Connect with us

Power rating of SMA connections?

Discussion in 'Electronic Design' started by Joerg, Jul 8, 2011.

Scroll to continue with content
  1. Joerg

    Joerg Guest

    Gents,

    SMA and lots of other connector technologies are rather lacking in
    specs. For example there are no currents or power levels at 50ohms
    listed. How much can they take at <30MHz?

    I found a 40W-rated product with SMA here but wanted to ask what this
    distinguished circle of newsgroupers thinks:

    http://www.instockwireless.com/power_divider_pd1120.htm
     
  2. Joerg

    Joerg Guest

  3. Joerg

    Joerg Guest

    That became blatantly clear early on in that article :)

    True. I just need to know where on the crappiness scale SMA connectors
    are. Whether they can safely carry 50W in the shortwave band.
     
  4. Joerg

    Joerg Guest

    That would be about an amp, and at HF where there's not much in
    dielectric dissipation. Looks doable but I'd like to know for sure from
    some manufacturer. Seems I'll have to write <sigh>. It's so tough to get
    an answer beyond the datasheet these days because often the real gurus
    have been laid off, retired or passed away :-(
     
  5. Joerg

    Joerg Guest

    Thanks. I had found some MIL versions like this but they are very
    expensive. I have written to AMP about their regular $3 series, will let
    the group know when they answer.
     
  6. Joerg

    Joerg Guest

  7. Joerg

    Joerg Guest

    Well, we will have a very fast bridge and a controller that shuts it
    down in the case the VSWR goes loco.
     
  8. Joerg

    Joerg Guest

    For plasma generation it may be better to design your own amps. Then you
    know the limits. Many things can be protected by artifical rails and
    diodes, things you won't find in the typical COTS RF amplifier.

    If it's really bad and it is a one-off experimental setup use a tube
    amp. Tubes are very forgiving, even the occasional blue hiss between
    plate and other electrodes can be survived if the amp is designed right.
    Same for current. I've had plates glow bright orange and the tubes
    survived without noticeable performance loss.
     
  9. Joerg

    Joerg Guest

    Phil Hobbs wrote:

    [...]
    The one with the secret stash of vodka under the front cover?

    :)
     
  10. Joerg

    Joerg Guest

    Gents,

    Here is the promised feedback. Got an answer from Tyco this morning.
    Peak power was stated as 20kW at sea level. Yes, that's kilowatts, had
    to look twice. Voltage breakdown is given as the limiting factor and
    peak power must be derated by a factor of 20 for "critical altitude"
    around 70,000ft.

    Another document they sent states a 3mOhm max test requirement for
    center and outer contact. Now that's very encouraging.
     
  11. Take into account that SMA is not made for infinitely
    many mounting cycles...
    So whatever the manufacturer specifies for a new product,
    the contact degrades over the cycles.

    Rene
     
  12. You could reasonably assume that most have more gold plating on them
    than a CAT 6 cable or connector, and those last through a LOT of
    connect/disconnect cyclings.

    I doubt it would change enough to require more than normal
    consideration. The per connection dB loss in a given path is likely far
    greater a worry than how many times a given cable was used or a
    connector, for that matter to comprise one such connection. The loss
    would stay pretty flat right up until the failure mode. (isn't that
    always how it is? :)
     
  13. There is a standard for the SMA interconnection. I also often see
    various sized coaxes (all 50 ohm) between elements which usually
    correlate directly to how much power is being pushed (or clamped in the
    case of termination). So the cable characteristics and raw connection
    count are likely more important than the connector or connection
    loss-over-number-of-cycles factor.
     
  14. On a connector, perhaps. But the cables that interlink things would be
    where the power bottlenecks are.

    Did you look at that log amp I sent you the link to?
     
  15. The center conductor (nib) of an SMA (M) connector only "inserts" about
    1mm into the center hole on the female. They are held mated by the
    threaded portion of the connector. There is little gap, even between the
    insulator sections.

    So there is no big long sleeve of captivation. It essentially is the
    tangency between the radius of the nose of the nib, and the split center
    hole, both being gold plated.

    All it takes is ONE loose connector. Remember to torque SMAs properly
    to 7 to 12 inch Lbs.

    Some of our sat gear has huge N type connectors and an SMA on the other
    end, but I don't think they are passing high power over them. They are
    more for signal or noise injection, etc. for test regimens.
     
  16. Joerg

    Joerg Guest

    Yes, Rene, but it's no problem here because this is for internal
    modules. So typically there will be only one mounting process in the
    lifetime of the unit. But that lifetime can be very long.
     
  17. Gold will outlast ANYTHING you call a lifetime.

    The album Carl Sagan put on the space probe is Gold plated Copper.
    It will NEVER oxidize, even after a BILLION years.
     
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

-