A/C Leak Detection method recommendations?
 

Background: Total system replaced 4/08. New barrier hoses, new compressor, new evap, new consensers (4 including Duehl Kuehl), Pro Cooler, high low cutoff switch.....i.e. every component has been replaced.

Although I'm not running the A/C this time of year I have noticed that last month or two that I could not get the A/C compressor to kick on and thought maybe it was the cold/ low pressure switch that would not allow it to kick on. Today I bypassed the H/L pressure switch and by doing so was able to get the compressor to cycle on and off regularly. Ran system and it did not cool at all so I must have a leak and am low on R-134a.

Recommendations on leak tracing? I don't have a sniffer but may need to get one.
I need to get the car up on a rack to visually inspect all the hoses, but assuming it passes the visual inspection, what is the best way to trace a leak?

I have not hooked up the gauges yet - What's the point? I already know I'm low and the system is not cooling.

Suggestions?

Rubber hoses leak....period. A good sniffer is going to set you back several hundred dollars. Some people put dye in the systems, but I do not like that.

My solution. Keep adding. R-134a is cheap.

Let me know when you want to run over to get the car up in air.

Look for a wet spot. Usually at a connection or seal. If it's wet with oil; it's leaking.

uv light and yellow glasses

I'm not wild about the idea of using dyes as I have read that they tend to contaminate A/C systems.

Look at it this way....you almost made it two years. You can do another vacuum and recharge and be good for another two years.

put gages on to confirm you are low on 134. if it is low, check all the connections first look for the wet spot. make sure they are tight, if the 134 is all gone, go ahead and replace all the o rings. ,

you can get a adapter for a propane torch that has a rubber hose attached to it and you trace the a/c lines. If there is a leak, the flame changes color. There cheap and if you have a torch (burner if your British) then you just need a adapter. Check at any a/c repair shop (home, industral, commercal, or automotive).

Yet some manufactures have it in initially?
I've heard a lot of things too. ;)

I haven't heard of an issue w/ dyes but with some of the stop-leak type stuff included w/ some refill cans. That said, my experience with using a dye was a pain cause it was so messy getting it in the system. Meaning, if there was a leak around the compressor and valve, I'd never have found it. The first place I would look would be the valves, especially if you used the adapter types on top of existing Schrader valves. That's where my leak was. Used some soapy water. Remember the caps are integral to a complete seal. Good luck.

I used the dye and it shows leaks well. I also have a strong black light flashlight to loan if you need it. That is a disappointment to only hold the charge a couple of years with all new components. I would suspect the Nippondenso compressor first. Do you have an obvious oily line on the decklid above the compressor? My system showed a bad connection at the dryer and another at a service port where I installed the trinary switch. Two years is not that unreasonable to do a vacuum and recharge since you have all the equipment. That would be the case if it is a slow leak that has been happening for 2 years. On the other hand if it is a fast leak that has happened over a week, that would be a problem.

I used the dye also and found the leak was the compressor, coming from the actual body of the compressor where it was bolted together. Hopefully Sam's will sell the 12 pack of the 134 for 35$ again and I will be back in business. The dye actually worked well and could not find any other leaks except that one and it is a slow leak. Good luck!

Hi Mark,
I'm the guy with the '80SC 3.6 gray metallic thats close to your car color. We've chatted several times at the Avenues Sunday cruise in.
If you haven't found your leak by the next Avenues meet up, I can bring my "sniffer" for you use to try find your leak. Let me know.
Grant

With slow leak like that there's no reason to re-evacuate it. It's not like air and moisture is gonna leak into it, the refrigerant is just leaking out untill pressure gets low enough and it stops so all you have to do is recharge it.

You only have to evacuate it if the system is opened up to atmosphere and then air and humidity gets in and replaces the refrigerant.

Oil will settle to the bottom areas of the hoses and heat exchangers from gravity so if the slow leak is near the top it won't be wet from oil.

I re did my entire AC system with all new rennaire and zims updated parts and barrier hoses that came to around 2 grand not including my labor.
It leaks down real slowly too, probably an oring somewhere fun to get at like in the smuggler box which is hard to open up with a strut bar in place..

I buy R134a at around $100 for a 30 lb botttle to maintain both my cars here in hot florida so it's really easy to just recharge it once or twice a year.

Mark;
Same issue with my Rennaire/Griffiths. Laeked down last winter and was fine when I stored the car. HVAC buddy & I pushed 170# of Nitrogen & pulled a vacuum for almost 2 hours. No leak at 85 ambient. Dyed it and recharged it and I WILL find that damn leak now taht winter is here. I suspect an O ring or fitting. $2000.00 & problems! ARRRGGGGHHHH!!!!!!

All good comments. Thanks guys!

I do unfortunatley have a little tell tale oil streak coming from the compressor clutch that I noticed shortly after the install. It is a Nippondenso Factory Rebuilt unit......I think when head temps get really high in peak season this happens to most compressors - at least the Nippo's. Griffiths states that Nippos are larger capacity units vs the Sanden replacements and are slightly better for our cars if they are working properly (Charlie - sorry if I'm miss-interpreting your comment but that was my understanding)....anyway that's why I stayed with the Nippo.

May take you up on your offer. I'll PM you the next time I plan on going. Ever get your hose connections re-done?

Good to hear from several that slow leak downs with new systems are not uncommon. Since I've got all the equip, I'd perfer to evac and recharge so that I know how much 134a I've got in there. After several recharges, I've calibrated my "optimum charge" using a 30lb jug and a scale at 39 oz. If I just top off/ recharge, I'll have no idea where I'm at.

Let me know what you find? I plan on sniffing mine when I get it charged back up.

"Good to hear from several that slow leak downs with new systems are not uncommon. Since I've got all the equip, I'd perfer to evac and recharge so that I know how much 134a I've got in there. After several recharges, I've calibrated my "optimum charge" using a 30lb jug and a scale at 39 oz. If I just top off/ recharge, I'll have no idea where I'm at."

With this formula from "KUEHL" I know exactly where I'm at and I can just top it off perfectly once or twice a year in a few minutes without wasting any R134.

These pressures are at idle speed with the lid down or if you have a fan(s) on the condensor(s) the lid can be up.

80-85F = ambient times 2.4 X 1.05 or the + 5% rule
86-90F = (ambient times 2.4) X 1.1 or the +10% rule
91-95F = ambient times 2.4 X 1.13 or the +13% rule
96-100F = ambient times 2.4 X 1.15 or the +15% rule
101-105 = ambient times 2.4 X 1.17 or the +17% rule
106-110 = ambient times 2.4 X 1.20 or the + 20% rule

"Here is how you apply the rules:
Say you have charged the system with a known quantity of R134a refrigerant (grams or ounces or whatever) and the ambient temperature is 95F. Take 95 and multiply it by 2.4 = 228 psi and then multiply that by 1.13 = 258 psi. Now consider if you really had this pressure well that would be in a perfect world. Depending upon whether you parked on black asphalt or if the expansion valve was opening or closing this mean target of 258 might swing up to 270 or so and frankly that is pretty good in a 911 with R134a at 95F. Or, if you are testing with the deck lid down and your services hoses are holding the deck lid up an inch or two that would reduce the effective air flow across the deck lid condenser and drive pressures upward. "

If it is flat (zero pressure) just find a regulator and charge the system with nitrogen as mentioned before. There should be enough residual gas in the system to provide a "trace" of refrigerant after nitrogenizing. Use the electronic leak detector and see what you can find.
You can always use a soap solution to leak check fittings while pressurized too...

I usually top the charge up once or maybe twice before doing a full evac and pull a fresh vacuum. I just don't know how much water is in there, so I don't really like to just top it up, a personal choice/ quirk/ eccentricity or whatever. Besides, JFairman convinced me to buy the Robinair 6CFM 2-stage vacuum pump a few years ago, and I love using that thing.

Mark,
Here is the typical procedure:

1) Evacuate/Recover (cough) whatever residual refrigerant is in the system.
2) For now just pull a 'standard' vacuum on the system (not need to do the Charlie 3 step) and charge the system; if you have not already added dye, you can inject after you have pulled the vacuum.
4) Run the system on a warm day, say 80F + to get the system pressures up, take it for a spin.
5) Take out your dye light and your electronic leak detector and check all the normal suspects following the flow with the system running at idle:
A) Compressor: nose seal, case seals, charge port adapters, manifolds on the compressor, hose connections at the manifolds and the fferrule crimps on the hose ends (this on all future mentioned hoses; not a common leak area but check). Inspect the charge port adapter valves to insure one is not hung up.
B) Deck lid condenser hose connections, the manifolds attached to the condenser.
Inspect the deck lid condenser for signs of a 'ding' in the fins; usual suspects that cause the ding.
C) Any auxillary condenser (aka Kuehl or other copy cat ... cough) hose connections and the manifolds attached to the condenser and fin dings like before.
D) Drier inlet and outlet hose connections (the drier orings are suspect as the fittings tend to be over torqued).
E) Front condenser and its fittings.
F) Expansion valve hose fitting, the base of the valve where it sits on the evap, the evap outlet hose... and your done.

I'm leaning towards your observation noted on the compressor nose seal, very common.

The new barrier hose set will not permeate enough refrigerant as you have noted you lost.

Call, email or send a carrier pigeon if you need a hand. A fresh six of Becks if the bird is worthy.

Griff

Well, my leak turned out to be a easy find. I topped off the charge this weekend just to get the A/C running and observe how long it held the charge while coninuing to look for the leak.

This AM I noticed oil on the floor under the engine oil cooler - clear A/C oil not engine oil. I traced the leak to one of the ferrule crimp connections at the high/low pressure switch which is located on my high pressure line running from the compressor to the rear deck condenser. Oil dripped down onto the tin exiting at the rear tin seam (lower bolt in pic) onto the oil cooler S hose then onto the garage floor.

I also see a fresh oil streak on the deck lid indicating that the compressor nose seal is still leaking. Two items need fixed but at least now I know where my leaks are. New nose seal kit and hose are on the way - thanks Griff!

A/C oil dripping from lower tin bolt onto S hose clamp:
http://forums.pelicanparts.com/uploa...1264547035.jpg

Leak source - left ferrule crimp here:
http://forums.pelicanparts.com/uploa...1264547103.jpg

Thanks to Griff, I have a replacement hose assembly and a Denso shaft seal kit in hand. I have been doing some research on the shaft seal replacement job. There is a tech article on the process here on Pelican but I found another excellent tech article here:

1988 911 Nippondenso A/C Compressor Rebuild

and another here:

Rebuild method for Nippondenso 10P15C A/C compressor

No problem, we covered your hose under the good guy warranty naturally. Don't know why crimp failed, after some 10,000 crimps I guess you gotta have one bugger in the lot. Call me is you need help with the seals. Enjoy.

Griff

Using dyes to locate a refrig. leak is a popular method. However, in newer vehicles, a smaller, precise amount of refrigerant is called for in most systems for max. cooling efficiency. Introducing dyes and stop leaks into these systems displaces the amount of refrig. needed and compromises what the system can do. Besides, how do you know how much leak dye was inserted by the PO?

Besides, flourescent leak dyes are not popular on passenger clothes and interiors. As others have suggested, using an electronic sniffer is preferred.

Sherwood

Hmmmm,

I don't agree with some of what you wrote here Sherwood.

Dyes and stop leaks are two different worlds.

1) Refrigerant dyes are typically introduced into the system as 'oil', either Ester (used for R12 or R134a) , PAG (R134a exclusively) or Mineral (R12 exclusively). The 'dye' is contained in the oil. You can add dye into the system a few ways: (a) after pulling a vacuum on the system you can use dye syringe, typically 1 to 2 ounces is plenty on a 911 or 930, and this amount becomes part of the total oil requirement, (b) using dyes already included in small cans of refrigerant or (c) when you add oil to the compressor before you install it on the vehicle.

If you have a complete 'dye kit system' it should include a cleaning agent to remove the dye from ac system and near by components: I won't say it will or will not come of you clothing or the interior, however I don't put my Sunday best clothes on normally when we do this work in our shop, nor have I ever encountered 'dye' in the interior of any car because the nearest component to the interior is the evaporator and that is typically enclosed
behind firewall or part of the forward dash and evaporator boxes have drains that lead outside the vehicle rather than inside of it.

The reason a 'dye method' is popular is because the auto, commercial and HVAC industry uses it all the time. Dye's are also used to find engine coolant and oil leaks as well, however the 'dye' is formulated for each particular use (you don't use engine coolant dye in an air conditioning system per se). The benefit of using the dye-method is that 'sniffers' or 'electronic' leak detectors do not always detect the leak at the moment. Electronic leak detectors are sensitive to rates of xx parts per million. So if you have a system that your checking and say for the moment the high side pressure is only 180 psi and the 'hole' or permeation is rather small, the electronic detector may not pick up the leak. However if the the hole is larger or the system pressure is up to say 250 psi or more it is more likely to pick up the leak.

The benefit of using a dye is that over time it is possible that the refrigerant oil, which contains the dye and is being circulated through the system by the refrigerant when the system is operating, 'may or can' come through the hole and you could have the opportunity to locate it later.

The normal procedure therefore is to use both methods and it usually works like this: you check with the electronic detector first, if nothing is detected you introduce the dye, the vehicle owner uses the ac system for a week or so and revisits the shop and they check the system with a dye light. At least that is what we have done for many years as well as our peers this industry.

How do you know how much dye was inserted by the PO? Well, if they "did it right" they would have documented on a sticker on the vehicle how much oil was injected or placed into the system or a qualified tech would document on the work order; so if the system takes 5 ounces of refrigerant oil, then at least 1 or two ounces would have contained dye; this amount is not 100% dye but rather an amount of oil that contained dye.

Dyes don't 'clog up' systems and a dye does not displace refrigerant. The oil that contains the dye is in a liquid state whereas the refrigerant is in one of two states (gas or liquid). Adding one or two ounces of oil will not raise pressures nor will it impact the performance of the system to any major degree noticeable (if you have too much oil in system you can saturate the evaporator's tubes however which will reduce the heat transfer, but an ounce or two extra oil usuall does not do that in our experience. Remember, if the vehicle lost refrigerant chances are it probably lost some oil too along the way, especially an older vehicle (in most cases we find cars that had refrigerant added and they did not add additional oil to replace that what was lost over the years; especially with DYI'r's.

3) Stop Leaks - we prefer to avoid them at all costs. Why? It is a known fact in this industry that stop leaks "can" do more damage then they are worth. We have seen this first hand on a few clients cars that tried to take the cheap route to stop a leak. Stop Leak additives can gum up and clog driers, expansion valves and compressors. Want proof? Contact MACS, or the Mobile Air Conditioning Society, they published an article on this a few years back and provided graphic pictures of the mess. I'm not saying that this quick and dirty procedure won't work, I'm saying its not worth the gamble.

Is this empirical data? Like someone says "with my perfectly charged 88 911 at this temperature I observe this high side pressure".

If not what is it based on? I've seen these rules of thumb and wonder how they work. Is the optimal r134 charge one that makes the high side full of liquid at the lowest ambient?

Well, I opted to go ahead and replace the hose and recharge to get my system up and running for Spring and then monitor the compressor shaft seal for leak severity. Its been roughly three weeks since the recharge with relatively little A/C usage but system is holding a charge nicely. Today was the first real A/C need of the season. 88 sunny degrees on the commute home with vent temps in the high 30's/low 40's at temp setting 5 o'clock at 50% fan speed. On max cold setting I got the vent temps down into the high 20's.:D System mods: Griffiths Duel Kuehl fender condensers, evap, next gen front condenser, variable fan speed controller and Hurricane evap fan upgrade, barrier hoses, three fans on rear deck lid condenser, ProCooler and DIY recirc hose from bowtie to evap inlet (all mods posted on various threads):D:D:D

Yes. The numbers are pretty close. But I'd leave the deck lid down 'gently' resting on the service hoses. (Did i ever tell you of the time I was working on a turbo, had the deck lid up supported by a broken 3 iron shaft and a strong wind hit the deck lid and the lid slammed down shut and locked and could not get it open by pulling the release latch..... oh boy... ).

Is the data 'empirical' ? hmmm, depends if you are a fan of Thomas Samuel Kuhn.

Don't forget to document your data.

The 'optimal R134a charge' would be stated as "the least amount of refrigerant, that won't be detrimental, needed to achieve the results you wish that won't be detrimental" ... too much and you drive up the system pressures and lessen the life of the system (compressor), too little and you don't get enough oil movement.

When there is talk of filling to a given pressure I assume the table above is what we should use. Right? Here is a graph compared to the standard R134a temp/press chart. Let's say the the temp. is 100F, the pressure target is 275psi, that seems close to the high pressure limit where the switch might open. Maybe that is the idea. Fill to just under the trip pressure.



http://forums.pelicanparts.com/uploa...1312857994.jpg