[time-nuts] Frequency comparison - LPRO
lists at cq.nu
Tue Feb 9 01:32:46 UTC 2010
There's a multi turn pot and reference built into the LPRO. It's already driving the C field tuning. I assume you are leaving it in there. You should be able to get the unit to less than 5x10^-11 without a lot of crazy effort.
With any luck, the LPRO TC plus a few months drift should be sub 1x10^-10. Net, you need 2x10^-10 or so of tuning range with a ~ monthly touch up.
The normal tuning sensitivity is obviously much to large (10X) for this. The answer is actually fairly simple. Just put a nice big resistor in series with the electronic tune pin. No extra pots, references or other sources of added error. Just make sure the resistor has fairly good TC and you should be ok.
On Feb 8, 2010, at 6:29 PM, Neville Michie wrote:
> I am setting up an LPRO in a steel box (for a little additional magnetic shielding).
> The power supply is a linear in a separate box (separate the magnetic fields).
> In the LPRO box is a 5V voltage reference (2 x LT1009) which is used across a
> pair of trim pots (coarse and fine) with two front panel jacks for a high impedance
> DVM to set the frequency trim voltage f the LPRO.
> The LPRO is insulated with more than an inch of foam plastic and a small fan
> (on the outside of the box) is used to control the 1/2 inch aluminium plate that the
> LPRO is bolted to to about 38*C. The temperature control is much better than 1/10 degree.
> Inside the box is a decimal divider with outputs down to 1Hz, and a phase detector
> composed of 2 D latches (CMOS) the output of which is filtered and fed to a strip
> chart recorder or data logger.
> I have a TBOLT connected to a decimal divider, so I can compare the phase with the LPRO.
> At 10kHz I get 1microsecond per month resolution ~ 2x10^12. As errors show up
> the trim voltage can be adjusted with the DVM by an amount that is easy to calculate.
> So with a bit of time and without much in the way of high precision timing gear I hope
> to be able to get the LPRO adjusted to the limits of its precision, and I avoid the short term
> fluctuations in the TBOLT output.
> After a year or so I will be able to tell you how well it works.
> cheers, Neville Michie
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts