Connect with us


Discussion in 'Electronic Design' started by Galsaba, Sep 23, 2003.

Scroll to continue with content
  1. Galsaba

    Galsaba Guest

    Anyone knows how to calculate MTBF and how to calculate response time?
  2. MIL-HDBK-217
  3. Reg Edwards

    Reg Edwards Guest

    If you can count, keep a record of the number N of failures you get in the
    next 12 months.

    Then, if its not a leap year, find a person you can trust, preferably
    somebody with two university degrees, willing to use his pocket calculator
    to work out the following mathematical formula -

    MTBF = 8766/N hours.

    Be very careful to press the "divide" button or you may end-up changing your
    equipment supplier.
  4. Galsaba

    Galsaba Guest

    Thanks, How can I access it. Where can I see it?
  5. Baphomet

    Baphomet Guest

    MIL-HDBK-217 used to be available (and probably still is) from the
    Government Printing Office:
  6. R.Legg

    R.Legg Guest

  7. Phil Hobbs

    Phil Hobbs Guest

    Aaugh! AAAAAAAUUUUUUGGGGGGHHHHHHH! (sob) Okay, okay, I can handle it, I can handle
    Sorry, MIL-HDBK-217 gives me FITS. And I'm not alone.

    This is great advice if the OP wants (a) to placate pointy-haired managers, or (b) to
    produce a random-number generator. The whole idea of 217 is that failures in
    electronic components are thermally activated, and so follow an Arrhenius curve with
    temperature. This isn't even remotely true. It's closer to say that silicon
    components are nearly bulletproof until you get them up to, say, 150 or 200 C, and
    then they all stop working at once. Wet electrolytic capacitors work OK until the
    electrolyte vents, and then they stop right away. That sort of thing.

    217 also predicts that adding a snubber to a relay driver will reduce its

    You're better off consulting the entrails, or waving a dead chicken over your design
    and announcing that the spirits have told you what the MTBF is going to be. It's
    just as accurate and much more satisfying.

    Honestly, I'm OK, I feel much better now.


    Phil Hobbs
  8. Active8

    Active8 Guest

    Electronic Engineer's Handbook

  9. red rover

    red rover Guest

    I recall that 217 also allowed you to use
    established reliability based on real failure
    data. So I suppose you could decide which
    is better and claim that.

  10. red rover

    red rover Guest

    Response time?

    I'm not sure if I'm taking this out of context but
    are you also calculating MTTR and need to
    know the response time of a repair person
    to implement the repair? Or is this response
    time in a different sense?

  11. Galsaba

    Galsaba Guest

    <are you also calculating MTTR and need to
    know the response time of a repair person
    to implement the repair?>
  12. red rover

    red rover Guest

    Generally you do not calcuate response time
    you are told response time (or assume response
    time). The ability to detect a fault, report a fault (alarm)
    and have it responded to. It could be hours or days if
    it is remote gear. Minutes or hours if it is at a manned

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