Are metal electron tubes made entirely out of metal or does the metal merely cap some glass which constitutes the real tube? Are metal electron tubes easier to make than glass ones? I would think they are, since one avoids the problems of making a glass-to-metal seal. But maybe there are other problems, such as getting a good vacuum, avoiding unuathorized circuit paths, etc. Which is actually cheaper to make? I realize the answer might depend on whether you are only making one or making 10,000, but in either case.
Ignorantly, Allan Adler snipped-for-privacy@zurich.ai.mit.edu
According to discussions on news:rec.antiques.radio+phono they were made both ways, some were metal cased glass tubes, while others were really metal tubes. You can use Google groups to search out the threads.
Thanks, I found one thread on rec.antiques.radio+phono using deja-news. I only found discussion of the metal-to-glass seal type of tubes. The thread contained references for some books that I'll try to look at: (1) Tyne, The saga of the vacuum tube (2) H.J.van der Bijl, The thermionic vacuum tube and its applications (3) Fred Rosebury, Handbook of electron tube and vacuum techniques (4) Herbert Reich, Principles of Electron Tubes
It seems to me that if one is interested in making one's own tubes, it must be a lot easier to machine a metal tube, seal it with flanges and O-rings, evacuate it and have electrodes passing through the metal than it is to learn to make reliable glass-to-metal seals. The discussion also focuses on the large number of parameters to worry about, each of which presumably requires a new tube. But this all metal construction presumably makes it possible to open it up and change things in it. So this approach seems to be more suited to experimenting with tubes if one is so inclined.
Since this boils down to a metalworking project, I'm cross posting this to rec.crafts.metalworking.
Ignorantly, Allan Adler snipped-for-privacy@zurich.ai.mit.edu
Suggest you consider the CRT in PCs and television receivers. These are probably the most widely used vacuum tubes being manufactured today. Considering size, strength of materials, ease of fabrication, insulation, etc., IMO you can consier the methods used in their construction to be state of the art.
And they had *metal* bells at least in the '50's, maybe early '60's.
...Jim Thompson
--
| James E.Thompson, P.E. | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona Voice:(480)460-2350 | |
| E-mail Address at Website Fax:(480)460-2142 | Brass Rat |
| http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
This is called a virtual leak and when you are tying for 10^-6 torr can be a real pain! Conflat is good, but you really need to be able to bake out. Of course you can use a stupid big pump to overcome some of these problems, a big sorbitron pump with short large diameter lines can pump really fast (as can a cryopump).
If I was playing with valves, I would probably use a belljar with the feedthru electrical connections made via the base plate.
Do not seriously consider a diffusion pump, they are a pain (especially when the line to the backing pump comes up to atmosphere by accident). Even when operating normally they need a cryotrap to keep backstreaming under control for most applications. A turbopump is a much more satisfactory pump for most uses.
Regards, Dan.
--
** The email address *IS* valid, do NOT remove the spamblock
And on the evening of the first day the lord said...........
.... LX 1, GO!; and there was light.
Experimenting yes, if you can leave your pump running.
But for manufacturing it is not so good. The glass-to-metal seal is easy enough: make sure your glass and pin material belong together.
I agree that blowing a nice multi pin glass base with the pins in correctly is not easy at all, but it lends itself well to an industrial process.
I'd think you can buy these for custom tube work, in a specified glass.
Now, if you leave your tube connected to the pump you could even use vacuum epoxy to seal everything.
O-rings are tricky. If you have a square channel with a round ring in, gas will be trapped where round meets square; this air will eventually escape. The same applies to flanges coming too close together, and really anything that you cannot heat very well.
There are (or were) all metal tubes, some of which were available in glass, too. The 6V6 was one of those, IIRC.
Nuvistors were all metal, too.
On the other hand, there were tubes such as the Mullard EF50, which looked like all metal, and even had a screw locking ring, but were really an all-glass tube inside an aluminum can.
All-glass tubes are (or were) much cheaper to make than all-metal, one you'd got the plant installed.
As to the question of making one or 10,000, most widely used tubes were made in their *hundreds of thousands*. remember that the replacement market was bigger than the OEM market, at least for consumer types.
One off tubes, such as large transmitter tubes were largely made by hand, and cost thousands of bucks each. The really big ones were demountable. You could take them apart and fit new cathodes (filaments, really, though they looked like a bent six-inch nail). Such tubes were run pumped all the time, with a diffusion vacuum pump pack nearly as big as the transmitter cage. Big transmitter types were a mixture of metal and ceramic, with either a forced-air or water cooled anode. There were some ingenious ways of stopping the cooling water circuit from loading the output, such as winding the pipes to form an RF choke.
The BBC in the UK were running 200KW tube transmitters in the MF broadcast band into the 1980s. May still be, for all I know.
--
Then there's duct tape ...
(Garrison Keillor)
nofr@sbhevre.pbzchyvax.pb.hx
If you want a very good vacuum, use stainless steel construction, flanges with knife edges and oxygen free copper rings for sealing. Avoid screw holes and such which can trap dirt or moisture on the inside. Flat flanges and copper wire also works for sealing. Brass construction is also ok, but beware of the zinc vapour pressure. Epoxy may be OK, but unless it is cured by heat, I would be vary of outgassing. Apiezon has a full range of sealing compounds for UHV.
Stainless steel is actually better than glass, as glass will allow hydrogen through in time. Bead blast or pickle to get rid of any oxides after welding. Clean everything in ultrasonic cleaner or similar equipment, rinse with distilled water, clean again with degreasing agent such as acteone or hexane. A vapour degreaser is excellent. Bake everything after assembly while you are pumping. This helps the outgassing. Heat can be applied with a heat gun or heating tape.
Z> Allan Adler wrote: >> It seems to me that if one is interested in making one's own tubes, it >> must be a lot easier to machine a metal tube, seal it with flanges and >> O-rings, evacuate it and have electrodes passing through the metal than >> it is to learn to make reliable glass-to-metal seals. The discussion also >> focuses on the large number of parameters to worry about, each of which >> presumably requires a new tube. But this all metal construction presumably >> makes it possible to open it up and change things in it. So this approach >> seems to be more suited to experimenting with tubes if one is so inclined.
Z> Experimenting yes, if you can leave your pump running.
Z> But for manufacturing it is not so good. The glass-to-metal seal is Z> easy enough: make sure your glass and pin material belong together.
Z> I agree that blowing a nice multi pin glass base with the pins in Z> correctly is not easy at all, but it lends itself well to an Z> industrial process.
Z> I'd think you can buy these for custom tube work, in a specified glass.
Z> Now, if you leave your tube connected to the pump you could even use Z> vacuum epoxy to seal everything.
Z> O-rings are tricky. If you have a square channel with a round ring in, Z> gas will be trapped where round meets square; this air will eventually Z> escape. The same applies to flanges coming too close together, and Z> really anything that you cannot heat very well.
The demountable tubes used in WW2 British CH radar (45MHz, modified TV transmitter) had no seals at all, just optically flat ground flanges. (This from a guy I knew who had worked on them. I'm not quite *that* old)
--
Then there's duct tape ...
(Garrison Keillor)
nofr@sbhevre.pbzchyvax.pb.hx
I'm not good at this stuff, so can you please explain what that means? The mental picture I get is of a cylinder fitted into the flanges, with everything machined as flat as one can test visually, with no o-rings other sealant. The thing might be pumped to maintain the vacuum and it all works because even if it leaks a little at the ends, the vacuum is more efficient at pumping out the gas than the end leak is at supplying gas. Is that correct?
Ignorantly, Allan Adler snipped-for-privacy@zurich.ai.mit.edu
I don't understand: do they supply something you can put your own anode, cathode, grid in or do you send them a design for the whole tube and they manufacture one for you? In either case, who does this (not that I can afford it)?
Ignorantly, Allan Adler snipped-for-privacy@zurich.ai.mit.edu
It may seem so, Allan, but the opposite case is often true.
First, in making vacuum tubes, you are always going to end up with either glass to metal or metal to ceramic seals, because the seals must maintain a hard vacuum over a wide temperature range, the differential thermal expansion characteristics of various materials, and and a number of other factors. These considerations pretty much rule out that use of any materials except for vacuum hard metals of specific types (these metals exhibit similar thermal expansion characteristics as that of glass or ceramic). O-rings are not an option, except for experimental configurations where the vacuum pump remains connected to the vacuum tube at all times.
Traditionally, glass headers with special metal connection pins typically made of Kovar or Invar are mass produced and sold to tube manufacturers so that the tube manufacturer does not need to address process details of producing the glass to metal seal. The tube elements are assembled on the glass header and generally spot welded together. The header mounted assembly is then covered with either a glass or metal envelope and fused together.
The tube can then be evacuated by pumping on a piece of glass tubing attached to either the top of the tube (typical of miniature tubes like a 6AL5 or 12AU7) or from the bottom of the header (as was done with most octal based tubes (like a 6L6 or 5Y3). Once a moderate vacuum exists in the tube, the pump down tube is tipped off with a torch, sealing the still incomplete vacuum. I should not that the tube is almost always kept in an over while being pumped down, because both the glass and metal tube structure ougasses while being pumped down.
The ultimate level of vacuum in the tube is produced by "gettering" once the tube has been sealed. Gettering involves the vacuum deposition of a special metal film on a portion of the tube surface inside the vacuum. The metal film attracts and contains most of the remaining gas molecules in the tube after pumpdown and tip-off operations are complete.
Large industrial and transmitting tubes are similarly manufactured, although the scale and geometry of these tubes are completely different. The same basic concepts involving glass to metal or ceramic to metal seals are employed. Many of these are large triodes and tetrodes, x-ray tubes, magnetrons, klystrons, etc. With the larger triodes and tetrodes, often the plate or anode forms the outside surface of the tube envelope, with grid, cathode, and heater connections being passed through a glass or ceramic headers at each end of a cylindrical plate or anode. Eimac type 3F2500F3 and 4CX250 are examples of external anode tube structures, while the Eimac 4-400A is an example of a glass power tube envelope structure.
Methods other than glass/ceramic to metal seals have rarely been used in the manufacture of vacuum tubes, and then only with moderate to poor success rates. An example of this would be the videcon tube (typically the 6198 family) whose sensitized faceplate was glued to the tube body, because it couldn't survive the heat of a glass to metal to glass seal. These actually lasted quite a few years in service, and were usually replace for reasons other than loss of vacuum. Metal envelope CRTs didn't do as well, prematurely became 'gassy', and were discontinued after a few years of trial.
I hope this sheds some light on what is now an increasingly extinct technology except for large power devices, extreme audio enthusiasts, and survivalist radio operators. :-)
Allan, I believe you will be hard pressed to find many suppliers of standard vacuum tube parts today, but back during the 1950s many standards tube components were available to tube manufacturers like RCA, GE, Raytheon, etc. Also, the tube manufacturers sold tube parts to one another, so once in a while you'll find a GE plate assembly in a RCA 6SN7, etc.
You could buy cathode/heater assemblies, grids, plates, etc., in standardized configurations for both experimental and production use. Needless to say, there was a pretty limited market for these types of components even back then. I personally suspect that most of these parts came from the end-of-run inventories of the various firms, which were then purchased and made available to the industry in limitied quantities.
You can still buy headers and tube envelopes from specialty firms searchable using Google, and maybe even tube parts and elements somewhere, since duplicates of some of the older and more popular tube types are being produced by new firms who seem unlikely to have resources to build all of the required tooling to produce individual tube parts.
In a nutshell, that's right. If you grind surfaces flat enough, they'll stick together so tightly that you can't pull them apart, only slide them. Ordinary workshop slip gages will demonstrate that. If you make the flanges wide enough, there's very little leakage. By the standards of the time, the vacuum maintained [dunno what it was, but certainly worse than
10^-6 :-) ], was adequate. I don't suppose they minded a bit of negative grid current, the exciter was something like 500W. I _do_ know that if you ran the finals up too fast, the things would arc over and trip the PSU down, and you had to start over again.
--
Then there's duct tape ...
(Garrison Keillor)
nofr@sbhevre.pbzchyvax.pb.hx
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.