[time-nuts] Better than Quartz?
magnus at rubidium.dyndns.org
Sun Apr 27 20:43:33 EDT 2008
From: "Rick Karlquist" <richard at karlquist.com>
Subject: Re: [time-nuts] Better than Quartz?
Date: Sun, 27 Apr 2008 17:21:58 -0700 (PDT)
Message-ID: <15220.127.116.11.178.1209342118.squirrel at webmail.sonic.net>
> Any decent crystal oscillator circuit will have a loaded Q that is
> a large percentage of the unloaded Q. The bridge circuit in the
> E1938A does not enhanced the loaded Q, compared to, for example,
> the 10811. It doesn't seriously decrease the loaded Q either, but
> in any event, Q is not the raison d'tre of the bridge circuit.
> What the bridge circuit does is make the oscillator circuit (as
> distinct from the crystal) very stable. If you don't have a bridge
> (ie any oscillator other than the E1938A), then the loaded Q of
> the crystal will affect how much the oscillator circuit can pull
> the crystal. In this sense it could affect stability. But if
> the circuit is sufficiently ovenized, and run on stable power
> supplies, etc., then the stability of the oscillator will be determined by
> the intrinsic stability of the crystal.
> There is no direct correlation between the Q of the crystal
> and its intrinsic stability. It is fairly easy to make a
> "junk" crystal with a Q of many 100,000's at 10 MHz. This
> Q is at least within an order of magnitude of the theoretical
> QF limit of quartz. Yet the stability is many orders of magnitude
> removed from a precision crystal.
I think your argument about stability needs to be clarified in terms of short
term and long term stability. Correct me if I am wrong, but short term, i.e.
jitter is fairly dependent on crystal Q value, among other things naturally,
such as drive level and oscillator noise.
Long term stability depends more on factors like environment etc.
> AFAIK, the loaded Q of exotic crystals like BVA is not significantly
> higher than other good crystals, since they are already close
> to the theoretical QF limit.
The BVA configuration adds a little bit of serial capacitance, as a side-effect
as I remember it. I could dig up the Q value and other measures on my BVA
(OSA 8600) if you like.
> There was an interesting paper given at FCS about 10 years ago
> about the quartz shortage during WWII. There were govt projects
> to search for a replacement material. It sounded like a good
> effort was made, but in the end nothing else even came close
> to quartz.
It is only more recent work in esoteric materials, configurations and
temperatures that provided competing Qs, but then for most part in a totally
different frequency range.
More information about the time-nuts