Mighty Xantia 1.9TD suffers power loss at Hill starts

[masood_ilyas]

Xantia 1.9TD Automatic 1999 model with a healthy 135K miles on the clock.
My beloved Turbo Diesel Automatic Xantia has been faithful for the last 5 years has just started behaving rather oddly showing signs of loss of power yet has plenty of grunt once it starts moving.
I have looked at the archive of our FCF forum but there is nothing that could explain this loss.
When there is a high load e.g starting up on a hill or doing a gentle stop start at a harsh road hump the car exhibits an effect similar to clutch slipping, but this is an automatic with clever electronics!. Some sort of protection appears to kick in and limits the rpm to just about 1000 taking no notice of the accelerator pedal position even when floored. This results in a dead stop with having to reverse back down the slope to build up speed and momentum for a take off at the ramp.
I have looked for obvious signs without much joy . The diesel leak off tubes and injectors are fine as the car start first time even on frosty mornings. Once it is moving there is a slight delay in acceleration but manages to get up to speed in no time suggesting no problem with timing etc. The cam belt was changed last summer with no ill effects, this development is a recent phenomenon which is present whether the engine is cold or hot after a long run. The same effect is experienced when using reverse gear. This can be very embarrassing for a proud xantia driver and annoying for the bloke in the Fiesta behind.
I can only think of two possibilities:
1. The automatic gearbox needs attention as it may be telling the engine to pull its throttle back or
2. The engine management system is kicking in to prevent me from destroying the planet from excessive emission and limiting the RPMs.
The exhaust pipes were renewed last summer with no sign leakages at either end from down pipe to the end box.
This car also has an EGR system with a vacuum pump at the end of camshaft near the hydraulic fluid reservoir, but I do not know enough to link this with the fault.


I would love to hear from some one who may have experienced this or any one who may spot what I am missing. I would like to know what is it that is preventing the fuel from reaching the compression chambers as otherwise the car manages speeds over 70mph and revs of up to 4K
The fuel pump is a Bosch variety.



Any advice will be greatfully received.
Thank you

Masood Ilyas

[RichardW]

I'm pretty sure that all the 1.9TDs have a physical accelerator rather than a fly-by-wire system - in which case you can rule out the engine ECU - you need to check though - is the acclerator cable attached to the injection pump, or does it go to an electrical device?

I suspect that, unless there is a fault with the brakes or something, that you are looking at a gearbox problem What is the fluid like in the box, and is there enough in there?

Edit: Looking at your picture, the accelerator cable is attached direct to the pump - I don't think the ECU can override this input, so something is causing the engine to bog down - sorry, but it does sound rather like a gearbox fault has developed

[masood_ilyas]

Thanks Richard, I was dreading that diagnosis.

I had changed the gearbox oil last month to rule it out, alas it made no difference.
Breaks are fine too. What is odd is the engine just revs at about idle position, but if you put the gear into N the RPM will roar away so I am inclined to believe that the pump is being prevented by gearbox or something else if not the ECU, though I believe these cars have common engine management and gearbox controls module which sits next to water reservoir bottle, may be some one can confirm this!.
I do not think these cars have any Lambda sensors but might have other sensors for antipollution measures which could be giving duff feedback.

It seems that I will now have to take it to the stealer to have faults read out on their confuser, the print out no doubt confirm it will cost about £1800 to fix i.e uneconomical.

[CitroJim]

I've just been looking on pr.net.

The Auto TD has a different pump compared to the manual and furthemore, the auto has an "EXCITION SWITCH" associated with the throttle pedal and cable. I cannot see how it interacts but it appears to be an electrical and vacuum device.

I'll take a punt and say this may be the root cause and the auto box may just be totally blameless. I hope so.

My guess is that this EXCITION SWITCH somehow controls the special pump in some way and it could all be related to kick-down perhaps.

Worth checking the connections, both electrical and vacuum, to this device. It looks like a big tubular thing, not unlike a throttle "fly-by-wire" pot as found on 2.1TD engines. Goodness knows where it is physically located though.

[Kowalski]

The pump pictured looks almost exactly like the pump on my '97 TD, the throttle is completely manual as pictured.

There are some electrics / electronics on the pumps that disable the afterglow if you press the accelarator down a certain amount for a certain period, could the switch associated with that be an "excitation" switch?

[masood_ilyas]

I took the car to have the fault codes read out but these did not help my understanding of the nature of the real fault.
Citrojim, you are right the tests did not reveal any gearbox faults only the engine parameters appeared on the report.
The codes reported were:
1. Engine speed check (intermittent)
2. Throttle position couple (intermittent)
3. Idle solenoid valve (intermittent)
The tester was able to clear the fault's memory but not remove the faults.

To my mind all these must be ok as the car works fine most of the time.
Speed restricting to about 1k suggest the idle solenoid is being forced on but how can I check this. Also the throttle position coupling, this I guess is the electrical potentiometer sitting on the the top of the pump with three wires going to it. Disconnecting this makes the car go immediately into limp mode i.e lamps flashing on the panel and the throttle range is restricted to 1200RPM. Kick down does not operate either. This surely must be ok as the car does fly once moving a little.

I notice there are three electrical connectors on the pump if one of them is for theft prevention, one for throttle position or coupling sensor what is the third one for?

Still searching for answers, once I find out I will update in case it helps other people who might find themselves in my sitiuation.

[masood_ilyas]

I took the car into the local dealer for another opinion, the friendly engineer came for a drive and is of the opinion it might be either the turbo, diesel fuel pump or timing of the pump itself.
For their 1Hr minimum charge they will investigate a bit and let me know to decide whether it is economical to pursue.

If it was the Turbo I could have a bash at changing the turbo. On the other hand if it is the fuel pump I have no idea how easy or difficult this might be. Can any one help with these questions
1. As it is an automatic will it accept a pump from a manual car? both are mechanical types.
2. Will it fail the anti theft checks for key code and pump match. I remember reading on this forum a while ago that these pumps get a signal from the ECU to disable the stop solenoid but I am not sure if there is an element of security bit inside the pump which is unique and matched with the key before the ecu generates a disable command for the stop solenoid.
3. How easy or difficult will it be to swap with one and time it, can it be left roughly where the old pump is and hope the management system will automatically advance/retard itself for optimum.
4. How is internal timing adjusted I thought it was just a case of setting the height for a tdc on third cylinder +/- the timing figure stamped on the pump. (as shown in B.O.L)
Any helpful hints please, I am hoping I can get this fixed rather than give it a redundancy notice on grounds of efficiency, it is too good to be junked.

[MikeT]

The codes reported were:
1. Engine speed check (intermittent)
2. Throttle position couple (intermittent)
3. Idle solenoid valve (intermittent)
The tester was able to clear the fault's memory but not remove the faults.

To my mind all these must be ok as the car works fine most of the time.

8) Most of the time? As in, when not intermittently faulty? 8)

I mean, it looks like we got a wholesale intermittent fault here, doesn't it?

Do Citroen's use Powertrain modules that talk with the ECU?

[CitroJim]

Those ECU diagnostic codes look more like ones you'd expect to see from a petrol engine. There is no idle control valve on a Bosch Semi-Electronic pump so on that basis I'd be inclined to disregard what the ECU diagnostics say except for the engine speed check error reported.

The third connector on the pump, the lowest, is the timing adjustment servo.

I'd not suspect the turbo as they don't fail like that.

Timing is possible. it is under ECU control and it works lke this:

The ECU takes inputs from the coolant temperature sensor, the TDC sensor, to measure engine speed, and the throttle pot on top of the pump. Using informnation from these three sensors, it calculates the optimum injection timing for the prevailing conditions and sets this by applying a signal to the pump timing servo. The ECU then measures the actual timing by comparing the signal from the TDC sensor and the needle lift sensor on No. 3 injector. It then trims the timing in a feedback loop to make it correct.

Idle control is achieved mechanically by the pump governor. The fuelling "trim" applied to compenaste for an increase in turbo boost pressure is also achieved mechanically.

If the ECU is unable to confirm that it has set its desired timing it lights the "K" (Engine Management) light and sets the pump timing to a failsafe retarded value which will cause a massive power reduction.

My guess is that you have an intermittenly failed sensor, either the TDC sensor of Needle Lift sensor or the wiring between them and the ECU. I doubt it is the pump itself but if not a sensor, it could be the ECU although they are normally very trouble-free.

From the diagnostics, the first port of call has to be the TDC sensor.

Do you ever see the "K" light come on? It should always come on at first switchon before the engine starts to prove the bulb but extinguish as soon as the engine starts.

Does the tacho always work normally? It works thrugh the ECU via the TDC sensor but will also function from Needle Lift signals. Does it jump occasionally? If so, that will give a clue that a sensor may be responsible.

[masood_ilyas]

Thanks Citrojim, an excellent insight into the working of the pump.
The K light does come on initially but goes out before the beru have even warmed up. Once started there is no K light ever.
The rev counter is not erratic,gives steady indication rock solid not intermittent.
What I find perplexing is on straight runs the car is fine every time though a bit sluggish when starting to move. Once you go above 10mph it then flies easily into the 90s leading me to think the cambelt timing is probably in the right area. Also rules out collapsing tubes feeding the pump.
On steep hills starts, it is as good as dead always, not intermittent.
This appears to be Load related which impacts on how the fuel is delivered when faulty, as the RPMs do not rise above 1200, as good as idle speed.
I remember reading on another thread (could even have been your contribution citrojim) explaining the fuel delivery and back pressure but that went over my head as I could not visualise the pump, not having seen the insides of one.
I have checked the lift sensor connector and sprayed copious amounts of WD40 on the ECU connectors and other wiring areas without much change.
To learn about this pump I am tempted to take apart an old pump off a mk1 xantia but will it be the same in principal or will it be a totally different beast so waste of my time.

[CitroJim]

The more I think about this, Masood, the more I'm thinking the problem is fuel starvation, given it happens on hills and they symptoms you are experiencing are exactly the same as I've seen with fuel starvation. I'd do all the normal checks for air ingress into the fuel supply, especially around the filter housing. If you have not already done so, replace the fuel filter and ensure it is a good one you have used. I recently found Champion L141 filters don't fit well and cause air leaks. They're about 3mm longer than standard and cause a poor seal in the filter housing. I fitted one and had fuel starvation problems.

Air in the system will also upset timing because the timing servo uses the fuel as hydraulic fluid to effect the timing setting and if there is air present, the timing setting will fall to the fully retarded setting for a few moments.

It'll also be well worth checking your tank strainer to ensure it is not clogged and checking your fuel return line from the pump is 100% clear. Disconnect it and blow down it. If you can hear bubbling in the tank then it's nicely clear.

A pump from an early MK1 Xantia is very much the same, just lacking the timing servo. In its place is the electrically operated cold advance device which operates on a similar principle.

Unfortunately, taking one apart will only teach you how to take them apart; you may not glean much insight into how it works except to be staggered at just how complex and well engineered they are Take one to bits by all means but have a study of the operating principles and then all the bits in front of you will all make sense. A pretty good write-up is in the Citroen Technical Guide

http://www.tramontana.co.hu/citroen/guide/guide.php

[masood_ilyas]

Thanks Citrojim your helpful advice is appreciated.
You may have hit on the possible reason i.e air in the fuel supply retarding the timing. I noticed when the RPM was raised by rotating the throttle cam on the pump I could see through the clear tube small air bubbles arriving from the fuel filter to the banjo union. I had missed this earllier in my eagerness to find more complicated reasons like the gearbox etc.
It may have been since I changed the fuel filter last summer as part of its service. The filter was one of Halfrauds variety.
I added another clear tube at the supply side near the turbo intercooler pipes to check and sure enough there were no bubbles going into the filter.

What are the chances of the filter chamber itself becoming cracked or porus, I am no gorrila, I am sure the securing allen screws were set to correct torques.
Time to change the filter and renew rubber seals.

[CitroJim]

What are the chances of the filter chamber itself becoming cracked or porus

That is quite a common problem Masood and you can't guarantee a replacement from a scrappy will be good either. They're not made of the most wonderful plastic I don't think they're staggeringly expensive brand new.

[Monkeyfeet]

Masood,
For all of us following this one, please let us know when you've cracked it!

[masood_ilyas]

This problem has now been resolved.
The removal of the air bubles in the supply did improve the running a little bit but did not cure the original fault.
The car was taken to a very nice Auto specialist in London (Stephens Engineering) where it was diagnosed as having pressure regulation faults and a broken band. Given the age of the car and 122K on the clock it was recommended that the gearbox should be replaced with a reconditioned one. Gearbox was replaced, all the faults have now disappeared.

The gearbox was unable to deal safely with Hill starts or high load at road bumps when moving from standstill, the interconnection between Gearbox ECU and Engine ECU together deemed it necessary to cut off fuel so as to prevent me from applying too much torque and thrashing the internal bits of the autobox into pieces. There was too much difference between Engine RPM and calculated RPM from road speed,gear,load information.
Any AL4/DPO specialist on this forum may be able to explain this better than I can.
For other members who might be in this situation, it was an expensive fault to diagnose and most of us would probably think twice before spending £1460. Still you do get one year or 12K miles warranty to ease the pain.