Connect with us

epoxy in vacuum

Discussion in 'Electronic Design' started by Jamie M, Apr 21, 2012.

Scroll to continue with content
  1. Jamie M

    Jamie M Guest

    Hi,

    I am making some DIY vacuum electrical feedthroughs into an aluminum
    cylinder wall. I was thinking of just drilling a hole and epoxying in
    a short length of copper plumbing tube, that has a copper wire epoxied
    in it. Would this work ok for vacuum sealing and also for not too much
    vacuum outgassing from the cured epoxy? Also how thick of epoxy is
    required for 50kV standoff voltage!?

    cheers,
    Jamie
     
  2. Use a polyurethane potting compound. They are several hundred volts per
    mil thickness. You can pre-evacuate it after mixing, but prior to the
    end of its rated "pot life". It has to actually be applied by that
    point. Then, you can bake it or time cure it.

    So one tenth of an inch should be well over 50kV. That's 100 mils at
    several hundred volts per mil. They differ, but not by much.

    Epoxies, on the other hand, do differ a lot more in insulative capacity
    and cured properties as well.

    The poly stays attached to surfaces well, and doesn't gas, and is
    rubbery, albeit very tough rubber. Usually doesn't need primer as well.
    Also NASA likes it, so it has to be good.

    The epoxies usually cure much more firm, and less serviceable, if at
    all, and it is very stiff usually, and can fracture and gas. They can
    also detach from surfaces, with or without primer.
     
  3. Jamie M

    Jamie M Guest

    Hi,

    Thanks, would this stuff work?

    http://www.homedepot.ca/product/polyurethane-construction-adhesive-300ml/947644
     
  4. Klaus Bahner

    Klaus Bahner Guest

    Depends very much on what you want to achieve. In a "low-tech", one-off,
    home experiment, epoxy works well. No need to go for exotic industrial
    stuff, ordinary two component expoxies (hardware store) perform well.
    The tricky part is to avoid air bubbles when mixing - the cured
    insulation layer should not contain any voids.

    It's not so much the absolute voltage than the field strength
    determining how thick the insulation layer has to be. In your coaxial
    feed through design a few (5-10) millimeters should bring you on the
    safe side, when you observe that inner conductor diameter is about 1/3
    of the outer diameter.

    Epoxy based glues outgass like hell. Thus it depends on how low your
    vacuum shall be and what your pumping speed is, whether the perform well
    vacuumwise. Down to about 10^-3 mbar there shouldn't be hardly any
    problem. If you want to achieve high-vacuum (order of 10^-7 mbar), even
    a tiny blob of newly cured epoxy will make it take days under constant
    high speed pumping to reach that vacuum level.

    Hope that helps
    Klaus
     
  5. mike

    mike Guest

    I'd check with one of the high vacuum groups.
    I tripped over some threads recently that discussed this subject.
    Apparently there are epoxies that work well as vacuum seals, but a
    random selection probably won't.

    I'd also worry about the thermal expansion differences between the
    aluminum and copper. Epoxy might need to be flexible.

    http://www.teralab.org/index.htm

    Here's a quote from here:
    http://www.fusor.net/board/download_thread.php?bn=fusor_vacuum&thread=1148666207

    For educational purposes, going against the grain of conventional wisdom
    on high vacuum technique is definitely worthwhile, because commercial
    stuff made for the purpose is unavoidably expensive. I think your
    website is proof that you're on this track already. Acrylic (plexiglas),
    PVC pipe, and most other plastics are decent in a microtorr vacuum as
    long as they don't get hot. Copper water pipe and 3% Ag / Sn solder is
    likewise just fine (and don't bother with the challenge of trying to
    make inside joints). Hysol 1C epoxy is the same as the much more
    expensive Varian Torr-Seal, and works great as a glue or a makeshift
    structural material in vacuum. Teflon ball valves from a hardware store
    are very respectable at high vacuum. Metal-to-glass and metal-to-ceramic
    seals can be scavenged off of microwave oven magnetrons, CRTs, and other
    vacuum tubes, and then you can epoxy or solder these into apparatus.
    Heavy-wall Tygon tubing makes a great makeshift (and flexible) "neon
    sign" through the medium vacuum range, and from experience I will tell
    you that school kids love playing with things like that.
     
  6. qrk

    qrk Guest

    Adhesives don't bond well to aluminum due to the oxide layer. If this
    is for a short period of time (months), you're probably ok. Otherwise,
    use the expensive glass-metal feedthrus. Be careful of partial
    vacuums, it doesn't take much voltage to flash over at a partial
    vacuum. Saw a guy pulling down a chamber with 120VAC applied to the
    heater in the chamber. Beautiful flash. He spent the whole day
    scrubbing the copper off the walls.
     

  7. The only compound NASA approves for space applications and HV is
    "CONAP" brand (DuPont, IIRC) polyurethane.

    The stuff is so tough, it makes you think it would stop a bullet.

    It certainly stops any knife you can drive at it within a quarter inch.
    That is how good it grabs knifepoints.

    Anyway, no outgassing, and adhesion remains. It is not very
    serviceable because it is so turgid (durable).
     
  8. WoolyBully

    WoolyBully Guest

    Total bullshit.

    As pressure goes down, dielectric resistance goes up because the air
    molecules are farther apart. So, it takes more voltage in a reduced
    pressure chamber than it does at atmospheric pressures. It MAY be far
    less than the first arc in a full vacuum takes, but it will still be
    higher than in normal air.

    The volts per mil to acquire an arc in a air lattice is 100% dependent
    on pressure. That is what determines atom spacing in ANY gas.

    A partial vacuum has air in the chamber. There is only one type of
    "full vacuum".

    A full vacuum does not have the same behavior as a chamber with an air
    lattice in it. Once an arc is created, there are metallic 'gasses in the
    chamber, and on surfaces and carbon trails and arc path can form and
    substantiate themselves. At that point, the "full vacuum" can no longer
    be claimed. At all.

    So the arc in even the best chamber draw down will require a higher
    voltage than it will once the very first arc occurs. After that, it
    isn't really a fully vacuumed chamber any more.
     
  9. Carl Ijames

    Carl Ijames Guest

    Um, ever heard of a Paschen curve? The breakdown electric field of a gas
    goes down as the pressure falls from way above atmospheric, to a minimum in
    the range of .1-50 torr depending on the gas, then rises again as the
    pressure falls further. At high pressure the mean free path of an electron
    isn't enough to let the energy rise to the ionization potential before a
    collision slows it back down so ionization is very inefficient and breakdown
    field is high, and at very low pressures the gas density is low enough the
    electrons make it all the way from cathode to anode without a collision with
    a gas molecule so even though they have plenty of energy to ionize something
    there just isn't anything to ionize so the breakdown field goes back up. In
    that fun in-between region you get pretty glow discharges and all kinds of
    other phenomena.

    The main question the OP never answered is "what pressure are you trying to
    achieve"? That will dictate the materials needed and the physical layout
    needed for his voltage.

    -----
    Regards,
    Carl Ijames
    "WoolyBully" wrote in message

    Total bullshit.

    As pressure goes down, dielectric resistance goes up because the air
    molecules are farther apart. So, it takes more voltage in a reduced
    pressure chamber than it does at atmospheric pressures. It MAY be far
    less than the first arc in a full vacuum takes, but it will still be
    higher than in normal air.

    The volts per mil to acquire an arc in a air lattice is 100% dependent
    on pressure. That is what determines atom spacing in ANY gas.

    A partial vacuum has air in the chamber. There is only one type of
    "full vacuum".

    A full vacuum does not have the same behavior as a chamber with an air
    lattice in it. Once an arc is created, there are metallic 'gasses in the
    chamber, and on surfaces and carbon trails and arc path can form and
    substantiate themselves. At that point, the "full vacuum" can no longer
    be claimed. At all.

    So the arc in even the best chamber draw down will require a higher
    voltage than it will once the very first arc occurs. After that, it
    isn't really a fully vacuumed chamber any more.
     
  10. Bob Quintal

    Bob Quintal Guest

    Oh, that it would be so simple.
    However there are conditions where a higher pressure also probvides
    quenching of the arc, deflection of the electric field and cooling of
    the electrode surfaces so that arcing does not occur.


    So, it takes more voltage in a
    Actually, when 'pulling a vacuum' the possibility of arcing remains
    fairly constant between 760 Torr(standard sea level pressure) to
    about 20 Torr, then increases rapidly until about 10^-3 Torr, below
    which the possibility of arcing drops away to much less.

    I'd like to see it. Even in intergalaxy space there are some
    molecules of gas.
    That is so true.
     
  11. The breakdown voltage has a minima which occurs at less than
    atmospheric pressure for reasonable gap sizes and common gases.

    http://www.highvoltageconnection.com/images/paschen-curve_clip_image002.jpg

    The pressure at which the minima occurs is not a constant- it's
    inversely proportional to the gap size over a wide range of gap sizes.


    Best regards,
    Spehro Pefhany
     
  12. http://en.wikipedia.org/wiki/Paschen's_law

    pmin = (1/d)*exp(1-b) where b is a gas-dependent constant

    As the page says, you can get this by differentiating Paschen's law

    V = (a * d * p)/(b + ln(d*p)) wrt p, which gives you

    (a*d)/(b + ln(d*p)) - (a*d)/(b + log(d*p))^2

    and setting the derivative to zero to find a min (or max).




    Best regards,
    Spehro Pefhany
     
  13. WoolyBully

    WoolyBully Guest

    No shit, asshole.

    That what you get for the "Um".
     
  14. WoolyBully

    WoolyBully Guest

    The initial arc in a pure vacuum is hard to initiate. The moment it
    is, however, the node that was struck by the electron stream gives off
    metallic vapors.

    The circumstance of your "vacuum" and the entire chamber at that point
    is now 100% different than the starting set-up.
     
  15. Yes, the thought did occur to me. He was just slightly too obvious in
    that last troll, but just enough to work (on a weekend).


    Best regards,
    Spehro Pefhany
     
  16. WoolyBully

    WoolyBully Guest

    I worked on the development of an HV oil filtration device. It used our
    20 kV supply, and our plate arrangement and our tank chamber and oil
    pump. We PULLED the oil through the chamber and filter media as this
    reduced the pressure in the chamber and that was enough to keep arcing
    events from happening. Operated with the lid off, it would arc
    immediately. With the lid on, the pressure would pull down almost
    immediately as the oil's viscosity took some work to pull it through the
    piping. That reduced chamber pressure kept it from arcing.

    Your curve is for pressures well below that.

    In normal air, the voltage required to establish an arc rises as
    pressure drops. This is NOT data related to a vacuum. It is for reduced
    pressure environments.

    The curve plot refers to vacuums and very low pressures.
     
  17. WoolyBully

    WoolyBully Guest

    Then the tanks in our design were pulling below 1mm and were on the
    other side of the slope.

    Either way, the result was that it did not arc when under negative
    pressure, and did arc in open air.

    This refers to the tops of the plates in the filter chamber, which was
    bigger than a gallon,but smaller than a cubic foot.

    A good question would be to see what the plot is on values above sea
    level. IOW, what kind of insulator is compressed air?
     
  18. The curve I linked to (air, 0.1" gap) plots from 760 torr (which is 1
    atmosphere, of course) downward, and the breakdown voltage of at 0.1"
    gap in air decreases with dropping pressure down to about 0.2 torr,
    which is a medium vacuum. For to go up, the gap size would have to be
    < 1 micron.

    High voltage clearances should be increased for devices that have to
    remain functional at high altitudes.

    Many, many years ago I worked with a suitcase-sized system that
    contained a subminiature Van de Graff generator enclosed in a
    pressurized SF6 atmosphere. It made many hundreds of kV, IIRC.
    AFAIUI, SF6 will likely be banned if they can find a workable
    alternative- it's a very effective greenhouse gas.


    Best regards,
    Spehro Pefhany
     
  19. Fred Abse

    Fred Abse Guest

    With waveguides, that's a fucking headache.
     
  20. Jamie

    Jamie Guest

    Other wise know as plasma.
    Hmm.
    Think about FS6, roughing and defusing pumps. Of course, it's nice to
    have an ion gauge and DOM's sitting in the chamber for safety shut downs.

    Jamie
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-