks, the items that require checking (as you see from the picture above) are not difficult to access.

Is there any credibility to the idea the ECU is having trouble 'seeing' its crank speed but while still maintaining enough of a signal to run well under load/revs.

I am starting to wonder if the fluctuating revs is ancillary loads and somehow the engine has lost its ability to control the ICV but then the ICV does move and i believe performs its two full sweeps of valve travel when the ignition is first turned on so.......

D

[xantia_v6]

Idle speed is governed 99% by restricting the air supply, so too much speed === too much air. There is nothing that you can do with fuelling or ignition to make a correctly tuned engine idle faster.

There is a possibility that the engine is not correctly tuned, and the air supply has been increased to compensate, but that seems unlikely.

Disconnecting the throttle pot will disable the fuel cutoff and remove it from the equation.

[Deanxm]

Fair point, I guess if it looks and smells brown, the chances are it's brown, or in this case a vacuum leak.

Given the guy has been over it and found no vac leak, not sure if he used propane, carb cleaner, smoke tested it or what but he says there is no leak, the only thing I am now thinking is the Canister purge valves, one opens when the ignition is turned on while the other is closed with ignition on, then when the manifold pressure is suitable the closed one fires almost like an injector open and shut fairly rapidly to allow the fuel vapour from the canister to enter the intake, now if this valve failed it would constantly draw air from the tank which normally would colapse the tank if the valve opens when there is high vac in the intake but the pipe in the wheel arch to the canister almost always cracks and leaks which you would never notice but it would create a vac leak if it does and you get a solenoid fail as well.

Worth a check I think as this would be hard to find with most methods and impossible to find with propane or some other form of volatile fuel sprayed around the intake area.

D

[CitroJim]

The purge valve, canister and piping to it is a good call...

This was implicitly included in my advice to check all south of the throttle body...

So often the purge valve and carbon canister is overlooked as both are often hidden...

The valve can be disconnected as a quick test and also an ELIT/Lexia session will show when purging is active.

[Jan-hendrik]

Was this problem ever solved?
I've just noticed my ES9 has started to idle faster than before and rpm does not increase when turning the steering wheel.