[time-nuts] 100 MHz Source
magnus at rubidium.dyndns.org
Tue Apr 29 18:11:36 EDT 2008
From: Bruce Griffiths <bruce.griffiths at xtra.co.nz>
Subject: Re: [time-nuts] 100 MHz Source
Date: Wed, 30 Apr 2008 09:24:17 +1200
Message-ID: <48179201.1000601 at xtra.co.nz>
> michael taylor wrote:
> > On Tue, Apr 29, 2008 at 12:46 PM, Richard W. Solomon
> > <w1ksz at earthlink.net> wrote:
> >> I need a 100 MHz source to lock up a F/W brick. I have 10 MHz available
> >> from my GPSDO. Anyone know an easy way to get 100 MHz using the 10 MHz
> >> source ? Maybe another PLO ?
> > Run a 100 MHz oscillator in a divide by 10 PLL (Phase Locked Loop)
> > configuration, compared to your 10 MHz standard? If the Motorola /
> > Freescale MC145170 (-2) was still in production, that might be a good
> > one to use. Maybe National LMK03002.
> > Or a PLL synthesizers that takes the 10 MHz as its input clock and
> > synthesizes an 100 MHz output clock.
> > An example might be, Analog Devices ADF4001
> > <http://www.analog.com/en/prod/0,,770_850_ADF4001%2C00.html>.
> You will achieve a much lower phase noise if you use low noise
> multipliers rather than using noisy dividers.
> Unless of course you use regenerative dividers or an injection locked
A regenerative divider tuned for 90 MHz and 10 MHz would do the trick.
Should be fairly simple to acheive. Some care in filter Qs and it should be
fairly stable in phase.
> An injection locked multiplier is another option.
Infact, in a PLL case you could bootstrap the regenerative divider with the
reference 10 MHz and then pull it of as the 90 MHz gets strong enought. But it
should not be needed. A straight regenerator should wake up quickly enought
anyways. A side-consequence of a bootstrapped variant is that the generated
10 MHz would quite likely to be within +/- 18 degrees from the reference,
which could be a useable startup condition for any PLL lockup.
More information about the time-nuts