Getting a good seal with NPS

For Short Term Learning Discussions ONLY. This area is for CURSORY questions and connecting with other users ONLY. ALL technical contributions need to be made in the appropriate forums and NOT HERE. All posts are temporary and will be deleted within weeks or months. You should have already search the extensive FAQs in each of the forums before posting here as your question may already be answered.
Post Reply
jameslatshaw
Posts: 4
Joined: Tue Apr 02, 2024 11:06 pm
Real name: James Latshaw

Getting a good seal with NPS

Post by jameslatshaw »

Hello,

Background:
I am hobbyists, attempting to build a sputtering magnetron. I have a simple vacuum chamber which I sneak some ultra thin magnet wires into, in order to carry my HV and produce a plasma (which I have completed successfully, I have a photo attached if you are interested). This setup was a cool proof of concept but I need to bring in a thicker gauge wire and provide some type of cooling water to my target to keep it from melting and ruining my magnets.

My question:
I am experimenting with galvanized steel pipe (NPS threaded with Teflon tape) and custom cut high temp silicone washers (shore 50). These work OK but I can't seem to get to ~1 Pascal like I could before.

1) Should I be using a different seal than Teflon tape? Ideally, one that can take high temp as this pipe will be near my target
2) Any good tips for identifying leaks?
3) Anyone else have success with NPS threaded piping holding a medium vacuum?

Best Regards,
James
Attachments
Plasma from an early test! :)
Plasma from an early test! :)
JoeBallantyne
Posts: 329
Joined: Tue Jul 20, 2010 4:08 pm
Real name: Joe Ballantyne
Location: Redmond, WA

Re: Getting a good seal with NPS

Post by JoeBallantyne »

I don't have any experience with NPS threads - which I guess you are supposed to seal using an o-ring, or washer.

My understanding is that the vast majority of galvanized steel pipe has NPT threads, not NPS threads. So, if you are actually using NPT, please consider the rest of this post. Also, you should make sure you are not using standard galvanized pipe with NPT threads with NPS threaded connectors. NPT threads are designed to work with other NPT threaded parts, not NPS threaded parts.

I have had a terrible time trying to get NPT connections to vacuum seal correctly. In general they just DON'T.

So, my advice to you, is to AVOID THREADED NPT CONNECTIONS LIKE THE PLAGUE.

Because otherwise they will plague you with leaks... :-(

Use KF/QF/NW connections - those are designed to hold a good vacuum, and they are very easy to use. (Of course, you can also use Swagelok, or VCR, or VCO connections too. All of those will hold a good vacuum as well, but are not as easy to use as KF and in some cases have one time use parts.)

When I was testing out my low cost mechanical vacuum pumps, I had to convert from an HVAC flare connection to KF, and the most straightforward way was to connect a couple of parts together with an NPT connection. But the only way I could get a good seal, and I do NOT know if it will last long term, was to use a bit of vacuum grease on the male threads, then wrap them very tightly with PTFE tape - at least 5 times around, maybe more, and then put a little vacuum grease on the outside of the PTFE tape, and then screw the parts together until I got to about 5-10 ft-lbs of torque. And this was on two separate brass to stainless connections, both with 1/4" NPT threads.

That seemed to work ok, but 2-3 wraps of PTFE only with no vacuum grease, leaked. Just vacuum grease alone worked at first, but then when things got hot, started to leak as well. My experience is you MUST wrap the PTFE tape a good number of times around the threads - more than normally needed for water pipe connections. I don't even know if it will work without the vacuum grease, as I did not try that. (Well, I did, but with only a few wraps, not many, and just a few did leak.)

One way you can verify if you have a leak on any threaded connection, is to drip some high quality mechanical vacuum pump oil onto the connection so that it completely surrounds the surface of the female threaded part, where the male threads are going in, and see if your vacuum suddenly pulls much lower.

Mine did, so I knew for sure the NPT connection was leaking. It dropped from 180 microns down to 12 microns in like a second or so. Huge difference.

NPT connections SUCK for vacuum.

Don't use them if you can at all avoid them.

My 2 cents.

Joe Ballantyne
jameslatshaw
Posts: 4
Joined: Tue Apr 02, 2024 11:06 pm
Real name: James Latshaw

Re: Getting a good seal with NPS

Post by jameslatshaw »

Greetings Joe!

Yes, you are right, they are NPT not NPS (whoops!).

Thank you for the feedback. I will look into trying additional wraps of Teflon tape and try to find some vacuum grease and see how that goes. How does the vacuum grease handle higher temperatures?

Also, maybe I am being too cheap trying to make my own rubber gaskets and I should just invest in some o-rings. I will also try this and see if that improves my seal.

Thanks again! I am so excited to have found this forum. It seems like there are a lot of really cool projects going on :)

Kind Regards,
James
User avatar
Liam David
Posts: 529
Joined: Sat Jan 25, 2014 5:30 pm
Real name: Liam David
Location: PPPL

Re: Getting a good seal with NPS

Post by Liam David »

I have gotten small (1/8") NPT threads to seal to 1e-6 torr, but I think anything larger than that will be much more challenging unless you use (vacuum-rated) epoxy. To be more specific, I sealed both stainless and brass MNPT's into stainless FNPT's using several wraps of the thicker teflon tape, no vacuum grease needed. Some non-negligible fraction leaked outright which was solved sometimes by applying more torque and sometimes by redoing it. If you can replace the threads with KF fittings, those are a much better option.

It's not advisable to use vacuum grease, and certainly not the cheaper stuff, for high temperatures. It will just pollute your vacuum and may degrade your coating quality. In the vast majority of cases, people use vacuum grease when they shouldn't. Cleanliness, smooth sealing surfaces, and proper compression are pretty much always the way to go.

I definitely suggest using standard-size Viton o-rings from McMaster, The O-Ring Store, or the like, for your seals. They really aren't that expensive and will pay for themselves in time saved. If your sealing surfaces aren't scratched and the o-ring grooves are properly sized (look up the Parker o-ring handbook), dry (i.e. no vacuum grease) Viton o-rings will seal into 1e-9 torr. Avoid silicone seals as they outgas a lot. Viton is best, buna-n is ok.

I also recommend replacing any galvanized components with non-galvanized versions. Zinc is one of the worst metals to use in vacuums, especially with high temperatures. Look up the vapor pressure of zinc.

Joe's suggestion to use pump oil to check for leaks is ok, as long as you don't mind oil mist in your chamber. Coatings beware. 99% IPA is better if you have a gauge that has a gas-dependent response (ex: thermocouple, pirani) since it vaporizes without residue. The measured pressure will jump when the IPA is sucked through the leak.

What's all the orange in your chamber? Are those bumpy rubber-looking sheets your seals? If so those are likely as much a problem as the NPT's.

More pictures of your setup, and especially the seals, will help us provide advice.
User avatar
Dennis P Brown
Posts: 3189
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Getting a good seal with NPS

Post by Dennis P Brown »

A few words; any permeant magnet (ring) will focus the sputtering plasma. DC voltages from 300 to 600 volts is all one needs. A clean system is essential and argon gas for good quality work. The vacuum system needs to reach low millitorr to get a reasonably clean film. Through setting the desire pressure depends on many system parameters.

Sputtering can be done with air but the films will likely be oxidized.

Teflon tape is a terrible material to have exposed to a plasma but is ok for threads used for vacuum connectors.
jameslatshaw
Posts: 4
Joined: Tue Apr 02, 2024 11:06 pm
Real name: James Latshaw

Re: Getting a good seal with NPS

Post by jameslatshaw »

Hi Liam!

Thank you very much for that wealth of information. That is very helpful and I especially appreciate your notes on the o-rings.

That photo was from my first attempt at trying to excite a plasma (no cooling water and no piping in that setup) and has since been decommissioned. The orange rubber is a soft shore silicone gasket but the bumpy purple sheets are my wife's yoga mat turned gasket. It worked well for 1 compression but not so good once it deformed. Yoga mats should be used for stretching, not for sealing :)

I am still in the process of rebuilding it and will try to get back soon with the results.

Best Regards,
James
jameslatshaw
Posts: 4
Joined: Tue Apr 02, 2024 11:06 pm
Real name: James Latshaw

Re: Getting a good seal with NPS

Post by jameslatshaw »

Dennis P Brown wrote: Wed Apr 03, 2024 9:43 pm DC voltages from 300 to 600 volts is all one needs.

Hi Dennis!

Thank you for the specification you mentioned. That is valuable guidance to me.

Do you find that you need to ramp up the voltage to something higher than 600V to form a plasma and then lower it after the plasma is formed?

Kind Regards,
James
User avatar
Dennis P Brown
Posts: 3189
Joined: Sun May 20, 2012 10:46 am
Real name: Dennis Brown

Re: Getting a good seal with NPS

Post by Dennis P Brown »

The issue relative to voltage is working pressure for a given system; most sputter systems work in the 20 to 50 milli-torr range and 600 volts is an upper max for a magnetron type system (higher for systems with no magnet.) Not that higher won't work but one gets better films. This is for gold, Nickel and Aluminum. Also, argon flush/leak to hold the desired pressure. For air these values will likely differ but not by large amounts. The ring magnet makes striking a plasma far easier and better concentrates the plasma. So, the voltage doesn't need to be as high. The supply I used could be varied from 50 to 1000 volts DC.

As for your question, no. A 30 - 50 mTorr chamber with a magnet should easily 'glow' under 600 volts DC. If the pressure is too high, then yes, one will need higher voltage. This can be done - say work at a torr - but certainly only under argon and the quality might be rather poor.
Post Reply

Return to “New User Chat Area”