If you’ve ever taken a close look at a vacuum tube, you’ll have seen the seals around the pins that keep everything air-tight while providing the the device’s electrical contacts. As [maurycyz] finds out, it’s not an easy process to get right.
The problem is one of both chemistry and thermal expansion, as while a good seal can be made between glass and red copper oxide, it remains very difficult indeed to stop the glass cracking on cooldown due to differing thermal expansion properties. We’re led through a variety of experiments including surface treatments and flattening the metal to a sheet, with varying pros and cons. The most successful seal on the page comes from very thin tungsten wire, though hardly the most practical conductor for a vacuum tube.
It’s a fascinating investigation for the casual reader, taking them into the properties of metal-glass bonds and the difficulties involved in making them. We have even more respect for the people who make their own tubes after reading it.
The interesting point is they learned to solve this very early one (with light bulbs) but also with million of vacuum tubes, that had in some cases a sizable number of pins passing through.
Pinged Google and got this:
Dumet Wire: The industry standard for decades. It consists of a nickel-iron core (often a 42% nickel / 58% iron alloy) clad in a copper sheath. The copper creates a strong hermetic bond with the glass.
Platinum: Used in early vacuum tubes (pre-1915). It expands and contracts at almost the exact same rate as soda-lime glass and bonds well to it (it would be an expensive option though!)
Kovar: An alloy of iron, nickel, and cobalt specifically engineered for glass-to-metal seals. It is widely used with borosilicate glass in heavier-duty or power tubes
This is why I read the comments . . . I’m six months into a new job, and have been scratching my head as to why we only use “dumet wire” for certain things – now it makes even less sense, but with historical implications of the “if it ain’t broke don’t fix it” type. No metal/glass, but fifty years back ic bodies were ceramic and certain industries move very slowly.
I was always under the impression that platinum was the early leader due to its similar thermal expansion and wettability – to be followed by tungsten for similar reasons.
I really don’t have much to contribute here – other than a thank you for inadvertently solving a mystery in my brain.
If I remember it correct, they used a special glas aloy around the pin. However, if I had to do something like this, I would use the lower part of an old tube. There are plenty in the world…
However, the idea is nice to make your own nixie. But why? I have a watch at my wrist and a mobile phone! Hm…probably for a chess computer? There must be a reason for building a stylish part of our living room that is NOT a clock! Some ideas?
Yes, on older tubes this was often Uranium glass.
This problem was solved generations ago by Kovar family alloys. See the book “Laboratory Scientific Glassblowing” by Paul Pinnet, that’s the bible behind the explanations. The key is wetting prep and the right coefficient of thermal expansion matching material to the glass.
The trick for tungsten wire is you need to oxidize it before you coat it in glass. You heat it in a flame to get it hot to oxidize and you are looking for “Kentucky fried Chicken brown”, as the glassblower I learned from called it.