[time-nuts] What type of Crystal?

Ed Palmer ed_palmer at sasktel.net
Sun May 31 16:20:34 EDT 2009


   Magnus Danielson wrote:

     Ed Palmer skrev:

        The recent discussion regarding the type of crystal in the HP
     10544A
        brought this question to mind.  We're always coming across
     unknown
        oscillators.  Usually we can figure out the pinouts and
     voltages.  Then
        we can measure stability, aging, etc.  But are there any tricks
     to
        figure out what type of crystal is in the oscillator?  How can
     you
        detect the differences between AT, BT, SC, etc?

     One thing which may be a hint is to look at what frequency they have
     cold, the detuning they have at room temperature is quite a good
     hint. This works best for OCXOs, since TCXOs at these frequencies
     usually is AT cut.

   Yes, I should have specified that I was talking about OCXOs.  Since a
   TCXOs purpose is to compensate for temperature changes, the concept of
   'warmup' is a bit of an oxymoron.

        I think that AT crystals have a broader tuning range than SC and
     that
        when warming up AT crystals tend to overshoot the final frequency
     and
        fall back.  Are these generalizations correct?  Are there other
     tricks
        to help differentiate the crystal types?

     The overshot by itself may not be a good indicator. An SC with wrong
     temperature may exhibit overshot as well.

   A defective oven controller could certainly confuse any attempt to
   characterize an oscillator.  Let's assume that - as far as we can tell
   - the oscillator is working properly.

     SC cut 10 MHz seems to be about 200 Hz low at room temperature.
     Don't recall the number for AT cut, but I think I saw something like
     1 kHz or so recently. Need to test to be sure.

   I happened to record the startup performance of an HP 10544A.  It
   started out ~1100 Hz low.  I was initially worried that it was
   defective, but it was fine once it warmed up.  It also appeared that
   the amount of frequency overshoot was dependent on the oven voltage.  I
   want to investigate that more on various oscillators.
   Ed



More information about the time-nuts mailing list