Try from here http://p38webshop.co.uk/index.php?route=product/category&path=71 (the last one), when you click on it, it should immediately download the pdf and open it in Acrobat. This assumes from a laptop or desktop, no idea what it will do on a phone though.

Marty's pdf of how to do it is very good, even my mate Danny who is as ham fisted as they come was able to follow and get a full set of pixels. Marty's instructions are here http://p38webshop.co.uk/RRdocs/P38%20HEVAC%20Pixel%20Fix%20v2.pdf and may be more comprehensive than anything from RPi.

It sounds as though you have dealt with all the likely places. Are you sure it is still coming in or if this isn't just water that got in there before you fixed the leak and hasn't evaporated away yet?

That is the advantage with a diff from Ashcrofts, all the pre-load and backlash is set up by them so it is just a case of fitting it. The main problem with the weight is that it isn't in one place. The end that you can't put a jack under is the heavy end and there's no really flat surface to sit it on.

This is what I use anywhere that calls for an RTV silicone https://www.autodoc.co.uk/reinz/982547. My local factors (Millfield Autoparts) stocks it. Not cheap but the Victor Reinz stuff is recommended by Porsche amongst others and it is very easy to use being in the aerosol type dispenser. The only problem is once opened the nozzle gets clogged so you may end up throwing half of it away.

Yes, I've got an Ashcroft on the rear and another from a different rebuilder on the front (a place up north who could do one for me in a couple of days whereas Ashcroft gave me a lead time of a month). The front has leaked slightly from the pinion seal from day one.

Correct, later 4.6 has a 4 pin as standard, earlier ones and diesel and 4.0 litre have a 2 pin. Diff ratio is 3.54 on all P38 versions.

Only if you spend more time crawling over rocks than driving on the road. You've got traction control that does the job on all but extreme surfaces.

Yes it is. Drop off the rear propshaft at the diff and tie it up out of the way, pull both hubs and halfshafts out by a couple of inches (you can even leave the brake callipers in place), undo the nuts holding the diff in place, whack it with a mallet to break the RTV seal and keep your hands out of the way as it drops as it is heavy. Hardest part of the whole job is lifting the new one into place..

Is it one tap per revolution? If it isn't peeing it down tomorrow, I'll get under mine and see if I can record the noise I have.

Mine is only there at idle and sounds horrible when I pull up at passport control when I'm getting on a ferry as it reflects off the side of the booth and the engine is nice and hot then. Give it a bit of revs and it goes away. I can just hear it from outside the car when started from cold and louder when warm but, as I say, it has always been there, never got better or worse. If I get an excuse to separate the engine from the gearbox again, I'll probably replace the flywheel and if it is still there then I really have no idea where it is coming from.

The spare also does it but to a much lesser extent and I've heard a similar noise from others too.

If a liner was slipping you'd be pressurising the cooling system. Does it sound louder from underneath the car? I've had a tapping noise from mine since I bought it at 205k and decided it was probably a cam follower, changed them and it made no difference. At 287k I had the engine rebuilt and it was still there but while doing something else realised it was sounding louder under the car like it was coming from the bellhousing. Decided it must be the torque converter but that was replaced, along with the flex plate, when I changed my gearbox about a year ago at 454k and the tapping is still there. With the lower cover off the flywheel (separate on a GEMS but part of the sump on a Thor) and the engine running, it looks like the toothed ring for the crank position sensor is fractionally offset so it could be one tooth on that just touching the crank position sensor as it turns. I've had a good look at it and none of the teeth looked 'shiny' like they had touching anything, they are all dead straight and the gap is perfect on all and there are no witness marks on the crank position sensor either. It sounds louder when hot but has never got any worse or better, so I just ignore it. It would be nice if it wasn't there though.....

Garvin wrote:

then it dawned on me that I could take the whole half shaft and hub out with the shield still attached!

and with the brake disc and calliper carrier too. For some reason rears usually come out easily enough but fronts can put up a fight, made more difficult as the ABS sensor has to be persuaded to come out in one piece as well.

I don't think any are actually clogged, but if the spray pattern is screwed or flow is restricted, then that could cause the lumpy idle. It doesn't sound to have a misfire as such, it just isn't smooth. I think taking them out and backflushing them with brake cleaner is my next step. I could just ignore it and do as I have since I've owned it and only run on petrol if I really have to. I've always said the petrol in the tank is regarded the same as the spare wheel, for use in emergencies only.

Pete12345 wrote:

If you usually run on LPG, maybe the £20 of petrol has lifted something from the tank bottom ?

Does Nanocom say which cylinder is misfiring ? If so, I would take that one out & do a bench clean with the 9V battery & carb cleaner.

Maybe, but it was running the same before I put the extra in so unlikely. Thinking about it, it ran much the same last year when I tried it on E85 (which Nigelbb has been running his Thor on) and I decided it didn't like it. But it may be that it wasn't the fuel but the fuel system. Nanocom doesn't give any P030x codes on a GEMS, so no idea which one, or ones, isn't running right.

I've still got the 3 Ohm wirewound resistor I used to limit the current when I did them last time, so may well do that.

Aragorn wrote:

i imagine a lot of these cleaners are somewhat snakeoily...

On mine, i found the injectors from a Rover K series were a direct fit and had similar flow characteristics, while being a much more modern design with better atomisation etc, so i just swapped them out. I can dig out the part number if you like? Was cheaper than paying to get the old ones cleaned.

That's why I asked in the first place, to try to avoid spending money on snake oil. I must admit I am a little sceptical anyway as when I did them with brake cleaner and a battery last time I found back flushing them got more dirt out and nothing you put in the tank is going to do that.

Interesting that K series injectors are a direct fit. Most K series engines used the earlier MEMS so that would suggest the electrical characteristics would be similar (if not the identical) and there isn't much difference in capacity per cylinder either.

Did that about 210k miles ago when the engine was rebuilt so it might need doing again but something I can chuck in the tank would be a lot easier. I think the most it has ever had in the tank since I've owned it is just under half a tank and that was only because I was setting off to drive across Europe and wasn't sure if I would be able to get LPG. As it turned out it is easier to find in Europe than here so I think that half a tank lasted the best part of a year!

However, one thought has just occurred to me. If I put something in the tank to clean the injectors, I'm going to have to run on petrol to let it do its stuff, so maybe taking them out will be cheaper and more effective......

Lpgc wrote:

It could be a petrol injector problem or could be the failed lambda sensor / learned fuel trims / less likely learned position for the IAV that's the problem?

First thing I did was reset the adaptive values in case that was the problem and it made no difference.

Or even something (Pitagora?) that's not connecting all the petrol injectors when it's running on petrol?

If that was the case it would be running on 7 when on petrol but it isn't. All cylinders are firing but the idle is lumpy rather than a consistent misfire.

I might give the stuff Dai recommends unless any has any better ideas.

As most of you know, my GEMS P38 runs on LPG. Unlike the later multipoint systems it has a single point so LPG enters just before the throttle body and the output from a lambda sensor adjusts a stepper motor valve in the LPG feed to keep the mixture correct. Because it doesn't slave off the petrol system, it can run on LPG from stone cold. It is set to change over at 1,100rpm on deceleration which means it starts on petrol, the revs rise initially and as they fall, it changes over and runs on LPG from that moment onwards. I keep some petrol in the tank for starting and in case I run out of LPG. Which I did a couple of days ago. My local Flogas charges 78p per litre and the two other filling stations nearby are £1.05 and £1.09 so I try to avoid them. I ran out of LPG a couple of days ago, so ran on petrol to Flogas but was too late, they had just closed. As I needed to use the car that evening, I bunged £20 of petrol in only to find it runs like a dog!

The idle on LPG is perfectly smooth but on petrol it is lumpy, accelerating up a slip road it could only just hold the same speed rather than accelerating as it would normally so seems down on power, although once I had got it up to cruising speed it felt normal. As I stopped at a roundabout, it stumbled and died but restarted without problem. The lumpiness clears once the revs are up to 1,000 rpm or so, so it is primarily an idle problem. As idle is controlled by the idle air valve, I gave that a clean (but thinking about it, that is still used when on LPG where there isn't a problem) and the petrol filter was replaced about 10k miles ago (I bought it when I first got the car 13 years and almost 300k miles ago and found it in the garage recently so decided to change it as I could). There is a separate 0-1V lambda sensor in the right bank exhaust downpipe solely used to drive the LPG system so the only part of the petrol injection system used by the LPG install is the throttle position sensor and that is only used to shut off the LPG on the overrun.

When running on LPG the petrol injectors are switched off but the fuel pump stays on and just circulates the fuel back to the tank. On petrol the RH bank lambda sensor switches as it should but the LH bank one stays pinned at 5V suggesting a lean mixture, yet the exhaust smells rich. Which means the only thing left would appear to be the injectors. My theory is that one of the LH bank injectors is partially clogged so the exhaust on that bank is showing as lean, the ECU is richening the mixture (Nanocom shows short term fuel trim on that bank pinned at +38.5%) but not all are getting their full quota of fuel so some cylinders are running lean while others are running rich. Hence me thinking the injectors could do with a clean. So the question is, has anyone used an injector cleaner, which one and did it make a noticeable difference? Or do I need to strip the injectors off and clean them for the rare occasion I actually run on petrol?

Marks could be signs of a liner leak but they could equally be slight distortion of the fire ring. One other thing I noticed is that the Elring logo and part number are still easily visible on the gasket which could suggest it has been apart before and wasn't torqued down fully. I assume it had stretch bolts rather than studs and doing the 90 degree turns can be a pain. You need a long bar so you can get enough grunt but then the bar hits against other things around it. If you put a long extension on the socket , it can twist when you are heaving on it so you don't get the full 90 degrees. Did they come out easily or did you need to use a breaker bar on them?

Gasket looks reasonable I'll agree but one of the fire rings is split and the stains on the head show a leak at one end. The pistons and combustion chambers aren't showing the normal steam cleaning you'd expect with coolant getting in although orange tip to the spark plugs is usually a sign that anti-freeze has been getting into the combustion chamber. It may be that you have a very slight leak around the outside of a liner, so it hasn't slipped but may be leaking. If that is the case, the options are a rebuild with top hat liners, or a dose of water glass in the coolant depending on how much mileage you do in it. The water glass will last for maybe 50k miles or so.

Tyres hit resonance at around 55-60mph so a rear wheel badly out of balance would cause it too.

mad-as wrote:

i could make it keep running if i pumped the throttle heaps but it would smoke out the shed in seconds

I bought a spares or repairs project P38 that was exactly the same. After pumping the throttle constantly, once running I could get it to run reasonably well. The pattern MAF sensors will usually work pretty well but if checked with an OBD reader, the airflow readings will be incorrect.

PC38 wrote:

I realise I should have put this in the electrikery section.

But then again, it's fuelling so that can leak....

Do you have an OBD reader that can show live values? If you do, check the lambda sensor and MAF outputs as it is, then disconnect the lambda sensors from the LPG system and see if it is the same. It may be that the LPG ECU is supplying a bias voltage.

The Zirconia sensors used on a GEMS work the opposite way round to the Titania sensors used on the Thor and most other cars, so 5V on a GEMS is lean with 0V being rich. That means it will be chucking more fuel in to try to get the voltage to come down. What diagnostics have you got? Anything that will show what airflow the MAF is reporting?