HN Debrief

Making glass-to-metal seals for home­made vacuum tubes

  • Hardware
  • Materials
  • DIY
  • Manufacturing

The post documents a home experiment in making the hardest part of a vacuum tube at home: getting metal conductors through glass without losing vacuum. The author focuses on copper because its oxide bonds well to glass, then explores whether thin wires and careful sealing can make that workable for a homemade device.

If you want a tube that lasts, treat glass-to-metal sealing as a materials-matching problem, not a glue or fabrication problem. For one-off experiments, reuse existing vacuum hardware like neon electrodes or proper vacuum flanges instead of inventing seals from scratch.

Discussion mood

Interested and respectful. People liked the craftsmanship, but the dominant reaction was practical and corrective: copper-to-glass is a neat experiment, not the proven recipe for a durable vacuum tube, and high-vacuum sealing punishes shortcuts hard.

Key insights

  1. 01

    Matched alloys made tubes practical

    Kovar, Dumet, platinite, and related alloys are the real story behind durable glass feedthroughs. They were engineered to expand at nearly the same rate as specific glasses, which kept seals from cracking over thermal cycles. That is why old vacuum tubes, gas tubes, metal-can transistors, and even lamp stems could be mass-produced reliably without relying on exotic handwork.

    If you are designing or sourcing a seal, pick the glass and metal as a pair. Looking for generic wire or generic borosilicate first is the wrong starting point.

      Attribution:
    • adrian_b #1 #2
    • tliltocatl #1
    • mmmlinux #1
  2. 02

    Vacuum failures come from slow leaks too

    A tube does not fail only when a seal visibly cracks. Rubber and plastics leak by diffusion, low-vacuum materials outgas, and a bad metal-glass pairing can reopen after a few heat cycles even if a repair initially holds. Getters are not optional decoration either. They mop up the residual gas that remains after pumping and sealing.

    For any vacuum build, screen materials for outgassing and permeability before you think about assembly. A seal that survives bench testing can still drift into failure in use.

      Attribution:
    • MisterTea #1
    • bluGill #1
    • adrian_b #1
    • Animats #1
  3. 03

    Vacuum tube manufacturing never vanished

    Audio tubes keep a consumer market alive, but that is the small and nostalgic corner of a broader industry. X-ray tubes, photomultiplier tubes, laser tubes, and transmitter tubes still need the same core sealing and vacuum techniques, so the manufacturing knowledge persists in active use rather than in museum lore.

    If you need process knowledge or suppliers, search adjacent tube industries, not just guitar-amp communities. The most relevant expertise may sit in scientific or industrial hardware.

      Attribution:
    • BoxOfRain #1
    • drum55 #1
    • musictubes #1
  4. 04

    Off-the-shelf parts beat inventing seals

    Premade neon electrodes already solve the glass feedthrough problem cheaply for some home builds. If you want modular metal vacuum hardware instead, proper ConFlat flanges and copper gaskets are the viable route, not improvised O-ring grooves. The practical message is simple: borrow existing vacuum components wherever possible and spend your effort on the active device.

    For prototypes, buy the vacuum boundary and build only the tube internals yourself. That cuts risk fast and gets you to a working experiment sooner.

      Attribution:
    • tyingq #1
    • ludicrousdispla #1
    • labcomputer #1

Against the grain

  1. 01

    Copper can still teach the right lesson

    Even though copper is the wrong long-term material for commercial-style tube pins, the experiment still demonstrates a valid point about seal geometry and process. The vacuum is normally pulled through a separate exhaust path, not through a magical live insertion of wires into an already evacuated bulb, so a successful copper-to-glass joint is still meaningful as a fabrication exercise.

    Do not dismiss the project just because industry used Kovar. As a learning step, proving you can make any sound glass-to-metal joint is useful before you optimize materials.

      Attribution:
    • CamperBob2 #1 #2
  2. 02

    A one-off tube does not need factory standards

    Judging the project by whether it meets commercial triode vacuum levels misses the point. Part of the value is simply pushing far enough to see what is physically possible with home tools and modest materials, even if the result is not production-grade.

    If your goal is exploration, set a lower bar than decades-long service life. You can learn a lot from a fragile success.

      Attribution:
    • rigonkulous #1

In plain english

ConFlat ↩
A standard all-metal vacuum flange that seals with a soft copper gasket for high-vacuum systems.
Dumet ↩
A wire alloy used for sealing electrical leads through glass, especially in lamps, because its expansion can be matched to specific glasses.
getter ↩
A reactive material inside a vacuum tube that captures leftover gas after sealing, helping maintain the vacuum over time.
Kovar ↩
An iron-nickel-cobalt alloy designed to have nearly the same thermal expansion as certain glasses, so metal-to-glass seals do not crack.
outgas ↩
To release trapped or absorbed gases from a material, especially under vacuum.
platinite ↩
A nickel-iron alloy developed as a cheaper alternative to platinum for glass-to-metal seals.

Reference links

Vacuum sealing and materials references

  • Getter
    Used to explain how vacuum tubes remove residual gas after sealing.
  • Materials for use in vacuum
    Shared as a primer on outgassing and material suitability for vacuum systems.
  • Parker O-Ring Handbook
    Referenced as practical guidance for proper O-ring seal design and why casual groove designs are wrong.

Component examples and DIY alternatives

Materials mentioned by name

  • Fernico
    Posted as another established sealing alloy that the article did not mention.