Regulator leakage

[DaveO]

I'm working with a new shielding gas regulator, and when I first fitted it to the cylinder I there was a leak at the nut where it connects to the cylinder. I tightened the nut and then pressurized the regulator only: flow valve off, cylinder valve open to show 2000 psi on the contents gauge, then cylinder valve off.

Then I left the room to get a cup of soapy water for additional leak testing. During the 2 minutes I was away the contents gauge lost about 400 psi, indicating a leak somewhere. I tested all the fittings on the regulator and found the connecting nut was still leaking, so I cranked down on the nut, applying more torque than I felt should have been necessary. This stopped most of the leak, but bubbles still form under pressure.

This is a very minor leak but my question is: what amount of leakage is acceptable? It's inert shielding gas, not going to flash, in the garage so not in a living space. Is the acceptable level zero leakage?

[zoama]

Yeah, zero. There's a problem withe the regulator or bottle.

[DaveO]

... or bottle.

Thanks- I thought about the bottle too, but with only one regulator and only one cylinder I have no way to test it. I'll see what I can figure out.

[CGCINC]

Could you use Teflon tape or is that not good for this application?

[CGCINC]

I've got the original regulator that came with my 250EX if you need a spare.....

[zoama]

Could you use Teflon tape or is that not good for this application?

No, it seals on the ball end, not the threads.

[jerky]

maybe the valve or the bullnose part has a scratch on it. really you want zero leakage, but if its only leaking say a few psi an hour, its not that much that you will even notice it. your only going to have the valve open when welding then shut it off. over the life of the cylinder, you may lose 20psi once your all said and done. really you wont know if it was the valve or your regulator until you test it on a new cylinder.

[geezer]

I'm working with a new shielding gas regulator, and when I first fitted it to the cylinder I there was a leak at the nut where it connects to the cylinder. I tightened the nut and then pressurized the regulator only: flow valve off, cylinder valve open to show 2000 psi on the contents gauge, then cylinder valve off.

Then I left the room to get a cup of soapy water for additional leak testing. During the 2 minutes I was away the contents gauge lost about 400 psi, indicating a leak somewhere. I tested all the fittings on the regulator and found the connecting nut was still leaking, so I cranked down on the nut, applying more torque than I felt should have been necessary. This stopped most of the leak, but bubbles still form under pressure.

This is a very minor leak but my question is: what amount of leakage is acceptable? It's inert shielding gas, not going to flash, in the garage so not in a living space. Is the acceptable level zero leakage?

One thing to remember about valves on bottles,,,,,,,,they only seal in two positions,,,fully closed and FULLY OPENED,,,if you have a pressure drop the regulator could be sealing properly but if the bottle valve is only half open it will leak around the valve stem and you will think the regulator is at fault..

[DaveO]

Thanks for the replies. And thanks, Geezer, I'd never heard of the full on / full off thing with the cylinder regulator. I had shut the cylinder valve off completely after I pressurized the regulator: the delivery valve was off and I closed the cylinder valve, leaving pressure in the regulator's contents gauge. I also dipped the flowmeter outlet into the soapy water to make sure there was no leak there.

CGCInc, thanks for the offer of your spare- still not sure I have a problem yet, it may be the cylinder.

[Spike]

i used to have the same sort of problem when i first started welding. On high pressure tanks (argon, co2, helium, trimix, etc) the valve on the wheel only seat when all the way closed (the ball seat is seated on the bottom), or all the way open (the ball valve is seated on the top).

The other thing i've had problems with is having dust or junk in the fitting between the regulator and the bottle. Before you attach the regulator to the bottle, check the mating surfaces of the regulator and the inside of the bottle, and be sure there are no burrs or scratches on either one. Also, be sure to crack the bottle's valve for a second before you attach a regulator - this will blow out any dust or junk that may have settled in the neck of the bottle that can make it hard to seal, or ruin your regulator.

You shouldn't have to really crank down on the fitting to seal it. If you put too much torque on it, you will damage the regulator fitting, and then it doesn't matter how clean or nice it is, it will forever leak a little bit.