Smoke Pen

[Jerry Bruette]

Anyone here have any experience using a smoke pen to find leaks in a vacuum system?

[Jim Koepke]

Back in my days of working on automobile vacuum systems those of us who smoked used a cigarette to see where the smoke was being sucked in.

jtk

[Jerome Stanek]

I use incense to find leaks

[Bill Dufour]

I would say a vape cigarette would do it safer then a cigarette. I have heard of party fog machines being used. For automotive they make fog machines that pressurize the vacuum system to locate leaks.
Bill D

[Jim Koepke]

I would say a vape cigarette would do it safer then a cigarette. I have heard of party fog machines being used. For automotive they make fog machines that pressurize the vacuum system to locate leaks.
Bill D

The OP, Jerry, didn't say what the vacuum system being worked on is. One company employing me as a drafter made very large diesel engines. These used logic devices for engine controls that were driven by vacuum. The tubing for these systems was measured in hundreds of feet. Some of the stainless steel tubing lines were in racks with as many as a dozen tubes or more.

A vape cigarette doesn't produce smoke without someone sucking on it. In this situation, Jerome's idea of using incense may be a better source of inexpensive smoke for an occasional use.

Having no idea of what Jerry is working on nor any experience with a smoke pen my reply was based on personal experience with an old habit that no longer plagues me.

jtk

[glenn bradley]

I use incense to find leaks

I also do this. Cheap and long lasting. I did a whole system with one stick.

[Jerry Bruette]

[QUOTE=Jim Koepke;2987637]The OP, Jerry, didn't say what the vacuum system being worked on is.

I'll try to explain the vacuum system we were trying to troubleshoot at work.

We make a product that is pressurized with a mixture of dry air and 10% helium. The product is put into a chamber which is evacuated by vacuum, alot of vacuum and quickly. Then some test valves open and let a sample of the air go into a molecular drag pump which pumps the sample across a mass spectrometer. The mass spec checks for helium. If our product has any leaks it shows up as excessive amounts of helium in the mass spec.

Any leaks in the system, piping, flanges, o-ring seals etc. will also show as excessive helium. The product may not have any leaks but the mass spec can't tell the difference between a system leak or product leak.

Our way of trouble shooting is to run an empty chamber and do the leak test. Then spray pure helium around the piping components and watch for a high reading on the mass spec. Problem with our method is some potential leak points are so close together we can't tell which joint the helium is being sucked into--can't see the helium. This is where the smoke pen would hopefully come into play, showing us which joint is the leaker.

[Leo Graywacz]

How about this?

Smoke in a can (https://superiorsignal.com/smoke-products/smoke-check)

[Jeff Body]

I made something like this to find leaks in my car's engine. I found out real quickly I had a bad intake manifold gasket.
Nichrome wire
Wick
DC power supply
and an air compressor
and some fittings
Works amazingly well

https://www.youtube.com/watch?v=g5W9v0qUF2s

[Jim Koepke]

Jerry, this sounds like something you will be doing on a regular basis. For that it may be worth your while to research the various products, read the reviews and then make an investment in the smoke pen or pencil to meet your needs. It looks like there is a wide variety of them on the market with prices from ~$30 on up into the hundreds.

jtk

[John K Jordan]

Jerry,

I understood a helium mass spectrometer leak detector is about as sensitive as you can get for tiny leaks. I can't imagine detecting tiny leaks with smoke unless the leak was relatively "huge". Due to the tiny size of the molecule (sometimes described as "slippery") helium will move through tiny leaks quicker than air. It will even slowly move through glass! - back before laser diodes were available we played with helium-neon lasers and helium "soaking" could sometimes be used to restore helium that had slowly leaked out of the glass tube.

More common and a lot cheaper is a halon or multi-gas leak detector. In the 70s I worked in a large government facility where we tested many things, including pressure vessels, and that was one of the methods, using a hand-held battery-operated detector.

If the leak is large, such as a poor seal in a cyclone dust collector bin, the smoke can work ok. But you really need a "draft" with significant air movement - you apply smoke and watch how the smoke moves. I never tried a smoke pen or smoke generator, just an incense stick and a smoker I use with my honey bee hives. (Fortunately I must have sealed the bin an cyclone body well and had no detectable leaks)

[Ole Anderson]

How about a stethoscope? A small leak in my dust collector will make a very small whistling sound. I presume most following this thread would be applying the comments to their dust collection system, not a mass spectrometer.

[Bill Dufour]

I thought the way my dad did it was to pressurize with He then use a He detector. He was running vacuums as low as they can go. This was for a particle accelerator. Then it takes days to get all the gas out of the system. Some leaks with seals etc leak under vac but not pressure so a pressure test is not a cure all.
Then you may have little pockets out gassing for days or weeks. But, I think that is not a big issue for you as long as the out gassing rate is fairly constant and you keep measuring it.
Are all your bolts vented etc? especially on the target support which is often a last minute design with changes not clearly thought through. Are you heating the chamber?

Bill D

[Steven Cooper2]

If you're able to use a helium leak detector, you can probably pump the system pretty far down already, somewhere below 1 Torr probably? I doubt that leak size would be enough to move a noticeable amount of smoke.

What we do a lot (HV/UHV) stuff it just mask off problem areas with electrical tape and you can selectively expose various suspected leaks.

You can also bag the system or suspect fitting either to selectively include or exclude helium.

If you can, another thing we do just trying to get systems to rough down into the sub Torr range can be to apply something like water (bad for vacuum systems, but not a big deal for some things) and it will briefly plug the leak and you can watch the pressure drop like a rock.

There is also the paint it with RTV method.

Helium also generally pumps poorly so it may show as an excess of helium because it lingers longer than heavier gasses.

[Steven Cooper2]

I should note that if you are running UHV, water or something else liquid is probably a bad idea. You can use Apiezon vacuum putty, it might be a little more difficult to get fast results, but it works well and won't gunk up you fancy stainless UHV stuff. It should help you narrow the target area.

And another note, your "mass spec" is it a helium leak detector or RGA system for system operation? If an RGA, you can probably use a different gas, I think neon may work and look for that.

You mention that you cant see the helium above, like no spike in signal or unable to suppress background enough to get a decent reading. If the former, you may be able to differentially pump the system and check that, if the latter, virtual leaks as mentioned above, could be an aggravating reason.

I hate vacuum leaks, troubleshooting can go on forever...

[Jerry Bruette]

I should note that if you are running UHV, water or something else liquid is probably a bad idea. You can use Apiezon vacuum putty, it might be a little more difficult to get fast results, but it works well and won't gunk up you fancy stainless UHV stuff. It should help you narrow the target area.

And another note, your "mass spec" is it a helium leak detector or RGA system for system operation? If an RGA, you can probably use a different gas, I think neon may work and look for that.

You mention that you cant see the helium above, like no spike in signal or unable to suppress background enough to get a decent reading. If the former, you may be able to differentially pump the system and check that, if the latter, virtual leaks as mentioned above, could be an aggravating reason.

I hate vacuum leaks, troubleshooting can go on forever...

Go on forever...no kidding! We just spent over 20 hours trying to find a leak. We normally have pretty good luck with spraying the helium on the system and watching for spikes on the mass spec. This leak was especially hard to find, not big enough to give us a "gross leak" but big enough to suck helium in from a distance which makes pinpointing the leak location a giant guessing game.

Different shifts tried different methods including pressurizing the system to about 85psi and using a Fluke leak detector. I haven't used it yet but I guess it takes a picture and looks for the frequency of sound where the leak is. In the end we disconnected all the piping from the mass spec and starting reconnecting one piece at a time while running a leak test after each piece until we found a suspicious elbow. Then we took that elbow and pressurized it to about 4psi and water bathed it, thing looked like an aquarium aerator.

I'm not familiar with some of your abbreviations is UHV ultra high velocity, and what is RGA. Our drag pump runs at about 1500 Hz and won't run up if the other vacuum pumps don't reach the proper vacuum level. I'm not sure what that is and it has to be met in a certain time period also, this is a production machine. We test the condition of our vacuum pumps by blinding them off with a vacuum gauge and turning them on. I don't know the CFM of our pumps off hand but our small pump hits 14 milliTorr in about 9 nine seconds and our big pump hits 40 milliTorr in about 40 seconds.

Our vacuum systems are usually pretty reliable but can be a real challenge when they get cranky.

[Steven Cooper2]

Glad to hear you found it. if you have the option of pressurizing, snoop (or soap water) works pretty well, cheap too.

Galvanic corrosion? Cast fitting?

UHV is ultra-high vacuum. 10^-9 torr+, by the time you have a vacuum system in this range, you are baking out equipment, turbo pumps, ion pumps, NEG coatings all kinds of fancy stuff. No more plastics or rubber allowed. a little water here and you would have to re-bake the system.

RGA is Residual Gas Analyzer. its a broadband mass spectrometer, used for can plot all sorts of gas ion momenta, you can view what sorts of gas species remain, hydrogen points to metal outgassing, other things to different things, oils, silicones, you can at least get an idea of whats causing problems. These are expensive and delicate. Very neat, nobody ever lets me use theirs cause my stuff is too dirty (Plastics, epoxies...) I don't blame them.

[Jerry Bruette]

Glad to hear you found it. if you have the option of pressurizing, snoop (or soap water) works pretty well, cheap too.

Galvanic corrosion? Cast fitting?.

Looked like a stress crack next to the flange on a long radius ISO elbow.