Question on my air com (1997 P38 4.0SE) - I wonder if anyone else has had this problem? My system has a persistent leak where the pipes join the compressor. There was a really big leak that the garage fixed with new pale green o-rings. It is still leaking however - much less, but still enough for the system to lose pressure and stop working after a week or two. Yes, the bolts on the connector are tight. Photo shows UV tracer. Any thoughts before I drop it back into the garage??
IMG_1321.jpg
 
Wrong thickness o-rings ? Or damage on the pipe fitting ? The single bolts are only there to hold them in place, so only need to be nipped up tight..
 
Wrong thickness o-rings ? Or damage on the pipe fitting ? The single bolts are only there to hold them in place, so only need to be nipped up tight..
Thanks, that is good info. Ref the o-ring size: I got some o-rings to take into the garage with the P38, but did not unpack them to check size in advance. Perhaps there were two different pipe sizes in early and later P38s - maybe different compressors? The garage said the o-rings I supplied were the wrong size and they had fitted the correct ones. Just measuring up now, the pipes are 12 and 18mm OD, and the o-rings I got were 9.5mm and 14mm ID (and 2.5mm thickness). So I guess they are too small for this model. The garage can have another go at sealing that joint.....but any further info gratefully received....
 
Online catalogs seem to only show the o-ring diameter, but not the thickness. Anyway I suspect stretching the 9.5 to 12mm will probably cause the thickness to shrink ? I'm not an expert in rubber rings though !!
 
Online catalogs seem to only show the o-ring diameter, but not the thickness. Anyway I suspect stretching the 9.5 to 12mm will probably cause the thickness to shrink ? I'm not an expert in rubber rings though !!
The garage didn't fit those rings - they told me they were the wrong size. I assume that JLR must have changed the compressor for later P38s, like 2000 onwards or something, and the new compressor pipes were smaller. I got the rings from Rimmer Bros so quite likely they were wrong. The leak is now very small but obviously still big enough to release the pressure over a few weeks and stop the aircon from working. Presumably the nasty compressant fluid will attack any sealant added to the joint, so you can't just bung some instant gasket in there when reassembling. Ah well, back into the garage she goes...
 
Sounds like whatever the garage used are wrong size as well ? One other small possibility is slight corrosion on either the pipe groove where the ring sits, or in the compressor orifice causing issues with the sealing ?
 
Sounds like whatever the garage used are wrong size as well ? One other small possibility is slight corrosion on either the pipe groove where the ring sits, or in the compressor orifice causing issues with the sealing ?
Interesting comment - thanks. Annoying not to be able to take a look myself, but refrigerant is not something to play around with...
Maybe the rings I got ARE the right size, given that you say they fit in a groove in the pipe and a recess in the compressor.
The dimensions of the rings that I provided, which were not used are:
Thinner pipe OD 12mm, o-ring ID 9.5mm, o-ring OD 14.5mm
Thicker pipe OD 18mm, o-ring ID 14mm, o-ring OD 19.5mm
Given that o-rings should be slightly smaller than the pipe on which they fit, those dimensions look to me as if these rings could be the right ones.
I'll discuss it with the garage but I think you might well have identified the problem for me - I'll report back in case the info is useful to anyone else.
 
If they properly extract the gas, they should be able to check the seals and either Vacuum test, or pressure with nitrogen. Then they can re-use the gas, so you don't need to pay for new gas !!
 
If they properly extract the gas, they should be able to check the seals and either Vacuum test, or pressure with nitrogen. Then they can re-use the gas, so you don't need to pay for new gas !!
He said it had gone down below the usable pressure for the car,that's why I was only getting a small amount of cold air,that could be why it was intermittent on the cooling side.
 
He said it had gone down below the usable pressure for the car,that's why I was only getting a small amount of cold air,that could be why it was intermittent on the cooling side.
Classic pressure loss - standard problem on all aircon systems. But it is not a variable, the pressure is either high enough or not. If not high enough, the pressure sensor simply stops the electric clutch from engaging and spinning the compressor - easy to check. So it either works or it doesn't.
 
Interesting comment - thanks. Annoying not to be able to take a look myself, but refrigerant is not something to play around with...
Maybe the rings I got ARE the right size, given that you say they fit in a groove in the pipe and a recess in the compressor.
The dimensions of the rings that I provided, which were not used are:
Thinner pipe OD 12mm, o-ring ID 9.5mm, o-ring OD 14.5mm
Thicker pipe OD 18mm, o-ring ID 14mm, o-ring OD 19.5mm
Given that o-rings should be slightly smaller than the pipe on which they fit, those dimensions look to me as if these rings could be the right ones.
I'll discuss it with the garage but I think you might well have identified the problem for me - I'll report back in case the info is useful to anyone else.

OK - I now think I understand what the problem is. As per attached pic (eBay manky old spares shown!) the GEMS aircon pump has larger o-rings where the (parallel) pipe clamps fit, whereas the Thor engines have pumps where the o-rings are on the ends of the (opposed layout) pipes and therefore much smaller.
So all I need to do now is find the correct larger o-rings.
If anyone wants a pair of Thor o-rings as per above sizing, message me and I'll pop them in the post to you. Pity to bin them.

GEMS and THOR aircon pumps.png
 
More research on the P38 GEMS a/c pump o-rings. My pump is a Sandon TRS105 / 3204.
O-rings don't seem to be readily available, but are listed on a German supplier site (https://st-ersatzteile.de/o-rings-suction-side.html?language=en)
OD 25.53 mm, ID 20.29 mm, thickness 2.62 mm, which sounds about right.
That German supplier price is €8.88 plus VAT plus shipping but that is for 20..!
I think I'll try to get generic a/c proof o-rings (HNBR, hydrogenated nitrile rubber) at that size.
Failing that, perhaps I'll buy 20 from the Germans and then sell the residual 18 on eBay...
 
Last edited:
Further checking shows that there are two different sizes of o-rings on the GEMS compressor: low pressure side (in) is OD 25.53 / ID 20.29 / id 2.62, high pressure side (out) is OD 22.36 / ID 17.12 / id 2.62. I am going to order, and after testing them, I'll sell the residual 19 of each on eBay and hope to get my money back.
https://st-ersatzteile.de/o-rings-suction-side.html?language=en
https://st-ersatzteile.de/0-rings-discharge.html
Do you have a tig welder?
 
The regas has worked ,for now anyway.
Nice and cold:).
Well, barely a month and it's packed up:mad::(. Had do go out today ,35 deg, kept burning my hand on the steering wheel. Waaaay to hot to investigate anything. I 'think' last time I was out, last Friday, it possibly wasn't blowing as cold as normal so I guess I've got a leak:rolleyes:.
 
Well, barely a month and it's packed up:mad::(. Had do go out today ,35 deg, kept burning my hand on the steering wheel. Waaaay to hot to investigate anything. I 'think' last time I was out, last Friday, it possibly wasn't blowing as cold as normal so I guess I've got a leak:rolleyes:.
That's what make me laugh as there supposed to check for leaks first and if it's all ok then charge the system.
 
That's what make me laugh as there supposed to check for leaks first and if it's all ok then charge the system.
Unfortunately, VAC testing does not always show very small leaks, nor does pressure testing with Nitrogen. I have had exactly the same on mine, no leaks on test done in my mates garage. One month later no aircon but a tell tale dye stain from the top offside corner of the condenser.
 

Similar threads