Valves. Why glass?

[GrahamN]

I've never understood why glass is necessary for valve manufacture. Surely there would be huge advantages in using a metal, ceramic or high melting point plastic case? Apart from the ability to see the filament light up, I can't think of any real advantages of glass, and I would have though that other materials can be much easier to work with, particularly if mass production was wanted.

Or am i missing something very basic?

[BGmidsUK]

A metal case would probably alter the characteristics of the valve - and would require an insulated base for the pins anyway.

Plastic - I don't think there were any suitable high melting point plastics at the time valves were in common use (apart from Bakelite maybe?).

As for ceramic, I'm not sure how well it could preserve the vacuum. Glass I would think is the easiest material to work with as far as sealing them is concerned - and it's also proven to stand the test of time



BG

[paulsherwin]

Metal cased valves were actually very common in the 30s and 40s, but it was cheaper to use glass in mass production once the technology had been sorted out.

http://www.r-type.org/exhib/aaa0015.htm

Paul

[GrahamN]

I'm not really referring to old valves - really just idly wondering why someone doesn't start production of popular valves using more modern methods, rather than the other thread about new valves being made in glass. After all, transistors used to be made with glass cases, but these gave way to metal, presumably in the interests of economy and ease of manufacturer. Whilst valves might be a bit more of a challenge, I can't think of any obvious reasons why an alternative to glass wouldn't be more economical, easier to manufacture, and more resilient.

[Station X]

The difficulty in making a valve is to get a seal around the pins. I suspect that even metal cased valves used glass for this, so why not make the whole thing of glass.

Military gear often uses metal cased valves, so I assume they were more rugged, more reliable and more expensive with glass types.

[John M0GLN]

Originally Posted by Station X

The difficulty in making a valve is to get a seal around the pins. I suspect that even metal cased valves used glass for this, so why not make the whole thing of glass.

Yes, otherwise you've got double the seals to make, pins to glass and glass to case, and double the chances of sealing faults.

John.

[Nickthedentist]

Originally Posted by Station X

Military gear often uses metal cased valves...

I can remember opening one of these up as a child, and finding... a pretty conventional glass valve inside!

So are "metal cased valves" ever glass-free?

Nick.

[Station X]

I've got a duff metal cased valve here. I'll open it up with a tube cutter and see what's inside. I will of course take the necessary safety precautions.

[Brian R Pateman]

Glass was cheap and the techniques were very well known. There were hundreds of skilled craftsmen who had been involved in the manufacture of scientific instruments or filament lamps.

That was where many of the early manufacturers started (Pye - was originally WG Pye, an instrument maker, Philips - started making light bulbs).

Even today glass is a cheap comodity compared to metals and it is an easy material to work with.

Regards,

[GrahamN]

Brian - I think that is my point. When valves started being manufactured, it was obviously sensible to use techniques and materials in good supply at the time. But if starting out now with a new manufacturing plant to produce valves for audio amps etc., why start with glass when cheaper and easier materials and techniques are now surely available?

I understand the point about sealing the pins, but surely glass isn't the only method, and I would think not even the best method considering what other alternatives are available now.

I also understand that valve characteristics may differ according to their construction, but surely it wouldn't take too long to redesign a valve to have similar characteristics to an older version, where such characteristics were desirable.

[XTC]

One thing you have to do in making a valve is to make the final seal when the valve is exhausted. This is very easy to do with a glass tube, just heat it gently and the tube collapses and can be nipped off. Even the metal octals used this technique.

Nuvistors didn't have any glass but they were made in a vacuum chamber and were quite expensive.

Glass was a good enough solution which would fairly easily make a good seal, lent itself to mass production techniques and produced a robust enough product for most purposes. The metal octals were widely used, but there was no metal equivalent of the B7G and B9A valves, (someone catch me out on that) although there were ruggedised versions. I'd say it was a question of cost vs benefit.

There are modern ceramic transmitting valves and magnetrons. I'm not sure how they're made.

As for modern production audio valves, they're making a traditional product with a well established process. Doing the R&D to make the same thing in metal, investing in the equipment to produce them and then probably find that the market rejected them , doesn't seem like a promising venture.

Pete.

[russell_w_b]

Originally Posted by GrahamN

I've never understood why glass is necessary for valve manufacture. Surely there would be huge advantages in using a metal, ceramic or high melting point plastic case?

Most, if not all, high-power radio valves now have ceramic envelopes (4CX250B upwards to - and including - the likes of Thales TH558).

Glass ones are prettier, though!

[GrahamN]

Glass ones are prettier, though!

Ha Ha - You got me there! Oh well, back to the drawing board...

[Station X]

Surely glass gives a more sparkly sound? Crystal clear I should think. The same goes for glass encapsulated transistors, especially germanium ones.

[paulsherwin]

Originally Posted by Station X

I've got a duff metal cased valve here. I'll open it up with a tube cutter and see what's inside. I will of course take the necessary safety precautions.

Somebody has beaten you to it Graham: http://www.r-type.org/static/6j5.htm

Paul

[Station X]

The valve I planned to disect was a 6J5 just like the one on the website. However I was planning to release the vacuum by drilling a small hole in the top of the can.

[Steve_P]

I thought the US Army valves were metal so that they would be more rugged and not radiate from the set - the enemy could pick this up. And laugh...

As for why glass - well, how about so we could see if they were lit up?

Cheers,

Steve P.

[mickjjo]

Originally Posted by Nickthedentist

I can remember opening one of these up as a child, and finding... a pretty conventional glass valve inside!

That might have been an EF50, a glass valve with an aluminium cover:-

http://www.r-type.org/static/ef50.htm#006

Regards, Mick.

[kalee20]

Glass, as well as sealing around the pins, is itself in bulk virtually impermeable to air. Many plastics aren't (even ridiculously thick).

Many transistors and IC's are indeed plastic encapsulated, but then the device inside isn't overly sensitive to contamination because it's covered in silicon dioxide (highly inert), glass, or silicone elastomer before encapsulation. So the fact that the plastic doesn't provide a true hermetic encapsulation does not matter. Germanium does react with atmospheric moisture, and that's why germanium transistors are glass or metal encapsulated - the few plastic types made, just died!

Plastic transistors, like TO92 or TO220 aren't up to the specifications required for many military applications. There is a hi-reliability version of a TO220 package. It's metal, with a welded-on lid, and guess how the leads are insulated? With little glass seals!

[ampzilla]

Here you can see The Blackburn Story ,tubes in production and perhaps get some answers ?

http://www.techtubevalves.com/about_us/film_reels.php