98 GEMS
So O rings done, I've now got full cabin heat, but the UK weather just went tropical (for a few days, anyway), so now I need the AC... Last summer the AC worked, after I replaced the condenser and added a 2nd hand drier.

This year the AC is not playing. The compressor works, when I give 12v and and earth direct to the connector (C166P) the clutch engages, and the cabin chills as expected. In normal use though nanocom says AC grant "no", but as far as I'm aware the nanocom doesn't also tell me the AC request status - if I could find that in the nanocom menu somewhere it could help with diagnosis, and if I could force it on that would be even better. I also need to disconnect the BECM diagnostic connector as it apparently interferes with the Nanocom's interrogation of the HEVAC ECU - I haven't done that yet - but I don't know if that wd screw up the readings or not.

It crossed my mind that one of the many sensors may be feeding in rubbish data to the HEVAC ECU.
I got ambient 12.7 (fine)
aspirator 18.2 (fine)
evaporator 16.2 (ok, just replaced with a 2nd hand one off a freelander, but same Valeo part number)
heater core temp -0.20 (although it climbs once I've turned up to Hi)

Is the heater core temp way off - seems so. What range should it read in? If that were the case is that going to explain why the AC is not being requested? I recently did the o rings so cd have disturbed something on the sensor

In general any HEVAC faults I have had in the last year are clearable. I have a rare "compressor open circuit" message (but I had that last year too, when the AC was working) and the right blend motor needs fixing at some point. I have a periodic service book come up, but I think it's usually triggered by me asking the jammed blend motor to move (ie by pressing the "wrong" button to adjust the air vent flow)

Seems my first job is to get to air con grant "yes" and then I can start testing stuff... And maybe pressure test the aircon system

Just read Rutland's post about wiggling the heater core sensor, I'll give that a go too

Heater core temp reading -20 may be why it isn't causing the AC grant to change to Yes as it is assuming that with air that cold you don't need the AC on as well. When we did Rick's (rcutler) O rings and checked it before putting everything back together, the heater core was showing -20 but was intermittent. Unplugging it and plugging it back in seemed to sort it and it started reading correctly so a bit of contact cleaner spray may be all that is needed. A couple of hours later we had the very same thing on Rutland's car too after changing his O rings so it does look like it is a connector that doesn't like being disturbed.

My Nano will not connect to the HEVAC first time but will if I come right out and go back in again, unless I've been into the ABS first, then it needs to be shut down and re-booted before it will connect to anything.

The drier should really have been replaced with a new sealed unit really, as it contains desiccant to absorb any moisture that manages to get into the system. If its holding gas okay at the moment, I wouldn't go rushing to change it - but if you do ever have to have it emptied/refilled in the future, it wouldn't be a bad idea to get a new Nissens part on it with new o-rings.

The nanocom doesn't show the request status, but the HEVAC itself should be requesting the compressor when the evaporator temperature is anywhere about about 7 degrees C. It cools the evap down to about 3, then shuts off, and requests again once it reaches 7. Assuming 'AC Off' isn't pressed. Temperature selection shouldn't matter, as the air is cooled and then re-heated by the heater core. When its set to 'hi' on both sides, it might disable the a/c, I haven't tested that.

When ambient is below 4-5c though, it doesn't request it. Consider it this way - if its only cooling the air coming in to between 3 and 7c, there isn't much it can do when its already cold outside. It just risks freezing the evaporator up.

It's entirely possible it is taking the heater core temperature into account too - I haven't come across a core sensor showing an erroneously low reading myself, at least not on a car with working air conditioning.

Do you have a multimeter to hand? You could check the request output and grant input on the HEVAC itself to see what is going on. Probably worth sorting that heater core sensor first though.

"Compressor open circuit" suggests your vehicle is one that drives the compressor clutch directly. This was changed to drive via a relay (and that fault detection/message was programmed out), as the direct feed from the HEVAC controller through all of the connections and loom wasn't always man enough to pull the clutch in. If its working, then its a non-issue currently.

The fault would also come up when the system was low on gas - the safety cut off switch is wired in series to the compressor clutch. When the gas pressure drops too low, it cuts off the feed to the clutch, and thus the directly-driven setup would detect open circuit.

The compressor open circuit fault can be caused by too large an air gap on the compressor clutch. It should be between 18 and 30 thou (the lower the better) or the combined resistance in the various connectors between the HEVAC and the clutch itself means that the reduced voltage isn't enough to pull it in. It gets worse the higher the temperature too, mine would work fine up to an ambient temperature of around 23 degrees but at anything higher it wouldn't. Easy test for that is to wait until it should be pulling in and tap the clutch with a screwdriver handle. If it clicks in and stays in, then you need to take a shim or two out from behind the front of the clutch to get the air gap back in limits. However, that does mean you need to get AC grant to show Yes. The fact that you have had it working last year means you haven't made the mistake my mate did when he first got his P38 and assumed the light on the AC button had to be on for the AC to work and that a button marked AC OFF really did turn the AC off......

From memory the grant signal goes out to the engine ECU and back. This is to ensure the AC clutch is only engaged when the ECU is happy. Be worth buzzing those two lines as well.

Thanks guys, I was being lazy and trying to avoid pulling the cheeks off and HEVAC out for the millionth time :o) but that may be next step... This eve i disconnected the evap sensor - open circuit goes to -40 bridging the connectors defaults to +80 - neither reading is much good to me, so I plugged it back in.

The heater core sensor was wiggled, no change, reads about 0, despite the pipe being too hot to touch. Unplugged and the "reading" is -20, when I bridge the contacts on the plug the default setting goes to +80. I figure that latter reading more closely reflects what the heater core temp would read when the vehicle is running, so I'll drive with the bridge in for a few days and see if the AC settles and kicks in. Sounds like the heater core sensor is mis-reading, but equally I'm not sure that replacing it is going to solve the no air con situation.

At 12V the clutch is definitely working, and immediately chills the cabin, so after I get an air con grant Ican definitely check the voltage drop (if any), I already checked the earth point for the condenser ground wire, and no probs. Last summer (ie after I bought it) the clutch wasn't working, so I removed it, wiped some of the crud off and it started working again, so gap seems ok, if marginal

Tomorrow I'll send 12V to the input side of the trinary switch and confirm that is functioning ok given that there seems to be sufficient pressure in the air con for cooling. And then I'll pull the HEVAC unit and see what I have on the request/ grant wires

Assuming I have air con request, but no grant, then that means some other condition is not being met at the ECM- has anyone stumbled across a list of what those are?

Also, if you haven't already - then it is worth checking the connectors inn the RHF Footwell, behind the kick panel.

These are the links from the fascia (dash) loom and the body loom, and are known for getting moisture in them and corroding. Both the AC request and AC grant lines go through these connectors. They go through C102/202 which is the bigger 18way White connectors pair in there.

AC Grant is Black/Green and on Pin 1, and AC Request is Black/Grey, and is on Pin 12.
They also go through C106/506 between the body loom and the engine loom, and AC Request is on Pin 6, AC Grant is on Pin 7

The feed from the HEVAC to the Trinary switch/AC Compressor also goes through the C102/202 connector aswell - and it a Black/Yellow wire on Pin

Something else that is worth checking, as those connectors are known for causing issues.

Lots of other things go through that footwell connector too. I've cut mine out and soldered and heatshrinked jumper wires in place. The one in the left footwell also suffers and carries the signals between the EAS ECU and the rocker switch and display. It's a fiddly job as there isn't a lot of slack on the cables but definitely worth doing.

Thanks guys. I'll check out those connectors too

Circling back to air con after covering off radius arms and panhard rods, and clearing a short on my electric seats.

I have put a 22k thermistor across the terminals for the heater core sensor, concluded the old one was scrap, replacement on the way. There's a corner of Nano in the GEMS ECU mode, which records the air con request from HEVAC ( I have a yes) and in the outputs section I can in theory force the grant to yes/ no...I just haven't got that function to work for me yet - but if there's something else interfering down-stream then I wouldn't necessarily see the clutch engage - next stop trinary switch and then I guess I'll get a pressure test on the a/c line and refill if necessary

if you know what signal it wants, just tap into the loom and give it grant for testing purposes. It'll likely either just apply ground or 12v to the wire for instance. I'm sure i recall that you can actually just connect the request line to the grant line and that works.

I took the front bumper off today and checked the wiring from pressure switch 1 through to the compressor, which is fine, I also changed the switch itself for a 2nd hand one I had spare - it reads about 1.5 ohms at 6 Bar of pressure. Then I went to 102/202 in the footwell. I removed it, and plugged it back in... clutch started working - hooray I thought.... except then it stopped again. Gives me a target for tomorrow - I may de-pin/ clean it - some green signs on the plug, I must be close : o)

Spent several hours messing around with 102/202, depinning the male upper plug at 1,7 and 12 etc and cycling the ignition. The Grant on #1 was 5v, #7 read 12v, and Request on #12 was zero. I put some silicone seal on top of this plug as water clearly gets in here.
Also unplugged c244 on HEVAC, nada.... Gave up, put it all back together, moved the car to park up for the night and it started working again...all v strange

I also adjusted the tracking in between, but fairly sure that didn't make any difference

12V on pin 7 should pull in the compressor clutch as it goes to the clutch coil. However, it does go via the trinary switch and when I first got my car I had the same problem you had. That turned out to be a bad connection on the switch, unplugged it and gave it a squirt of contact cleaner and it has worked ever since.

If 102/202 is showing signs of going green, chop it out and solder the wires directly. It's fiddly but well worth doing (same goes for 104/204 on the other side).