No, the knurled thumbscrew is the bit you screw in to close off the gas outlet, the bit you remove is the bit sticking out facing towards you where the solenoid coil was. You tried removing the manual valve so would have emptied the tank if you'd left it long enough. Bear in mind that 1 litre of liquid becomes 270 litres of vapour so you would have filled your farm.....

Had you asked in another 6 months time we'd all have known how you spent lockdown.....

Can't help, my daughter is older than a P38 so never had need to fit one.

What sensor on top of the reducer? Can't see one in the pictures, only the inlet solenoid and filter housing. Temp sensor may be screwed into one of the bolt holes holding the two parts together, small, brass with a couple of thin, usually orange, wires coming from it.

and the gasket sits on top of the seal but under the metal thing. Your picture seems to show the gasket over the top.

The triangular bit on the end of the seal should push down between the top of the block and the head. From your photo it looks like it isn't seated properly. I put the ends in first and push them down then push the rest into place.

Resistances sound about right. Try putting 12V on them and see if they go clack.

Profess......

HaHaHaHaHa, it's never worked properly then.......

It's a Dream XXI-S system so the engine running is detected by ignition pulses, the MAP sensor does what it says on the tin, it's a Manifold Absolute Pressure sensor so gas pressure is relative to manifold pressure so will vary to keep it the same with respect to manifold pressure (or vacuum). MAP sensor fault won't cause no changeover but it will cause poor running.

So they can use the same hub casting on both sides. Depending on what side it is going on depends on which hole gets blanked off.

Is it rust or oil from the half shaft seal?

If that solenoid is OK, do the same with the front one. If they both test OK, is the system actually trying to change over? Common problem is the wire breaks off the temperature sensor on the reducer so the system thinks it isn't hot enough to change over. If it attempts to change over then immediately changes back, that is likely to be a stuck solenoid (no gas pressure being detected), if it doesn't attempt to change over then I'd be looking at the temperature sensor.

Fuse 41, 20A, supplies power to the contacts of the relay.

The 'relay' is the solenoid coil that operates the shut off valve on the tank. Check the resistance it should be around 11 Ohms and look for any cracks in the plastic body. If you put 12V onto it, you should hear the solenoid make a very audible clack as it opens. If you remove the small bolt in the centre of it, the coil just lifts off laving the valve body behind so no gas will leak out, but you'll be able to test and inspect it much better. Just behind it is a knurled knob which is a manual shut off valve. If you screw that in fully clockwise, that will stop the contents of the tank leaking out when you do the next step. At the base of the brass spindle that the coil sits on is a hex, usually 17 or 19mm. With a deep socket, slacken that off, some gas will escape (the gas in the pipe between the tank and the front shut off solenoid) so let it bleed out slowly. Once it stops, unscrew the spindle and take it out. There's a spring loaded plunger inside it that can get gummed up, clean it and put it back together.

On a pre-2000 model and SRS fault will still be there until it is cleared with diagnostics even after the actual fault is cleared. It's only on the later ones where the light goes out after you've fixed the fault without it needing to be cleared.

Horn button applies a ground via the rotary coupler. One side of it will have a ground on it while the other side should go to the relay so the ground causes it to operate.

I thought it would only go in one way. Have a close look at it and see if it looks like it has been butchered.

From the picture in RAVE, it looks like it is angled upwards?

Good call on only one fan running. That will also give the book and naff all air as what is sucked by one working fan will just go straight out the other side rather than through the heater box. If you take both pollen filters out, you can see them down the hole and see if they are both running.

You've put extended shocks on with standard air springs? Don't jack it up on the chassis or try any offroading, you will pull the air springs apart. It's the shocks that limit the amount of axle articulation you have.

Gorilla Glue is good for some things but it sets brittle so can break with movement and also expands as it sets.

Noit just the blend motors, the HEVAC will not try to use anything that it detects a fault on so if it detects a fault at self test when the ignition is first turned on, if a blend motor is detected as faulty it won't try to operate it. Same goes for the blowers, if it detects a fault with either blower, it doesn't try to use it. Ideally you need diagnostics to confirm what the fault is and which blend motor, only the faulty one will not be operating. You should be able to adjust the temperature on one side or the other. Unplugging the blend motors from the HEVAC, switching the ignition on for 10 seconds, then off and plug them back in sometimes resets the fault but not always.

The ducting is plastic with foam seals which rot away so most of the blower output can get out before it even gets as far as the heater box, using duct tape for what it was originally designed for can be used successfully. Blend motors can be changed by going in through the hole after you've removed the instrument cluster or by dropping the glovebox out, no need to take the whole dash out. First you need to identify which blend motor is at fault and what the fault is.

There never was sealer, the rubber seal was stuck to the bodywork with a strip of double sided tape. Unfortunately, with age the rubber shrinks slightly and pulls to tape away. I've stuck it back on both my cars with Evo Stik contact adhesive.