Xantia Glow Plugs Stopped Working

[BalloonFish]

Hello everyone,

My 1996 Xantia 1.9TD SX suddenly became really hard to start. Here's the full story:

Having always been a good, clean starter the car suddenly became a hesitant starter back in May - needing a bit of churning on the starter, lumpy idle for a few seconds and lots of blue smoke. Had the local Citroen specialist pull the glow plugs while I watched and two were duds - complete failure, no glow at all even when jump-leaded straight to a battery. Four new NGK plugs later, back to normal and correct operation - quick, clean, steady starting.

Used the car on Wednesday afternoon, no problems at all.

Thursday morning - went to start, the glow plug light came on for the usual amount of time but the engine wouldn't catch. Cycled the plugs a few times, some signs of life from the engine, sometimes one cylinder would fire but not enough to catch and run. Finally I got it to run by cranking it for a good five seconds or so with full throttle - lots of blue smoke, lots of diesel knock.

Drove about 30 miles, everything else is fine and the engine is running and pulling absolutely fine (I had wondered if it was a fuel problem). Left the car for 1.5 hours, so when I got back it was 'cool' but not properly cold. Again, the glow plug light came on but with no obvious effect - it still needed cranking on full throttle to get the engine to start and keep running. Just to go down the fuel supply angle - the primer bulb is full of fuel, giving it a few squeezes before going to start makes no difference, the car drives and performs perfectly once the engine is running and it gets more difficult to start the colder the engine is; I really don't think it's a fuel issue.

I've done some basic checks with the multimeter. All four plugs seem to have roughly the same resistance reading in the 1.3-1.6 range. With the ignition switched on and the glow plug light glowing I can get a 12-volt reading from the #2 plug (I think that's the one if PSA engines count their cylinders backwards - the second plug in from the flywheel end) which is the easiest to get to but also the one which has the power cable attached to it to feed to the loom that joins all four plugs. I can hear the two clicks from the relays under the bonnet - one when the glow plug light goes off and one a few seconds later when the voltage reading at the #2 plug goes off, which I assume is the plug controller dropping the plugs into their reduced-power 'after start' mode and then turning the power off altogether.

I cannot get a meaningful voltage reading at any of the other plugs, but it's hard to get the probes properly onto the plug heads while holding the meter and going back and forth to turn the ignition on and off in the time available before the relay switches off.

I'm going to have another poke around this weekend with an assistant so I may be able to trace things better, but I thought I'd ask your advice. Any ideas what I should be looking for? Can you have sudden overnight failure of all the plugs? Is it possible to start an XUD without any of the plugs working or do I have at least one working as my haphazard initial multimeter readings would suggest? If the light is on and the relay is clicking then must the plugs be getting power? If so am I looking at some problem either with the wiring loom to the plugs or the plugs themselves?

Any thoughts, ideas and suggestions appreciated!

[white exec]

You've done some good investigation and voltage checks. Everything points to GPs, not fuelling.

NGK plugs are not regarded as the best in terms of performance; general preference is for BERU brand.

Note - To check each GP resistance individually, the cable connecting them together must be disconnected, or you will just be reading all four together, in parallel.
_____

The GP relay is responsible for three things
- switching on pre-heating
- maintaining post-heating after the engine is started
- operation of GP dash light (which extinguishes to indicate when ok to start, and not when heating stops)

It computes the heating time (pre- and post-) by looking at the coolant temperature. Post-heating time can be anything from a few seconds to 3 minutes. It always applies full available battery voltage to all four of the GPs simultaneously; there is no 'reduced power' mode.

If the GP relay is ever replaced, important that the correct one (by model/part no.) is fitted, as functions on these relays vary according to vehicle.

The current taken by (OE BERU) GPs is approx 9-10A each at 11v. (11v is specified because that is a typical battery voltage when the heavy load - at least 45A - is operating, starter motor on top of that.)

This heavy current requires all the contacts of battery, GP relay and GPs to be in good condition. Over time, these oxidise and deteriorate.

Would suggest:

- Check out resistance of the GPs individually, just in case one has failed. Reading these very low resistances isn't easy, but it should enable you pick out any duff ones.
- Checking the condition of the 'bus bar' cable linking the four GPs, cleaning up the ring contacts there, and renewing the cable if needed.
- Check the HD contacts on the underside of the GP relay; unbolt and clean up.
- Check the HD contacts (can be two pairs) inside the GP relay. Relay cover can be prised off.
- Check the main feed cable from the battery/engine bay fusebox to the GP relay.
- Check relay base eye connectors (and adjacent insulation) for signs of overheating.
All eye connectors on GP circuits should be soldered, not just crimped.

- Check out the long supply cable between GP relay and the GPs. For XM cars (and I'm guessing maybe for Xantia too) built around 1996, this cable was inexplicably reduced to just 2.5mm², which is extremely marginal for the load it has to cope with. The cable arriving at the GP relay is 4mm², and that leaving it really ought to be at least the same. (JFI, originally (on XM) there were a pair of exit cables, feeding two GPs each. Somehow, this got reduced to one... ) **

As you've found, starting with no or partially-working GPs is possible, but only by using the heat produced by lots of compression/cranking. With properly working GPs, starting should be more or less instant on the key.

Hope helpful.
Let us know how things go.

** In my own case, I ran a new 4mm² cable (neatly clipped to the outside of the loom) to replace the 2.5mm² one inside the loom. This dramatically improved GP current and produced instant starting in all conditions.

[BalloonFish]

Thank you for that detailed response, white exec!

My early findings:

- Confirmed I could not get a voltage reading at plugs #4, #3 and #1, but a steady 12-point-something volts at #2, with the relay clicking on and off as I'd expect. This is with one multimeter probe on the threaded part of the glow plug terminal (above the ring connector nut) and the other on a nearby spot on the cylinder head.

- Took off the loom connecting the four plugs (man, what an awkward job - I think Citroen deliberately made every fuel line, injector pipe, ancillary bracket and intercooler hose to get in the way!). It looks OK - I mean, it's been on the car for 23 years but no obvious corrosion, breaks, bad connections/crimps or broken insulation. The ring connectors could probably be cleaner but I wouldn't say they were suspiciously dirty. I'll clean them up and see if it makes a difference. Since I had the multimeter out, I tested the loom for resistance and continuity (an inexact test, I know) - all four ring connectors showed the same resistance as the meter (0.6) and there was full continuity from end to end and at each ring.

- Made individual resistance readings with the loom off. All showed 1.4 ohms apart from #2 which showed 1.3. Again, taken between the threaded part of the plug and the cylinder head.

- With the power cable from the relay now 'free' at the engine end, checked the voltage there, which was 12.4V with the glow plug system active. Inspected that cable back to the relay as best I could and it seemed fine.

- If I'm right the plug relay is the one in a plastic bag attached to the inner wing squeezed down the side of the battery. I chickened out of removing the battery, unbolting the bag and looking at the relay itself since I'm working on the street outside my house and I'm wary of both dismantling too much of the car and disabling it at the roadside (I have no off-street parking). The voltage readings and the clunks/clicks from the relay suggest to me that it's working as it should, although of course that doesn't speak for the condition of the contacts as you suggested.

- I connected #2 plug directly to the power cable from the relay (no GP loom) and the engine 'started' in the same way as before - 5-10 seconds of lifeless cranking at full throttle before it chugged into life with a lot of knock and blue/white exhaust smoke, just as you'd expect from an indirect injection engine with only one plug. So I don't think doing that has made the problem any worse and that I've probably only had one working glow plug since Thursday morning.

- While struggling to the get the loom nut off #4 plug (behind the injector pump) I noticed that the little 'arm' on the microswitch which sits on top of the throttle arm on the pump and hits a cam as it goes above a fast idle was jammed at a funny angle, and as soon as I prodded the arm it snapped off altogether. I'm sure I read somewhere that the purpose of this thing is to switch off the after-start phase of the glow plugs above a certain throttle setting (the engine is an all-mechanical one in other respects - waxstat fast idle system, immobiliser acting on the stop solenoid and so on. No ECUs other than the one for the climate control.). In any case, I checked the voltage at the engine end of the GP power cable and it retained 12 volts regardless of whether I pressed the contacts on the microswitch in or out. This may be wholly irrelevant but I thought I'd mention it.

It's interesting that NGK plugs aren't highly thought of. I wonder if I'd already lost one or two since they were fitted back in May and last week I lost a third which was enough to make the engine hard to start on a colder morning than normal? I'm still slightly concerned if I'm burning through glow plugs at such a rate - they've lasted four months and under 2000 miles.

[white exec]

Sounds like you found the GP relay. Distinctive appearance, and looks like this (#1 in the diag):

You're detecting voltage arriving at the GP rail (glow plug cable, #3), but it's possible that the GPs are not pulling the proper current each (approx 10A), either because they are faulty, or down to a bad contact somewhere.

If you have an ammeter (which can read 10A d.c. or so), you can check each GP in situ with a direct 12v connection from the battery. Without that, you'll need to remove the GPs and bench-test them. Tip should glow orange-hot within 5 secs or so. (Do not exceed 10 secs.) If bad, replace with BERU. (You can't wear GPs out; but poor quality ones don't have a long life. BERU normally last 10-15 years+ minimum.)

If GPs are ok, go to work on the external and internal contacts of the GP relay. One poor contact is all it takes to block the high current (c 45A) needed for all four plugs to work properly. A contact (or cable) with resistance might well show voltage present, but be unable to deliver the necessary current.

I know the XUD pre-heating arrangement fairly well, but am not familiar with the microswitch contact on top of the pump. Someone else here should be able to advise.

IIRC, removing three of the Xantia GPs is straightforward, just iirc No.4 being a sod. Even bench-testing just three will be a step forward, though. When replacing, put a decent smear of copper grease on their threads to enable future removal.
______________

One of these clip-on d.c. Ammeters will allow you to check what current is being delivered along the line...

https://www.amazon.co.uk/Sealey-BT98-11 ... 592&sr=8-6
Avoids the need to cut into a cable; just clips over it.

[Gibbo2286]

I have to say that in my experience nine times out of ten the relay is a goner if the plugs test out ok, burned internal copper connections similar to those on a starter solenoid.

[BalloonFish]

I had time to do a little more exploration today.

I took out the battery and un-bagged the glow plug relay. Even allowing for the fact that it's been protected by the plastic shroud it seemed very new-looking, with what I would say appears to be more modern (and cheap-looking) plastic than a 1996 factory-fit part. It is branded as a Bosch 'Gluhzeitsteuergerat' so it's clearly a genuine part but it's also stamped 'Made in Malaysia' so I'm guessing it's a more recent replacement part? Or maybe my standards are too high and Bosch were making control boxes out of cheap plastic in Malaysia even in the mid '90s.

Digression aside, the two ring connectors on the bottom both looked pristine - no dirt, no corrosion, no signs of burning or overloaded and no problems with the connectors or insulation. The same went for the multiplug connector as well. The other thing that made me wonder if this was not an original relay was that there seemed to be no way of getting the cover off to look at the inner workings - it seemed to be a sealed unit with a rivet visible next to the ring connectors on the bottom. I did give it a squeeze, prod, pull and poke but couldn't see any signs of the cover and the bottom being separate mouldings and I didn't want to break it, so I have no report on the innards. I did give it a sniff though and it didn't smell of burnt-out insulation or overheated plastic (which I have smelt on other electric parts when they've given up the ghost).

I had a go at getting the glow plugs out but the access defeated me with my selection of tools - it's just not possible without cranked/offset spanners and/or suitable deep sockets and I have neither. I did manage to tease out #2 glowplug with a normal spanner a few degrees at a time, and it glowed properly when wired direct to a battery. So that one at least works...but that was the one that was getting a voltage reading and the fractionally lower resistance reading anyway. The others I just couldn't get to properly - even #1 I couldn't get the spanner to bite well enough and all I was doing was chewing the corners off the hex bit of the plug.

[white exec]

OK. Best not to chew up the GPs.

Time for a definitive test:
1. make sure all 4 GPs are decently connected together, as per original wiring
2. unbolt one of the large eye-connectors from the underside of the GP relay - the one that goes to the GPs
3. extending it if necessary (stout cable, at least 2.5mm²), so that it can be briefly connected to + on the battery
4. with IGN switched on, touch the cable to batt+ and keep it there for 10secs, then, still connected, try starting on the key
5. see how well it starts
6. disconnect from batt+ after no more than 30secs after starting (or failed attempts).

What you'll have done, manually, is 10secs of pre-heat, and 30secs of post-heat.

If engine starts well, GPs are ok.
If not, replace the GP relay.

If you can post your VIN number, I'll find the correct GP relay for you.
As you say, what's there might not be original.

(If engine doesn't start properly, the only thing that test isn't checking out is the cable running from GP relay to GP 'rail', but failure of the cable is a bit unlikely. You could check the cable out by temporarily running another one (min of 2.5mm² again) from batt+ to GP rail.)

[BalloonFish]

My problem seems to be solved.

Long story short, I didn't have the time or tools to properly carry out the excellent advice given in this thread, so took it back to the specialist who replaced the glow plugs in the spring. He went through them with his multimeter and was able to get decent readings (voltage and resistance) off all four plugs - I think because he had better multimeter probes with longer 'needles'. Anyway, he pulled two of the plugs (#2 and #1) and they both glowed properly when jumped straight to a battery.

He agreed that the loom connecting the four GPs was in decent condition and that the relay was functioning correctly. We scratched our heads for a bit and decided to put it all back together and go from there. In doing so he noticed that the ring connector on the cable from the relay to the loom (that sits on #2 plug) was slightly dirty. Not corroded, not mucky, not verdigrised, just a bit tarnished. Since it wouldn't do any harm, he cleaned it up with a bit of emery paper and put the cable/loom back together with new securing nuts on the GPs.

And now it starts as good as new - instant, clean, smooth starting. I guess that slight layer of dirt on the ring connector feeding the entire system was enough to restrict the current flow, but I was still getting good voltage at that plug. So the lesson (as already identified by white exec) is that you need really, really good electrical connections all through the system, especially on the hard-worked single cable between relay and engine.

Thanks for the advice everyone!

[Pug_XUD_KeenAmateur]

cor, one of those simple issues that's easy to overlook. Made a note of that in my references in case it ever happens to me.

[white exec]

Glad you got to the bottom of it in the end. Well done in seeing it through.

A tiny bit of resistance can play havoc with the way high current circuits operate.
Common black spots include main battery connections, starter solenoid and motor, alternator connection, the IGN switch, and good old glow plugs. Oh - and grounding/earth connections.