Connect with us

Can this be done?

Discussion in 'Electronic Design' started by Phil Hobbs, Apr 2, 2005.

Scroll to continue with content
  1. Phil Hobbs

    Phil Hobbs Guest

    Dear All:

    I have a possibly very interesting instrument application that requires the
    application of a very large ac voltage, about 3 kV p-p anywhere between 20
    and 50 GHz--it doesn't much matter where.

    The actual power dissipated in the instrument is quite small, so I don't want
    to use a radar transmitter to drive it. Pulsed operation is fine--duty
    cycles of at least 1% are needed, though.

    Waveguides and striplines for these frequencies are quite small, which leads
    me to worry about corona, rf arcs, and so on.

    I don't mind using vacuum if necessary.

    Any ideas for less than about $20k in parts?


    Phil Hobbs
  2. James Meyer

    James Meyer Guest

    I *was* going to tell you *exactly* how to do it for about 100 dollars
    but I changed my mind when you posted *exactly* the same message *three* times.

  3. Jim Thompson

    Jim Thompson Guest

    Getting crabby in our old age are we ?:)

    Phil is probably cursed with Outhouse Excuse.

    ...Jim Thompson
  4. John  Larkin

    John Larkin Guest

    This would have to be inside a cavity or a waveguide; if a signal like
    this existed in free space, it would radiate kilowatts. And if it was
    a microstrip or stripline, it would probably fry any reasonable

    The Cebaf (now Jefferson Labs) electron accelerator used
    superconductive niobium cavities, each pumped by a fairly small
    klystron, to get megavolts/meter fields at something like 3 GHz.

    Sounds like magnetron territory to me, not difficult at 1% duty cycle
    if you can use a field inside a cavity or waveguide. Who is that
    company that has scads of old radar gear and antennas and stuff? Radio
    Research or something.

  5. Bob

    Bob Guest

    I'll give ya about 200 dollars if you'll tell me, and, I'll only ask once.

  6. Pooh Bear

    Pooh Bear Guest

    That's possible but I've had to uneccesarily resend messages ( not using OE ) when
    the news server doesn't acknowledge the message sent. Probably happens when the news
    server's a bit too busy.

  7. Phil Hobbs

    Phil Hobbs Guest

    Nah, no such excuse, I use Mozilla--it turned out to be a SMTP relay problem.
    I was sending a bcc: to my email account, which the relay complained about,
    but silently sent the NG posting along each time.

    I deeply apologize that I have inexcusably reduced the wisdom-to-noise ratio
    of this august NG so much--we almost never get repetitive messages here. ;)

    Doing this deep in the guts of a magnetron is a good idea, though it might be
    hard to get the resulting soft X-rays out of the cavity. (The X-rays are
    what I'm interested in--details must be left vague at present.) How big a
    voltage swing do you think I can get in, e.g., an evacuated waveguide
    resonator before it turns to lava?


    Phil Hobbs
  8. keith

    keith Guest

    Nah, he's not even running WinBlows (OS/2). From his headers:

    User-Agent: Mozilla/5.0 (OS/2; U; Warp 4.5; en-US; rv:1.7) Gecko/20040617
  9. Mac

    Mac Guest

    What is a soft X-ray? What is the relationship between these soft X-rays
    and the 30 GHz (or whatever) radiation you mentioned in the original post?

    I always thought that X-rays were gamma rays with frequencies far beyond
    the optical. 30 GHz is far below the optical, obviously. No doubt there is
    something I'm missing, possibly due to my ignorance concerning magnetrons.
    -- Mac
  10. Guest

    How soft is soft? Would a beryllium, boron or diamond window be
    transparent enough?

    The 30GHz radiation is probably exciting a plasma, and the X-rays could
    come from free electrons encountering positive ions, and dropping into
    an orbit close to the nucleus, though they could also arise from
    energetic electrons hitting the walls of the chamber.

    The volume may be "evacuated" but it isn't empty, and the residual gas
    molecules are the source of the plasma. Try searching on "electron
    cyclotron resonance" for a bit more detail on what might be going on.
  11. Phil Hobbs

    Phil Hobbs Guest

    I'd love to be more specific, but it's a partly-baked idea at
    present--it doesn't work, I'll post it so we can all have a good laugh,
    and if it does work, you can read about it in Nature. Odds are probably
    60% laugh, 20% inconclusive, 20% important. It's a way to make a bright
    soft X-ray source fit on a tabletop, instead of requiring a synchrotron.
    Early simulations are encouraging, but they rather require this big E
    field at high frequency....

    So how big a signal can a 30-GHz waveguide (say) take before it starts
    arcing in vacuum? Probably it'll be the wall dividing the resonator
    from the feeder that will melt first, being thinner. My machine shop
    will love being asked for a platinum-plated tungsten waveguide...


    Phil Hobbs
  12. Mark

    Mark Guest


    I think I would start with a "conventional" high power Tx then think
    about a cavity or waveguide based impedance transformer (like a 1/4
    wave line) that steps up the voltage (and down the current).

  13. John  Larkin

    John Larkin Guest

    OK, Pozar page 201. At 10 GHz, a typical air-filled waveguide can
    handle about 2.5 MW, e-field peaking around 3e6 v/m. Power capacity is
    inverse with frequency, but max e-field stays the same (it's just the
    air breakdown limit.) He recommends a 2:1 safety factor from that. One
    can improve things by pressurizing with dry gas, or filling with SF6
    at atmospheric pressure. The medical xray people who make
    microwave-pumped linacs like SF6 so they can make their exit windows
    very thin.

    Vacuum is interesting; look up the Farnsworth Multipactor effect. It's
    caused some serious problems in satellites.

    You will do more than publish a paper if this works; the lithography
    guys will be all over you. I can introduce you to some folks at Cymer.

  14. Phil Hobbs

    Phil Hobbs Guest

    Brilliant, thanks.

    The litho application had crossed our minds. I'm not sure we can go
    quite that long in wavelength (litho wants ~100 eV, I think, whereas
    we'll probably start out at around 1 keV), but we're certainly going to
    try, once we get it working at all.

    I'll go calculate some waveguide septa. Tally-ho.


    Phil Hobbs
  15. John  Larkin

    John Larkin Guest

    Seen this? Not the same, of course, but weird and interesting.

    "The use of the COOL-X in practical applications will challenge the
    user's imagination!"

    That's funny!

  16. Calculate the current needed to drive ANY amount of capacitance at all
    at 50 GHz, and you will find that the power needed is insanely enormous.

    UNLESS you can use some sort of precise tuning or matching.

    The problem sounds extremely ugly.

    Many thanks,

    Don Lancaster
    Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
    voice: (928)428-4073 email:

    Please visit my GURU's LAIR web site at
  17. John  Larkin

    John Larkin Guest

    That's not a huge field inside a waveguide or a resonant cavity.

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