Guidance with CO2 Leak Testing

[Alan B]

All,

I've been chasing a leak on my 1993 RX7 with 134a conversion for the past year. The conversion was working well for the prior 15 years. After a fresh charge we saw no film/vapor coming onto the windshield, so our assumption is that the evaporator is OK. We have used dye and a refrigerant sniffer without success. While the sniffer test showed a leak surrounding the service ports, replacing the adapters and core valves showed no improvement.

I recently put together a tank/regulator to pressure test the system with CO2. Testing at 125 PSI showed both sides losing 10 psi over a 35 minute period. I used a commercial liquid to check for leaks at all connections. I also removed the manifold to check the service ports with no visible leaks. I realize I need to still check the condenser core.

My question is what is a safe pressure and for how long can it be applied. I'm concerned that since the low side was never designed to handle the same level of pressure as the high I'm concerned I'll damage the system. A search I did online hasn't given me much confidence to increase the test pressure. Lastly any other techniques you might recommend to better identify the offending component?

Regards,

Alan

[tbirdtbird]

Have you sniffed the evap drain? Looking for film on the windshield is not foolproof. Refrigerant is heavier than air.
Are you using the plastic caps on the schrader connections? They must be in place if you do not want leaks
Have you removed the belts and put a shower cap over the comp for a day and then sniffed inside the cap?

You should not need such high pressures to locate the significant leak you are describing.
Remember you can put a small refrigerant charge (like 20 psi ) into the hi and lo side and then using this as a trace gas, sniff everything

[Alan B]

Thanks for the reply.

Caps are always on and secure. We did not test inside the car with the sniffer. No shower cap test. The tool belongs to a mechanic friend so I'll have to see if I can borrow it. The motor is now out of the car although the AC system is intact. I'd be using 12 oz cans. Rookie question: Will I be able to generate enough pressure without the compressor running?

Regards,

Alan

[tbirdtbird]

Compressor need not be running for the tracer test I described.
Hook up your manifold gauges, and puff in 20 psi refrigerant with both red and blue hand wheels open to the charging hose.
If the compressor is just hanging there now, off the motor, the shower cap test will be easy.

Sniffers are now not that expensive. Suggest you obtain one. They are less than 100.00. Most are sensitive to 1/4 oz leak per year

Do not forget sniffing the evap drain pipe under the car

This statement confuses me
" While the sniffer test showed a leak surrounding the service ports, replacing the adapters and core valves showed no improvement. "
This is why I asked about the plastic port caps being in place. A schrader will typically never accomplish a complete seal.
Be sure the port caps are on, with good rubber gaskets inside, and proceed with everything else. Sniff the port caps once they are on, you should not detect a leak. If you do, the port caps are faulty.
You have to sniff everything and take nothing for granted. This is what the pros do. Be sure to sniff the bottom of the connections, refrigerant is heavier than air.

You have not yet exhausted conventional testing

[Alan B]

Thanks. I will give the sniffer another try. Any inexpensive models you would recommend?

We replaced the R12 to R134 adapters several times. It continued to leak with the caps on. When we did the one sniffer test we assumed it may be the oring around the port. With my recent pressure test I put the bubble liquid around the Schrader valves and it looked fine. With the manifold connected unless my gauges are leaking the valves should be fine although the adapter's connection to the R12 port could still be a fault.

[bohica2xo]

You seem to have a good sized leak. Be sure to check your gauge hoses, adapters, manifold valve stems etc for leaks. Nothing worse than bringing the leak with you to the car...

Regarding the static system test pressure. A 134a vehicle in the sun in Phoenix is well over 160 psi just with the refrigerant charge. 125 psi CO2 will not hurt anything.

Follow Thunderbird's suggestion and charge a couple of ounces of 134a as a tracer gas, and use the sniffer if you have one. You can boost the static system pressure with N2 or CO2 while the 134a is in there if you think the leak requires more pressure to show itself.

[tbirdtbird]

Tim, do you stock sniffers?

[Cusser]

I agree: I would use R134a for leak testing. A combination of visual examination, UV dye, electronic detector, and bubble solution can be needed for difficult leaks.

[Tim]

Tim, do you stock sniffers?

This again comes with knowing most would laugh and say I'll never pay that for a leak detector. I don't offer them anymore. But for the best, look for a Bacharach H-10G. I have a 30-year-old GE, the first version of the H10. Took great care with it, and it will still find leaks no other device will. Tek-Mate is an okay device at a less expensive cost

As for many eBay options. Cusser's soap and water would be a better choice. That's the point I quit addressing, People wonder why they keep losing refrigerant. When their trusty $59.99 leak detector shows no leaks.

[tbirdtbird]

tough to use soap and water on an evap drain tube...

A Florida buddy of mine got one from HF (do not know which one) and believe it or not it has served him well.
I use a TIF (15 yrs old) and I wouldn't want to be without it