Broken A/C troubleshooting
The pump was running for about 30 minutes. It has been sitting with the pump off for over an hour. I’m planning to do the actual recharge this Wednesday when my low side fitting gets in, and I’m wondering if the system is dry enough to leave it how it is until then
‘83 Platinum N/A 944
-
Tom
- Site Admin
- Posts: 8971
- Joined: Fri Jun 25, 2021 2:04 pm
- Location: Silicon Valley, CA
- Has thanked: 943 times
- Been thanked: 4032 times
- Contact:
Yeah, probably ok. Keep it in vacuum state and monitor the gauge to see if it's leaking. I would pull a vacuum with the pump on for 45-60 minutes just before filling the system, after the low side adapter is on.Poorsche44 wrote: Sun Jun 22, 2025 1:06 pm The pump was running for about 30 minutes. It has been sitting with the pump off for over an hour. I’m planning to do the actual recharge this Wednesday when my low side fitting gets in, and I’m wondering if the system is dry enough to leave it how it is until then
I’m currently working on it at a friends driveway and need to bring it back home so I won’t be able to keep the gauge on to monitor, but I will pull vacuum for At least an hour when the fitting is installed. Thanks for all your help answering my silly questions, I’ve just never dealt with automotive ac before
‘83 Platinum N/A 944
Just received my 90 degree fitting in the mail today and I noticed that the o ring on the inside is black. Does that mean it’s not neoprene/r134 compatible?
- Add Pictures/Files
-
- IMG_6963.jpeg (3.89 MiB) Viewed 927 times
‘83 Platinum N/A 944
-
Tom
- Site Admin
- Posts: 8971
- Joined: Fri Jun 25, 2021 2:04 pm
- Location: Silicon Valley, CA
- Has thanked: 943 times
- Been thanked: 4032 times
- Contact:
It's a virtual certainty that the o-ring inside an r134a adapter is compatible with r134a. Neoprene o-rings do come in black, especially when part of an assembly like this. They just make consumer o-rings in green so that they can be identified as r134a compatible, but color-coding isn't really necessary in a manufacturing environment. I think you're good to go.
-
dr bob
- Moderator
- Posts: 632
- Joined: Thu Jul 08, 2021 9:30 pm
- Location: Central Oregon
- Has thanked: 255 times
- Been thanked: 251 times
Reading the 'vacuuming' part, I'll share some hints --
-- While convention seems to say that running the pump for 45 mins is sufficient, I think that short time is a compromise made by shops that don't want customer cars sitting in their work bays for longer.
-- I managed to scrounge a lab-quality many-stages piston vacuum pump that's capable of deeper vacuum. Oiled piston HF 2-stage pumps are certainly better than the ones they sold originally. I have yet to use mine here "at altitude" (about 4k' here) so can't say what the gauge readings end up at. I did end up buying an electronic gauge (Robinaire) that reads to microns/torr for the last bits of the process.
-- It took me two tries to get my R-134a conversion to work well. I didn't realize there are o-rings where the manifold plate fits to the compressor, and it took only a couple days for them to fail in the polyolester oil for the R-134a.
-- Each time I pulled vacuum, I left the pump running and connected for at least 24 hours before adding oil and refrigerant. Ambient temps were in the low 80's daytime high, and I wanted to make sure there was no moisture at all in the system. Depending on your ambient conditions and how much moisture made its way into the open system, a full dry-out with vacuum can take from hours to days.
-- Moisture still in the system will limit the vacuum you can pull. You can compare the deadhead pump suction pressure to your actual pressure, and see if you have residual moisture (or a leak...). Detecting the difference takes enough pump-down time to evaporate and extract the moisture.
-- Be sure to use R-134a compatible seals and oil on your gauge manifold and hose connections.
-- Be sure to lubricate every seal and connection point with refrigerant oil before you assemble.
-- I seem to find more connections leaking from 'too tight' than from too loose on others' DIY efforts. If you sniff a connection and detect a leak, loosen it slightly before re-torqueing. Yes, the fittings do have torque specs. Follow them religiously, and you won't be flattening or tearing o-rings with 'too tight' or even disturbing a connection after it's been snugged up.
-- Be sure to remove the old Shrader valves from the original charge ports before installing the R-134a adapters. Yes, there are pin depressors in better adapter fittings, but I'm maybe 50% in getting them to work correctly.
-- Be sure to upgrade all the Shrader valves in the system to R-134a-compaible parts. Besides the obvious ones where the hoses connect, they are under the pressure sensors and safety switch, plus there were two on that compressor manifold plate. Use o-ringed port cover caps too, with HNBR o-rings.
-- I've been using a 4-valve charging manifold for a couple years now, as it allows me to evacuate all the hoses including the one to the supply cannister, while the pump is connected. Saves juggling hoses full of air, or trying to do a liquid hose purge to push air from the hoses. Prior to that, I evacuated via the Shrader port under the safety switch. OK but not perfect.
-- I deliberately under-charge the system slightly with the scale, then sneak up on perfect using a couple thermometers, the gauges, and the sight glass. I look for maybe 20 PSIG suction pressure with engine at 2500, fan on speed 2, at 80º ambient. With the freeze protection disabled, center vent air temps are mid-teens while cruising in 90+ SoCal low-RH conditions. The new Pacific Northwest description for that is Pretty Falcon Cold. The goal is to have barely-liquid flow to the expansion valve, so the evaporator sees only liquid but the pressure is as low as possible.
In industrial environments like steam turbine condensers in power plants, we place a helium detector at the vacuum pump outlet. Then go put helium at every possible connection, gauge, test port, water drain, all the way to any fill valves and vacuum-breakers. Car systems are compact so easy to push helium at everything, point by point.
-- While convention seems to say that running the pump for 45 mins is sufficient, I think that short time is a compromise made by shops that don't want customer cars sitting in their work bays for longer.
-- I managed to scrounge a lab-quality many-stages piston vacuum pump that's capable of deeper vacuum. Oiled piston HF 2-stage pumps are certainly better than the ones they sold originally. I have yet to use mine here "at altitude" (about 4k' here) so can't say what the gauge readings end up at. I did end up buying an electronic gauge (Robinaire) that reads to microns/torr for the last bits of the process.
-- It took me two tries to get my R-134a conversion to work well. I didn't realize there are o-rings where the manifold plate fits to the compressor, and it took only a couple days for them to fail in the polyolester oil for the R-134a.
-- Each time I pulled vacuum, I left the pump running and connected for at least 24 hours before adding oil and refrigerant. Ambient temps were in the low 80's daytime high, and I wanted to make sure there was no moisture at all in the system. Depending on your ambient conditions and how much moisture made its way into the open system, a full dry-out with vacuum can take from hours to days.
-- Moisture still in the system will limit the vacuum you can pull. You can compare the deadhead pump suction pressure to your actual pressure, and see if you have residual moisture (or a leak...). Detecting the difference takes enough pump-down time to evaporate and extract the moisture.
-- Be sure to use R-134a compatible seals and oil on your gauge manifold and hose connections.
-- Be sure to lubricate every seal and connection point with refrigerant oil before you assemble.
-- I seem to find more connections leaking from 'too tight' than from too loose on others' DIY efforts. If you sniff a connection and detect a leak, loosen it slightly before re-torqueing. Yes, the fittings do have torque specs. Follow them religiously, and you won't be flattening or tearing o-rings with 'too tight' or even disturbing a connection after it's been snugged up.
-- Be sure to remove the old Shrader valves from the original charge ports before installing the R-134a adapters. Yes, there are pin depressors in better adapter fittings, but I'm maybe 50% in getting them to work correctly.
-- Be sure to upgrade all the Shrader valves in the system to R-134a-compaible parts. Besides the obvious ones where the hoses connect, they are under the pressure sensors and safety switch, plus there were two on that compressor manifold plate. Use o-ringed port cover caps too, with HNBR o-rings.
-- I've been using a 4-valve charging manifold for a couple years now, as it allows me to evacuate all the hoses including the one to the supply cannister, while the pump is connected. Saves juggling hoses full of air, or trying to do a liquid hose purge to push air from the hoses. Prior to that, I evacuated via the Shrader port under the safety switch. OK but not perfect.
-- I deliberately under-charge the system slightly with the scale, then sneak up on perfect using a couple thermometers, the gauges, and the sight glass. I look for maybe 20 PSIG suction pressure with engine at 2500, fan on speed 2, at 80º ambient. With the freeze protection disabled, center vent air temps are mid-teens while cruising in 90+ SoCal low-RH conditions. The new Pacific Northwest description for that is Pretty Falcon Cold. The goal is to have barely-liquid flow to the expansion valve, so the evaporator sees only liquid but the pressure is as low as possible.
In industrial environments like steam turbine condensers in power plants, we place a helium detector at the vacuum pump outlet. Then go put helium at every possible connection, gauge, test port, water drain, all the way to any fill valves and vacuum-breakers. Car systems are compact so easy to push helium at everything, point by point.
dr bob
1989 928 S4, black with cashmere/black inside
SoCal 928 Group Cofounder
928 Owner's Club Charter Member
Former Ex Bend Yacht Club Commodore Emeritus
Free Advice and Commentary. Use At Your Own Risk!
1989 928 S4, black with cashmere/black inside
SoCal 928 Group Cofounder
928 Owner's Club Charter Member
Former Ex Bend Yacht Club Commodore Emeritus
Free Advice and Commentary. Use At Your Own Risk!
Wow Dr Bob, thank you for all that info on the system! I m currently performing the final vacuum on the system before fill up right now, and things are looking good. I haven’t shut the pump off for the leakdown test yet but the pump is pulling more vacuum than I’d ever seen before. Around 26 in Hg and like tom said 25.5 seems to be the highest I can achieve at my elevation. I’ll keep you guys posted as things go on
‘83 Platinum N/A 944
So I tried to fill the A/C but it did not go as I planned. The ac compressor kept shutting on and off even after I had put two cans (24 oz) of refrigerant in the system. That would mean that’s the low pressure switch right? I tried bypassing it but It either didn’t work or I jumped something else entirely. I had to undo the system and put it away as I needed to move the car from the garage I was working in.
‘83 Platinum N/A 944
If you had a leak, it would shut off, but "on and off" makes me suspect a flaky temperature sensor, expansion valve, or relay/switch. First, go over the manifold gauge readings, see if the system is holding the charge over time, before you begin troubleshooting electrical issues. Once you know the lines, etc. are properly under pressure, you can go from there.Poorsche44 wrote: Fri Jun 27, 2025 10:40 pm So I tried to fill the A/C but it did not go as I planned. The ac compressor kept shutting on and off even after I had put two cans (24 oz) of refrigerant in the system. That would mean that’s the low pressure switch right? I tried bypassing it but It either didn’t work or I jumped something else entirely. I had to undo the system and put it away as I needed to move the car from the garage I was working in.
Jer