😀 I need to re-chassis a 90 Defender, in case you fancy a commission. Body comes off in one piece

I had to do mine when I first got the car. They definitely don't need welding. IIRC I think they are just a bolt-on job - i think the 4 horizontal bolts you see are what holds the crush can on - definite angle grinder job, that one

Even when you know the technique, you're pretty much guaranteed to break something - so buy some spare plastic fitting rings at least

If you have contact cleaner in yiur tool kit...i'd remove the xyz switch and find the little breather Richard mentions, and squirt some contact cleaner in.. And then let it drain/ breathe
Later models of the switch dont have the hole. While youre there take down the zf part no from side of the switch, so you can buy a back up off ebay. If you're stuck i have a spare one with no breather

Mine are tinted...maybe 10 or 20pc

No accounting for taste...probably worth a few quid in parts though

I think the change in note is just when ecm grants the a/c. After grant, if the clutch has not engaged within about two minutes the service book symbol will come on, and at that precise moment you will get another change in tone. So i'd say your prob is between hevac and compressor, or the compressor clutch itself.
Check fuses and relays (esp fuse 42 10A, which is the battery feed to the ecu, without that one you' re going nowhere)
Put a 9v battery across the spades of the compressor...does clutch engage?
Engine off and 9v connected to spades... using a ferrous spanner, does the central spindle/nut on the clutch magnetise? If it does then check the gap on the clutch with a feeler gauge.

If the clutch IS working then pull the hevac ecu and test the BS line on the 16 B connector for ac on. When you turn to Ignition II it will start at 12v, and go to near zero when the ac is requested. If you toggle the ac switch you can watch this go high (off) and low (on).
Test the BG line on the 16 B connector for ecm grant. It will start at 5volts and should drop to zero ( grant) within 30s or so of start up. If not, try raising the revs to about 1500.
And lastly take your 9v battery and put the +ve to the BY pin (no 7 on the 8 W connector) and the negative to a good earth eg back of the cigar lighter...clutch should lock... If not then do a resistance check from pin7 to the +ve at the ac round connector, it should not be more than 10 ohms or so.
If resistance is high, then is your ac charged, if yes then you need to go to the rh kick panel and spray the hell out of positions 1,7,12 with contact cleaner, and if that still doesnt work then test the resistance from pin7 ( lower) to the +ve at the round compressor connector...if you get high resistance there then you prob want to pull the bumper and bridge the trinary switch and/or check the resistance across it. Shd be a few ohms only when closed. If the switch is toast then you can change it without refilling the ac.
Hope it helps

Super reference KCR, thanks for that.

I dug out the HEVAC unit the other day and it turns out I have JFC102550 which as far as I can make out for my car (a 1998, chassis 395xxx) was the "wrong" / later model (it's been there since I bought it 3 years ago). I'm thinking through the implications of that before I make my next move. At the ECU I can see Request (Low/0.7V on a 12v line) and the incoming Grant (Low on a 5v line) but after a couple of minutes the HEVAC gets bored, service symbol comes on, and both lines go high. So far I have not been able to detect any voltage on the black/ yellow pin at the HEVAC ECU. I pulled apart the ECU the other day and there was no visible signs of corrosion or damage. So I'm just trying to work out if something has failed in the ECU, or there's some other condition not satisfied.
Having cleared the distribution flap jam the other day, HEVAC now reports a short to front windscreen (despite me having pulled both the fuses and relays)...I can go chase that next... but I assume that is not linked to a/c not working.

I wonder if either the fusebox or HEVAC ECU are messing with me?

Cleared the distribution doors error (was a stiff mechanism) - now back to just "a/c compressor circuit open" which is not true - as I measured the resistance manually as I mentioned above. I can read on nano that GEMS ECU is receiving an ac request, until after a few minutes of engine running and set at LO, and then it shows the book symbol.

I wonder if this has to do with the ECU determining that the resistance of the circuit is somehow out of spec. I'm still holding out for an a/c systems document, if anyone has ever seen one. I have an EAS one, that is more detailed than rave, i thought there might be a hevac one too

Tnx Richard. It's not the most elegant fix in the world, but it definitely works...amazing

I'm looking for a good way to free up my distrubtion flaps, and have found reference to putting a self tapping screw in the heater box "somewhere", but does anyone have a ref or pic? Seems the definitive guide was on rr.net once upon a time, all i get is dead links, but maybe someone here has actually done it?

If you're not in a rush, ask 9 and get 6-8. Nice looking car...

I measured the resistance from pin 7 at the RH kick panel through to the positive at the round plug for the a/c compressor - which is negligible (it was 28 ohms before i jetted pins1/7/12 with contact cleaner). I no longer get a ghost 12v signal at the condenser (which was there at ignition off, too),. After I cleaned the RH connector the clutch even kicked in for a couple of brief moments, so i think my issue is that one or more conditions for grant and request via pins 1 and 12 are not being satisfied.

I'll keep looking tomorrow, but does anyone have a detailed systems document for the HEVAC ECU - with a list of the conditions required for a/c grant, and request Rave is a bit vague, so I'm looking for the next level of detail My sensors all look to be reading fine. Temp is at Lo. Engine on

In the last couple of weeks (after i tried to calibrate the blend motors, in fact) my distribution flaps have frozen at 83%. I can't really be bothered to free them up at this point, unless we think that that would be enough for the ECU not to apply a/c? So it does have a service book symbol, which registers during the blend flap test cycle at start-up - so most likely the distribution flap being frozenas opposed to anything else.

I just saw today that the GEMS ecu reports that the AMFR correction is at the maximum positive value, so MAF may need a clean, but I assume that's unconnected

Happy days (the first drive, that is...)

pics?

WIfe just got a Tesla (I haven't even been in it yet... too busy fixing the P38) but she tells me that you subscribe to a provider, and pay that way

Tnx Richard, gap's good, measured with a feeler gauge. Will try the RH kick panel connector for sure

V good to know about the HEVAC programming for current draw, so I shan't waste my time on the relay option.

As the temperature is slowly improving my thoughts turn back to a/c. The clutch doesn't kick in, which i think is electrical. The clutch kicks in fine with 12v jumped from the fuse box and also with a 9v battery (who knew ? draws alot of current though). The thing is, if I measure across the terminals of c166 (round, male a/c connector) I see 12 volts, and I can see it at any ignition position - but it's not enough to draw the clutch in... I'm pretty sure that this should read zero until the engine's on at least. So I'm a bit puzzled why I see 12v when I don't expect to, and that the 12v doesn't pull the clutch in... I'm guessing something has failed/ a wire is damaged...next stop trinary switch and pin 7 of the C102 connector, unless anyone has any other thoughts?
I was thinking of knocking up a harness with a relay in it to take a positive signal from the dual pressure switch and then use a fused 12v supply to power the clutch. I think there was an official harness to do this - ie for the earlier P38s where the HVAC ecu is expected to power the clutch directly... does anyone have the official diagram? Not v difficult I suppose - I just need to route a live feed in for the relay, and find a couple of those round a/c electrical connectors

i gave up and ordered an ebay replacement... which according to Sod's Law means i will definitely have fixed the existing one by the time the replacement arrives or that Marty will announce plans to return to the UK 🙄

Fair comment. 3 of the "off"readings are coming off resistors that appear to be set up in parallel and adding the 1/r values gives me the right resistance for that as a circuit.
Leaves me with one now, the circuit diagram would be useful

Thanks for the replies. The holes were cleaned manually by drill bits 0.5mm and less, so I'm hoping the through tracks are no worse than when I started, under 25x magnification there's no obvious issues. Capacitors seem ok. I have tested every resistor multiple times with different meters and have 6 out of 35 that do not match their face value...have you ever seen resistor failure Marty? At least they would be a quick solder job, if I could source replacements. Have you ever managed to locate a circuit diagram?

Richard - nitro moors ? really? won't that cause major damage? I was wondering acetone, but that may eat something too