Low cost hermetic seals

[splin]

Does anyone know where  in the UK I could buy low cost hermetic seals in small quantities (around a dozen)?

I’d like to encapsulate some resistors in a small metal box; Kovar eyelet seals were suggested here in an earlier thread, whereby a Kovar tube is hermetically sealed axially in a metal ring with a glass seal between. This can be soldered into a hole in a tin or copper box and a copper wire fed through the Kovar tube and soldered to it to complete the seal. This avoids the thermal EMF problems of Kovar leads.

Are feed through capacitors an alternative? The ones sold by Digikey etc. don’t seem to specify they are hermetically sealed or the lead material and can be quite expensive at $5+.

So what are the options and who sells them? What about re-purposing second hand components with appropriate seals? I'd guess most suitable components would be expensive as new but perhaps there may be some which go cheaply on Ebay?

[rvalente]

Panel mounting BNC connectors would fit your needs?

They're are shielded, can be soldered in the metal box or add some RF shielding.... and are CHEAP!

[TimFox]

Another style of BNC connector achieves reasonable hermeticity through an O-ring seal on the shell, and a non-permeable inner connection.
See  http://www.amphenolrf.com/031-4238.html
I remember using these at reasonable vacuum levels in grad school, down to maybe 10^-6 mm Hg.
For real vacua, < 10^-8 mm Hg, we bought ceramic feedthroughs from the usual suppliers (Ceramaseal and Varian).

[rvalente]

May these would fit your needs...





http://www.keyelco.com/category.cfm/Terminals-Terminal-Boards/Terminals-PTFE-Insulated/p/509/id/515

[AF6LJ]

Are you trying to keep moisture out, or a gas such as dry nitrogen in?

[splin]

Panel mounting BNC connectors would fit your needs?

They're are shielded, can be soldered in the metal box or add some RF shielding.... and are CHEAP!

Interesting but I'd be surprised if these provide a hermetic (gas-tight) seal as it wouldn't typically be required for most applications. Could you solder the body to a metal box without subjecting the plastic (PTFE?) to excessive heat thus damaging the seal?

[rvalente]

Nope, probably will not keep the gas inside under pressure.

Was not mentioned gas hermetic, thought would be moisture, dirt or RF tight.

Fluid hermetic and low cost are words that simply wont match, better reach top vendors like Amphenol and Harting connectors...

[NiHaoMike]

Make your own.


Not sure how much pressure you need to seal against, but HVAC stuff typically operates at up to a few hundred PSI.

[AF6LJ]

Panel mounting BNC connectors would fit your needs?

They're are shielded, can be soldered in the metal box or add some RF shielding.... and are CHEAP!

Interesting but I'd be surprised if these provide a hermetic (gas-tight) seal as it wouldn't typically be required for most applications. Could you solder the body to a metal box without subjecting the plastic (PTFE?) to excessive heat thus damaging the seal?

You can get gas tight BNC connectors but you might not like the price, we use to use them and TNC connectors that were rated to provide a hermetic seal. We tested every one that was installed on one of our transmitters.

[splin]

Are you trying to keep moisture out, or a gas such as dry nitrogen in?

Good question. It looks like I misunderstood the term hermetic which I assumed meant gas tight but actually only seems to mean water-tight (or better). I want to keep both water vapour and oxygen out to prevent corrosion of the resistors to minimize drift for many years. They wouldn’t be pressurized beyond those caused by normal room temperature variations so no high vacuums but I’d prefer a seal that I could have confidence in for many years which suggests a glass-metal seal. I guess I could pressurize the enclosure with something like dry nitrogen and periodically top it up if it lost pressure but it would be a pain to have to do and would still risk the ingress of fresh oxygen if not done carefully.

May these would fit your needs...





http://www.keyelco.com/category.cfm/Terminals-Terminal-Boards/Terminals-PTFE-Insulated/p/509/id/515

I’m not sure how you could provide a gas tight seal between the PTFE and the hole in the metal box - I guess it would have to be a very tight interference fit but as PTFE doesn't have much elasticity I would not be confident of its integrity.

Another style of BNC connector achieves reasonable hermeticity through an O-ring seal on the shell, and a non-permeable inner connection.
See  http://www.amphenolrf.com/031-4238.html
I remember using these at reasonable vacuum levels in grad school, down to maybe 10^-6 mm Hg.
For real vacua, < 10^-8 mm Hg, we bought ceramic feedthroughs from the usual suppliers (Ceramaseal and Varian).

I'm not sure the BNC connectors would do the job which although specified as hermetic don't specify they are gas-tight.

Can small quantities of the Ceramaseal/Varian seals be obtained easily at reasonable prices?

[Kleinstein]

Depending one the degree of gas tight, it might work to use emaneled copper wire and expoxy glue. The wires should be long enough, not to heat the glue to much on soldering. Here it helps if the material the wires pass through is rather thick, not just 1 mm or less. In the old days sometimes a kind of tar is used for seals.

The ceramig feed throughs specially for vacuum are often quite expensive.

[Gyro]

If your application is DC, then solder-in feedthrough caps might give you a combination of reasonably gas-tight sealing and EMC filtering.

Edit: Eg.
http://www.ebay.co.uk/itm/Ceramic-Feed-Through-Capacitor-100pF-FOUR-CAPACITORS-/400979029100?hash=item5d5c366c6c
(one of many listings)

[splin]

Make your own.

Not sure how much pressure you need to seal against, but HVAC stuff typically operates at up to a few hundred PSI.

Yes that looks tempting; however I guess the leakage of the refrigerant through the epoxy would be pretty low and so it wouldn't be noticed given the relatively large amount of refrigerant available. In my case the metal box would cause positive and negative pressures to be applied to the seal as the room temperature and air pressure varies. Standard epoxies are permeable to oxygen so I'd have to source one that is but I still couldn't be sure that I'd made a perfect seal without getting them helium leak tested or similar. Or I'd have to ensure the box was positively pressured at all times which could be a pain to do.

[splin]

Depending one the degree of gas tight, it might work to use emaneled copper wire and expoxy glue. The wires should be long enough, not to heat the glue to much on soldering. Here it helps if the material the wires pass through is rather thick, not just 1 mm or less. In the old days sometimes a kind of tar is used for seals.

My understanding is that special oxygen-impervious epoxy would be required which might not be easy to source. I guess it might work if the epoxy seal was sufficiently long and narrow by passing the wire through a thin copper or brass tube but it would be tricky to completely fill the gap between the tube and wire with epoxy without getting a lot of trapped air bubbles which would defeat the purpose. I'd have no idea how to calculate how long the seal would need to be to be effective though.

The ceramig feed throughs specially for vacuum are often quite expensive.

I suspected that would be the case - but the pressure difference shouldn't exceed 2 or 3 psi I'd have thought if the the metal box is thin enough. Any idea how expensive?

[daqq]

I've heard of spark plugs being used in a similar manner - they are abundant, cheap, and must withstand a lot of abuse in environmental terms.

[voltsandjolts]

Search ebay for "glass feedthrough", there are a few options.

Or cut the can lid off of a TO3 package and use the two glass-metal feedthroughs 

[splin]

Search ebay for "glass feedthrough", there are a few options.

Ah-ha - thank-you - Perfect!

http://www.ebay.co.uk/itm/Russian-military-miniature-glass-Feedthrough-insulators-lot-of-100-/251041737116?hash=item3a7340e99c

Or cut the can lid off of a TO3 package and use the two glass-metal feedthroughs 

Except that I'd have to drill axially down the centre of the Kovar leads without damaging the glass-Kovar seal to allow me to pass copper wires through - not impossible but very tricky! (I want to avoid the very high thermal EMFs which would arise from the resistor lead to the Kovar lead connection on the inside of the enclosure and the Kovar lead to copper wire connection on the outside).

[LaserSteve]

Get a Minicircuits mixer in the metal can, such as a SRA-1 or SRA-7.  Unsolder can, which takes some heat as they use a solder with a bit of silver in it.  Voilà!  8 PIN solderable seal with  pins on a multiple of standard 0.1 inch spacing.   Plus the kovar or 52 alloy shell is nickel plated so you can solder it as well.

Steve

[langwadt]

a pcb with few pads on? solder on both sides, bolt or solder to box

[splin]

Get a Minicircuits mixer in the metal can, such as a SRA-1 or SRA-7.  Unsolder can, which takes some heat as they use a solder with a bit of silver in it.  Voilà!  8 PIN solderable seal with  pins on a multiple of standard 0.1 inch spacing.   Plus the kovar or 52 alloy shell is nickel plated so you can solder it as well.

Steve

But wouldn't I have exactly the same problem as using a TO3 package? Presumably the feed through pins are solid Kovar so how would I get my copper wire through the seal? As I said previously I don't want to have two copper - Kovar junctions in the circuit because of thermal EMFs.

[splin]

a pcb with few pads on? solder on both sides, bolt or solder to box

A simple solution, but I doubt that the glass-fibre/epoxy PCB material is impervious to oxygen.

[LaserSteve]

Prepare for a confusing post, you will have to browse some web sites... Also low cost and hermetic in the same sentence doesn't happen unless you like nickel plated solder cup DB9 connectors with a glass core or BNCs.   Those can get down to 28$ each or so,  Digikey or Mouser stock them.

Surplus Sales of Nebraska has some solderable hollow capacitor FTs in stock cheap:

http://www.surplussales.com/Feedthrus/FTsolder3.html




If you need pure copper to glass or copper to ceramic  seals, those CAN be made, but they are NOT easy to make, nor are they reliable. For the most part they  went away when Dumet,  Kovar and Nickel to Glass  seals were perfected.  With the exception of some specialty applications, pure copper hermetic seals went away long ago.

For copper hermetic seals thru glass, at first,   Houskeeper Glass to Metal  Seals were used. Not fun to make as the oxidation state on the copper must be exactly right. Which is why Dumet Glass to Metal Seals were invented for light bulbs and vacuum tubes. On a Dumet seal, drawn  copper is the outer layer but it has a nickel-iron  alloy inner.  So basically these were for soft leaded glass, and not very useful for military and aerospace connectors. The copper melting point is very close to the hard glass (Pyrex, Nonex, Aluminosilicate) working point, thus good seals are difficult to make. So pure copper feedthrus are reserved for TC and High Power, High Vacuum work.

Omega has some low cost ones, but not that low cost:

http://www.omega.com/pptst/PFT2.html?gclid=CPGJw7iB88cCFQmpaQodXhkB6w

As for the copper EMF  portion, Sorry,  I  Missed that part.  I have some expensive ones here in a drawer here at work. Actually a common part in university research. They are made, and they show up from time to time on Ebay.  However they are not cheap. You can use the copper halves of TC vacuum  flanges or solder a wire into a hollow feedthru or flange.  High power Vacuum Flanges  have copper leads, but are not inexpensive.
 
http://mpfpi.com/stock-products/thermocouple.aspx

https://www.lesker.com/newweb/menu_feedthroughs.cfm?section=tctypek&init=skip

http://www.mdcvacuum.com/DisplayContentPage.aspx?cc=d68eacb2-f915-42dc-bc1e-9496cfe70ffa


If your looking for solderable hollow tubes thru glass or ceramic, you can try contacting any of the classic feedthru makers:

http://www.ceramicseals.co.uk/netscape/index.html

http://www.ametekinterconnect.com/
http://www.teknaseal.com/
http://specialtysealgroup.com/ceramic.php

Usually I ask if they have over-runs laying around. Sometimes they do, Sometimes they don't...

If it were me, I'd order the Ebay bag of 100 Ex Soviet seals...

Would you be happy with a ceramic loaded epoxy?  Hysol 1C White is what we use for vacuum work.. It is basically the core material used for making Torrseal Epoxy which is used on high vacuum chambers . As well as for sealed glow discharge tubes that only need to last a year or two.  Its basically 70-80% MgO dust.

Steve
 

[babysitter]

Go for hollow feedtrough caps, I just think they are fine - Stettner used to make them, you can get surplus from Oppermann Elektronik, germany, for example.


edit: http://www.oppermann-electronic.de/html/durchfuhrung.html DFK types

[TerraHertz]

http://www.ebay.co.uk/itm/Russian-military-miniature-glass-Feedthrough-insulators-lot-of-100-/251041737116?hash=item3a7340e99c

Ah, that's nice, thanks. Ordered two boxes.

Does anyone know what the best solder and bulkhead materials for these would be? Aim is to make high vacuum feedthrough ports.
Something I have not done before.

I'd assume a brass bulkhead, silver solder, and copper wires soldered only in the extended bit of feedthrough tube. But I don't know how critical expansion matching and soldering temp would need to be to avoid fracturing the glass seal.

[SeanB]

Not silver solder, ordinary 63/37 lead based is best. It is soft and will not whisker, plus will not crack the glass metal seal. Those feedthroughs are vesigned for it in any case, and are pre coated with it as well. Case material best is brass plate, drill it and then dip in molten solder to coat it, then wipe down while the solder is still molten to leave a thin film and clean the holes, then solder in the feedthroughs, then do the leads and finally seal with solder. You probably will want to fill with either vacuum degassed water free oil or with wax.