How To Create Hermetically Sealed Electrical Connections

[Eric Strebel] is no stranger to pressure and vaccum tanks, regularly using them for all manner of resin casting jobs for his product design business. However, sometimes it becomes necessary to run equipment within a pressure tank, such as for rotomoulding or other similar jobs. In order to get power into a tank under pressure, [Eric] built a special plug with a hermetic seal to do the job. (Video, embedded below.)

The build starts with a large metal plug which screws into the pressure vessel, into which a square recess is machined. For the electrical passthrough, [Eric] selected GX-16 aviation connectors, in this case packing six conductors. The connectors are hooked up back-to-back through the hole in the metal tank plug, using bare copper wire. This is to avoid insulation on wires acting as a channel for gases to pass through. With the connectors wired up and an acrylic disc in place to stop overflow, the metal plug is filled with resin to create the hermetic seal.

Results are good, with the connectors functioning electrically and the resin acting as a perfect seal. There’s a small risk of short circuit with the exposed copper conductors, but [Eric] is exploring some easy solutions to avoid issues. We’ve seen his work before, too – like this great discussion on cardboard as a design tool. Video after the break.

24 thoughts on “How To Create Hermetically Sealed Electrical Connections

  1. LEMO connectors are pressure / vacuum-tight. Or at least I recall reading so. (I can’t find a reference now)
    I have had one fitted to the oil-filled spindle of my milling machine for about 10 years with no hint of any oil egress. So they ar, at least, fluid-tight.
    This only applies to the solder-bucket versions, as crimp-then-fit pins clearly are loose in the housing.

    1. Can confirm that LEMO connectors have vacuumtight variations, I’ve personally used them and (more importantly) they’re available in the LEMO catalog :)

      Most of the solder-bucket versions can be ordered as vacuumtight (with certificates)

          1. The IP rating has nothing to do with the pressure/vacuum tightness of the socket itself. That is the protection of the contacts in the mated pair.
            With no plug in the socket the IP rating of the socket is 0[1] but the socket itself is still gas tight (if the mounting flange is)

            [1] Not really, probably 20, ie “finger safe”

    2. Another cheaper option, for low levels of pressure-tightness might be Neutrik Speakon connections (great for stepper motor connectors, 4 poles @ 20 or 40A / 250V)
      “The chassis connectors are airtight for audio applications (leakage 30 cm3/s at 10 kPa differential pressure) unless otherwise stated.”
      (I don’t know how that compares to other types of connector, but that’s not a huge test pressure)
      I like Neutrik connectors, they have a nice solid feel, but only cost a few quid each.

    3. I know this round military connectors, The solder variants are only watertight in connected state, they have loose fitting pins in a plastic housing. The crimp pins are inserted in a rubber inner housing . They are tight also without cap.

  2. Doesn’t acrylic produce a lot of out-gassing? If so, seems like a bad idea to have that on the inside of the tank.

    I’d be concerned about some outgassing on the plastic parts of the connectors as well.

    1. That depends of course on the requirements. For really high vacuums most polymers are not suitable any more. But for applications like the degassing of resins or some gas laser experiments epoxi is usable.

      1. In one of my previous jobs I was seriously considering £20,000 for an injection moulding tool to make only 200 parts. This turns conventional injection moulding economics on its head, but the idea was to replace a machined glass-ceramic part in an X-ray tube with a polyimide injection moulded part.
        I think that the problem was that we thought it ought to work, but making a single prototype to prove it out was a big commitment.

  3. Might also be able to source some surplus mil spec bulkhead connectors either loose or on cheap surplus equipment. A lot of them are used on aircraft to move from unpressurized to pressurized spaces which is exactly what is needed here.

      1. Of course a filler is a good idea. It increases strength and probably reduces outgassing as the mass of organic resin is lower. But I would use a mineral filler like silica (quartz sand) instead of a metallic one.

  4. Why not use isolated magnet wire? That would remove the possibility of accidental short circuits. Yes I know you have to mechanically remove the isolation before soldering or use a special solder-through type (which will still require a pretty high soldering temperature). Don’t think the thin isolation layer would cause any problems with the sealing.

    1. Done this method before. Needed to get power into a vaccume chamber to run a heating element and lighting. Used three, 2mm enamel coated copper wires through a brass fitting filled with epoxy. Could use the brass body of the fitting as one terminal too.
      Worked pretty well.

  5. This won’t work well in the long run. The resin and the copper have different expansion coefficients. So after it sets, if it gets too hot or cold, the copper will work the resin away and create a pathway for gasses to pass. You need the packing material to be semi-soft and under load to keep contact with the wire. They commercially sell electrical passthroughs with PTFE or rubber packing that work at all sorts of pressures and temperatures:

    https://douglaselectrical.com/products/studseal

    1. Most epoxi materials are not glass hard, so depending on the required temperature range this could well work. Of course if you get to higher temperatures, >100°C PTFE is much better than epoxi.

  6. This is technically not hermetic. Any diffusion path through a polymer (epoxy in this case) will allow the slow movement of gas. To be hermetic, glass or ceramic insulation is needed around the conductors. This might be good enough for the application, but is not hermetic.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.