Please let us know your experience with glass to metal seal

[Richard Hull]

One of our local HEAS guys, Tim Raney, has worked extensively with both house keeper seals, graded seals, (U glass), and nitrited tunsten rod.. the bulk of them leaked until he started using delecate 20 mil tungsten rod and backing all seals with TorrSeal. Making a good glass to metal seal that is forever proof against leaks is an art for the amateur and a science for the manufacturer.

I have found that using old vacuum tube contacts by first breaking of the tip to kill the vacuum and then diamond sawing the lower half of the tube with contacts away. This will yield a nice feed through once the tube element are cliped away. Melting and joining a good finished, manufactured product to a home built project of glass always works well. This is not a solution for all projects, but will get many jobs done fast and forever.

HV feed throughs are best done with a top part of the classic and commonly found 1B3GT or any anode capped HV rectifier or dead transmitting tube. Why fight city hall when 50 cents will get you a factory done glass to metal sealed tube at a hamfest. Don't over look ceramic tube options.

Richard Hull

[Jerry Biehler]

There is also the method of using indium metal to solder glass to metal. Indium will wet just about anything.

But going from 70mm glass to metal? Yikes. You could do a transition from kovar to glass and then weld the kovar to your tubing like on an ion gauge.

[steve_rb]

Yes I noticed but Hysol 1C looks is good up to 10^-6 but Torr seal is good up to 10^-9. This was the reason I focused on Torr seal.

[Richard Hull]

I have made many glass to metal and ceramic to metal seals with indium and indium-tin alloys. Beware! Low melting point critical. If your project gets hot as in a target electrode, anode cap, or even the envelope, the seal can fail.
Not good. If it stays cool, it is hard to beat an indium to glass seal.

Richard Hull

[steve_rb]

Hmm. Great. I really liked this indium and indium-tin thing. Richard could you please give some more info how I can do this? What critical points should be considered to have a good seal? Also I would like to know which type of metal can be sealed to which type of glass and ceramic?

Also how this seal acts at high voltagees (150 kv) and high temperatures (at least 200 deg C)?

[Carl Willis]

Indium melts at 157 degrees C.

[Richard Hull]

Pure indium has a much higher melting point than tin-indium. As Carl notes, even pure indium melts at a relatively low temp. Seals are made by cleaning the materials to be joined by soaking in very dilute HCl and then rinsing in de-ionized water. Molten indium is painted on the surfaces. Painted with a glass fiber brush. This is, effectively, "tinning" them with indium at this point, molten indium can be used to solder the seal closed. Again, if your item gets hotter than boiling water, forget indium seals.

Interestingly, indium remains molten and has no real boiling vapor pressure until 2000 deg C! A stick of indium can be used to write a conductive line on paper or wood as it is one of the softest metals known. It makes lead look like case hardened steel. There are no reports of indium poisoning or hazardous industrial results. If you are "that type person" some contact dermititis has been reported, as with all metals. A most intriguing metal.

http://www.indium.com/technical-documen ... d-its-uses

Richard Hull

[Jerry Biehler]

Indium is also used as a heat sink junction material especially in places like laser diode module mounts and getting good thermal conductivity on laser crystals or other optics and their mounts. It is also used by places like intel to interface heatsinks to cpus for testing, this way they dont need to mess with pastes.

The alloying of indium solder effects the melting point so you can use two different solders to solder things like laser diode dies to their mounts much easier than using one single melting point solder.

If you need to go to that kind of temp you will probably have to look at an o-ring gland seal.

[Rich Feldman]

Isn't there a fundamental dichotomy among the glass-to-metal seal applications being discussed?

On the one hand, we have apparatus that operate while connected to vacuum pumps.
On the other are sealed devices that are processed once, then expected to hold their vacuum forever.
Both had practical solutions 100 years ago.
Some materials and methods suit one, some the other, some both.

Which sort do you need, steve_rb?

[steve_rb]

I need the second sort. The one that holds the vacuum for years.

[Bob Reite]

Most hard vacuum "thermionic valves" aka "vacuum tubes". have a "getter" that continues to get rid of excess gasses after the tube has been sealed off. You may want to consider that for your project.

[steve_rb]

Yes I have noticed this. It looks getter material should be heated after tube sealed? Any experience with getter materials?

[John Futter]

You can get samples from Seas getters in Italy.
They are usually heated by inductive heating of the getter part that is on its own somewhere within the vacuum
Look at any Vacuum tube (ValvE) you will see a circular thing usually near the top of the valve and the mirror metalisation that came from the getter ring after it was heated to cherry red hot.
They are made from Zirconium and various reactive rare earths that are selected to getter various gases ie nitrogen hydrogen oxygen

[Jerry Biehler]

I need the second sort. The one that holds the vacuum for years.

Then all bets are off. You will need a commercial metal-glass transition and probably a getter like other have mentioned, There are getters that are fired electrically, the materials are deposited on a filament and you apply power to evaporate and activate the strip. Of course they need hermetic electrical feedthroughs.

[Richard Hull]

I agree with Jerry. Go commercial. They have years in the biz and have made reputations on good, long lasting seals. Gettering is really important if you hope to hold better than a 10e-5 torr seal forever. Common old receiving tubes are a prime example of how it is done right!

Richard Hull

[steve_rb]

I am OK with getters but no knowledge regarding this "metal-glass transition" thing. If I have got it right you mean all seals above not suitable for sealed tubes and I only have to use metal-glass transition (the one was at the above houskeeper video) right?
In that video copper gets hot and oxidized. Then glass melts and sticks to the copper oxide. Do you mean this sort of seal with "metal-glass transition"?

[Chris Bradley]

Might be easier to describe what your objective is, than the means you think you need to use to achieve it.

[steve_rb]

I am trying to build a custom made vacuum tubes like X-ray tube for a home made simple X-ray diffractometer. Also try to repair a few broken X-ray tube we have (1 target melted, 2 filament burned). Be windows are all OK. I might also give a scintillation detectors a try too.

[Jerry Biehler]

Good quality metal to glass transitions use something like kovar which has a similar CoE to glass and will bond and not shatter when it cools down. http://www.lesker.com/newweb/flanges/fl ... cfm?pgid=0

You can get x-ray tubes pretty cheap on ebay. And be darned careful with that beryllium, that stuff is pretty toxic.

[steve_rb]

Tubes in eBay are not suitable for X-ray diffractometers.
Thank you for your notice. Yes I am aware of Be toxicity.
Anyone has practical experience with Kovar seals?

[Bob Reite]

My experience with Kovar seals is in the day of glass transmitting tubes. You don't want them to get too hot or they fail. The glass transmitting tubes seemed to go "gassy" if sitting on the self for too long because the seal was never perfect. That's why in the old days we would alternate the working tube(s) with the spare(s) every six months or so.

For some reason, the metal ceramic seals used in modern transmitting tubes don't have this problem, so I run those till the rig won't make licensed power anymore, install the spare and send the weak tube off to the rebuilder.

Would a ceramic metal seal be suitable for your project?

[steve_rb]

Yes. As a matter of fact ceramic to metal might even be better. Have you got experience with ceramic to metal?

[Roberto Ferrari]

Steve and rest of posters,

I used to repair X-ray tubes and also diffraction X-ray tubes. Now do similar thing with special lamps.
Working with a glassblower, we messed in kovar to glass seals. It is a hard learning.
You will need to follow Dennis Brown' comments about CTE of glasses. Usually you have to do transitions, that means joining a couple of intermediate glasses to arrive to borosilicate.
As mentioned by Richard, recovering parts of radio tubes is wise and exciting.
When talking of wires, be careful with tungsten and molybdenum wires, they are made by pulverimetallurgy and many types doesn't stand high vacuum, due to micro channels between particles. You need a vacuum tube grade.

One point related to rebuilding tubes with beryllium windows. They are brazed to the anode and many times the brazing material gets corrosion...and you get leaks, very hard to correct. My suggestion in that case is to replace the whole window. I never tried alternative procedures -highly experimentals- like electroplating the punctured area or covering it with vaporized metal under vacuum chamber.

Getters come in two main families: evaporable and non-evaporable. Evaporable-ones are activated by RF or electric heating; non-evaporable can be activated in some cases (tube type and design) bombarding them with low pressure noble gas under approx. 300 V dc. That slowly vaporizes the active metal which deposits in the cold areas a film, very reactive to residual gasses. Others only are mounted close to hot spots.
Diffraction tubes in the past used electrically-activaded getters because the position inside metallic anode avoid the RF option. Some X- rays tubes use zirconium or tantalum disks, close to the heated anode, so when hot they start to suck gasses.

Please feel free to ask anything about these subjects.
Regards
Roberto

[Roberto Ferrari]

Steve,

I forgot to comment the hard time that is degassing the X-ray tubes.
You need to do it over the maximum power of the tube, generally taking the anode to dull red. And you do that simultaneously while pumping hard and applying filament and high voltage to the tube... Yes, you got it... generating X-rays.
We used to have an elevated furnace, cubic, approx. 1m by side, shielded with 1/4 " lead and 1/8" SS, with propane heaters and high voltage cables through it.
Everybody in the lab used dosimetric badgets.
Was pretty Frankenstein's lab! High voltage sparks were not uncommon.
I used a modified evacuation system inspired in the 40's, when pumping down transmission tubes. Go to the Electronics magazine from the WW II times, looking at the tube ads. They talk of a "charcoal trap", between the diffusion pump and the tube to be evacuated. Was a cage with charcoal and an electric heater. When starting the system you heat a lot the charcoal and pumps only with the mechanical pump; very slowly the charcoal releases mainly water and gasses retained. Then after the pressure improves, you start the diffusion pump, achieving 10-6 torr, degassing the evacuated tube thoroughly. Then you turn off the charcoal heater and the vacuum improves... two orders of magnitude! I did it replacing the dirty charcoal with molecular sieve pellets. When breaking the vacuum, I used dry nitrogen and after several cycles, the system arrived to 10-9 torr.
Regards
Roberto