Interesting read. Where are you based?

I would put an oscilloscope on the o2 sensors and confirm that they're switching properly and at the right rate (roughly once a second).

You also mention the injectors are ok. Can you elaborate on that? It could be one or more of them is leaking slightly, which would result in over fuelling at idle, but much less of an impact under load.

I'm still puzzled over the 39% STFT at start up. As far as I'm aware, as there is no output from the lambda sensors before the engine starts (and not for a few seconds after it has started) it should be running in open loop so the trims should be 0%. I'll have a look at the STFT on mine when starting it tomorrow if I remember (neighbours might object if I do it now.....).

Thanks Gents.

I'm SW London

Re injectors, I did a resistance test on them. and they all read consistently, but slightly higher than the range given in RAVE. I didn't write it down at the time, but RAVE talks about 13.3k +/-1.3k - i think they all came in at 16k, I took consistency to be a good thing, but since I didn't really get the significance of the resistance test left it at that. At the time i hadn't heard of any of the techniques of blasting the injectors etc, so have never physically tested for leaks.

I haven't used an oscilloscope since school :o) wiil it tell me anything that nano doesn't?I can see the switching on nano - seems fairly regular now, voltages 0-5v, and generating s.t trims of +/- 10%, but usually +/-5%. For the first few trips on new sensors they were giving rise to some wild ST trims, but ok now.

Thanks for looking into the 39% Richard, my neighbours must hate me - I just finished fitting a new alternator to the wife's Audi. I wonder if the 39% is a default strategy, but then if it is, why...what does it think has failed to execute that. It's a very specific 38.75% - same on both banks @ start

O2 readings from on nanocom are laggy, so will show it switching, but not how quickly, unless its ultra slow.

Resistance is really just checking the solenoid coil is intact. Good to check but is purely an electrical test.

38.75% is its maximum adjustment on the short term trims. Will wait and see what gilbertd's testing shows

Actually it occurs to me that it might be correct that its up at max at startup, that could be the way it does the choke. I'll plug my nanocom in later when I get a moment to check that, but it vaguely feels right.

Ok tested on my daily, from cold.

From initial start for ~ 5 seconds, trims are at zero, open loop.
For the next 30-60 seconds, it goes closed loop, and the short term fuel trims go up to max (38.75%). O2's are nailed at 5 volts as you'd expect.
After that the trims slowly reduce down towards zero, and the O2's begin switching.

So, it's normal.

Thanks Mace, good to know I can eliminate the O2 sensors from the investigation.

Mines the same. On a cold start, ignition in position 2, showing Open Loop and STFT at 0%, start engine and it stays at 0% but after 20 seconds or so, it changes to closed loop and the STFM starts to climb until it reaches +38.75% with the lambda sensor output staying at 5.07 V. After another minute or so, the lambda voltage starts flipping between 5V and 0V and the STFM drops down towards zero.

I'll be going out in it shortly so I'll see if it is the same on a hot start. As long as yours shows Open Loop and 0% STFM at start then it is correct but if it is showing the +38.75% before it starts that is the source of the richness.

As you said Richard. Open and zero stft at Ignition II, and then climbs fairly rapidly after start up to 38.75%. So I'll look elsewhere for a fix.

I was at the tip the other day, and had the same restart problem. Then I got to thinking that the last two times i'd been to the tip I had the same thing...so it's obviously not the fact that I'm at the tip - but the fact that I've run the car under full load to get there, turned it off for maybe a couple of minutes as I've chucked stuff in the skip, and then attempted a restart. Could the fuel rail super-heat in that time? Or is that not a "thing " in a P38?
Still haven't managed to get my old fuel filter off yet, but I'm hoping that will get me a few more psi of fuel pressure.

Ok, another difficult restart this evening. Car had been left about 10 mins. In that time the fuel rail temp was up to 70 degrees C. Presumably there 's an increase in pressure at this point too, but the fuel pressure regulator won't release until it has a vacuum ie engine running ? (I'm postulating, feel free to cry "bull s***" ). I need to find a way to keep my fuel rail a bit cooler when it's just sitting on a warm engine... It's dusted with black paint at the moment (though it's been there a while) getting rid of that might help a tiny bit, perhaps?

70 degrees isn't that hot, I've restarted without any problems after 200 odd miles, towing a trailer in ambient temperatures approaching 40 degrees so I suspect the fuel rail temperature would have been much higher. There's a connection on the back of the fuel pressure regulator to the inlet manifold so that fuel pressure is always kept at a fixed pressure relative to manifold depression. Is it possible that there's a bit of muck floating around in the fuel return line blocking it every so often? Fuel pressure will be higher when hot due to thermal expansion anyway but if it can't bleed the pressure off through the return it will be rich on a restart..

It's not impossible that there's some obstruction, and I still haven't managed to get the fuel filter off, that would tell me a lot about the state of the fuel system I suspect. I froze the filter/ fuel pipe coupling, that didn't work, and next I have some flare spanner-like vise grips coming. Unfortunately the whole connection is too narrow for me to get a crows foot on otherwise it could have been a quick job - but thus far I am in no hurry to strong arm it and end up ripping the end off the filter cannister. In the background I stuck some Redex in the tank today in the hope that that might help clean the injectors, rather than strip things down.

I have compressed air, is there a procedure for clearing the fuel lines/ rail ? or just blow....return to the rear, and feed toward the rail...

ok i have read this post front to back , i think your chasing your tail with all the trims and variables, you have forgotten the basic fundamentals off fuel injection, PRESSURE IS CRITICAL . running at 26psi at idle is not good, its 25% below optimum running pressure 36psi . all the variables are trying to work within their parameters but the fuel is not there so there all over the place . yes do the fuel filter first and recheck pressures , if you need to apply heat try and find an old electric plumbers soldering iron and use that , no open flame. if no improvement then its the pump. worn pumps have a habit of cavitating especially when the fuel is hot and this is where i think your problem is?
PS this my 2 cents worth and is a generic answer for all injection systems ,( not gems directly related) as i look at all systems to be the same . you must have noticed a drop in power or fuel usage to be different and in saying that these things can creep up on you and not notice that its not performing as it was until you have trouble towing the boat over that hill that you use to fly up .
PS don't forget the fuel pressure regulator as this can be an issue to

Mad-as, Thanks for having the patience to got through the whole thing. I'm trying to work through things using data as much as possible, without changing parts at random. So the only thing that's changed on the air side is IACV steps - everything else passed muster (with the exception of the IAT sensor, but perhaps that's due to wiring resistance). Once the new 02 sensors (exhaust needed changing, and old sensors were rusted in) settled I haven't needed to worry about them. As you say, then there's fuel - the fuel filter's held me up for a few weeks now, the restart issue's not bad enough for me to have gone nuclear on the fuel filter, mainly cos I don't want to lose the fuel line if i don't have to, and the fuel pump is last on my list because it falls into PITA-to-fix territory, relative to the others, at least. Sooner or later I'll declare war on the fuel filter - probably when I next get close to E - and I think I'll do another pressure test on the fuel rail as I know the data sounds low (although the GEMS data for rail pressure and FPR pressure seems hard to validate definitively) - but i really don't have any power issues with the engine. It sounds better now than it has ever done - although still probably runs rich, and hence I opted to Redex the injectors, in case they were dirty. I've had the car a year and a half now, and so don't know how it sounded/felt 10 yrs ago, but I do have the 4.0L V8 in my Defender to go by and it compares quite well (but richer as noted). Sadly I don't have a boat to tow over hills :o)

After a retest, pressure at the fuel rail is a constant 31 psi, but now there's no pressure response from either taking the vacuum pipe off the regulator, or pinching the return feed (with a brake clamp) Sadly I still haven't been able to get the filter off, the 3-way Irwin vise grip pliers were a bit of a waste of time. I'm starting to resign myself to getting new tank-end fuel pipes, but haven't tried the non-flame heat source yet, as I don't really have anything suitable lying around

Pressure test today was 26 with the FPR under vacuum, and 34 when not (pulled the pipe off the inlet manifold this time, rather than the FPR ). Clearly the pressure from the pump's not the greatest, but the regulator seems ready to cap it at 26 psi on occasions, so I'll keep my eye on that. Meanwhile the temp at the rail rises from the ambient temp of the fuel to about 70 degrees within 10-15 minutes of shut down, but there's no corresponding increase in the rail pressure - it stays steady.
The restart issue seems to occur when the rail is 60 degrees plus, since I'm not going to be able to get my fuel filter off any time soon then I'm going to put some reflective tape on the fuel rail as a heat shield, which may reduce the heating up on the fuel rail. I'm also going to change the thermostat which might help the cooling system manage the engine temp better. I had an over-heat the first weekend I got the car, which was mainly due to the PO having placed a second thermostat (which was clogged up) in the housing used for the 4.0 ie the front nose-piece on the inlet manifold. But I never replaced the lower thermostat, so possibly that's not as free flowing as it should be. If i'm bored I'll pop the water pump off too, to see how it's looking.

You fuel pressure is definitely low, clamping the return usually sees it shoot up to in excess of 50 psi. There never is a stat at the base of the top hose on a P38, it just looks like there could be, stat is always in the globe. Low fuel pressure could cause a problem with a restart but all the other symptoms point to a rich mixture and that would make it lean?

If you can't get the filter undone, just chop the bugger out and use a push on replacement. There on some on the other side that will tell you that is highly dangerous and it will leak out everywhere, catch fire, burn the car, you and anyone within a 10 mile radius. However, a Jag V12 fuel system runs at 40 psi and all the hose connections are push on. Fine as long as you use decent quality fuel hose and proper clips.

Just a thought here, is it possible the low pressure is a leaking injector (or possibly more than one)?

Thanks Gents. Leaky injectors ? Possibly, although it held pressure today for the half hour that I was watching it after I turned off- still haven't discounted the injectors though. Richard, as you suggested I'm now looking at the "chop it out" options, if you know of any filter set up that works, let me know... I was thinking of just cutting the steel line between tank and filter, and then putting a rubber hose over that and running forward to a male fitting (thread to be confirmed, possibly the filter is M16x1.5) to link up to the OE filter? However, then the male nut doesn't turn when I want it to (although the front pipe is perfectly serviceable), not my area of expertise, however. Is there a good website of fuel pipe repair fittings/ options? Am googling as we speak