View Full Version : A/C Compressor Manifold Leak

05-17-2016, 06:51 PM
Just repaired (almost) leak around compressor manifold oring on my 91GT. This was the second one of these that I had done and proceeded as normal - but it leaked immediately. I'm about to tear into it again but thought I would check if there is some trick I missed.

I have done plenty of AC repairs (Have vacuum pump, manifold guages, freon detector and 30lb R134 tank) on toyotas and cadillacs - so this should have been an easy fix.

1) Is there a reason why the system will hold a vacuum overnight - but leak immediately while under pressure?

2) Awhile back, I saw a vid showing a Celica compressor rebuild - where the rebuilder commented that the manifold oring is the common leak point. I was wondering if there is a proper repair procedure for this oring. i.e. OEM vs aftermarket oring, bolt sequence or torque specification?

05-22-2016, 12:59 AM
Vacuum pulls on things, pressure pushes on them. So things can leak with one and not the other.
Also the pressure difference between vacuum and atmosphere is much lower than the pressure difference between pressure and atmosphere, so you're more apt to leak there as well.

R12 is a larger molecule than r134a, so O rings designed for R12 won't necessarily seal r134a.
Other than that, make sure the Oring is undamaged, lube it with AC oil, make sure the sealing surface is clean and undamaged.

05-22-2016, 12:07 PM
If you've done a R12 -> R134a conversion it is possible that your low pressure side valve is the culprit.

I've seen/heard about this on two cars now with a cheap conversion valve pressing juuuuuust ever so slightly on the low pressure valve and causing a leak. If you used a refrigerant with a dye you might be able to chase it down to the culprit area.

05-22-2016, 06:25 PM
System was converted 10 years ago.

The manifold leak apparently started 2-3 years ago. AC would bold a charge for months. I had seen oil on the block next to the compressor and had been thinking it was a motor leak that I could not track down. The system lost all charge quickly some 3 weeks ago. On recharge and leak check, the manifold leak had gotten large enough that my suffer could pick it up. I eventually found that the pressure overload valve on the manifold was leaking. I was able to blow some AC oil through it to clear it out - now no leak. Found a post somewhere else that described a manifold installation procedure that effectively put about 10 lbs torque. (Gradually tighten until the bolts start to feel snug, then turn a 1/4 turn.

Bbbuutt. Put the system back togather and vacuumed it. It held overnight. Started filling it and immediately heard a leak. Tracked it down to an area on the condenser behind the drier. Slight touches on the drier would cause flexing of the condenser and stop the leak. I had seen other posts where components like the evap or condenser would hold vacuum but leak under pressure. Mine must have only leaked when the vehicle was driven add the vibration would strain the condenser enough to open the leak. I'm lucky that it leaked this time right when it filled out as this one had been confusing me for a long time. Having the front end jacked up must have introduced enough flex for it to leak. Normally procedure to find these types of leaks is to pull the component, hook it up to air compressor and submerge in a tub of water - looking for bubbles. Requires access to hose connectors that fit the components. Ugh

I'm guessing the leak on the manifold, and valves (all fixed) where really small after all. Oh well.

Thx all.


Sent from my XT1053 using Tapatalk