[time-nuts] Trimble TB replacement options???

[Said Jackson]

Hi Don,

Seems the easiest would be to capture your 5370B output to a file and use Ulrich's plotter to calculate ADEV.

With ~20ps noise floor on the 5370B you can assume a 1s indication of no better than 2E-011.

The 10811A will be better than that assuming you did your efc low pass filter properly.

Also, I would disable GPS disciplining on one of the two units to prevent artificially good performance results due to both oscillators tracking similar gps errors.

Plotter lets you remove the resulting drift later on.

The 5370B may also give better error averaging if the two clocks drift slight versus each other.

Bye,
Said

On Oct 12, 2013, at 12:22, "Don Lewis" <dlewis6767 at austin.rr.com> wrote:

> 
> I just completed my second scratch GPSDO and would like to share a few of
> its details and ask a question or two at the end)
> 
> 
> 
> I think the point of the below is that it did not cost very much, either
> time or money.   It was very easily designed and built.   I think the
> GPS-locked 10MHz output frequency is pretty good .but need to better
> understand how to use (Lady Heather?) to calculate ADEVs.
> 
> 
> 
> 
> 
> I started with a couple of HP part-donor units that came my way from eBay.
> (TRULY part units -like they were missing entire pwbs and cables and such
> ...no way to easily rebuild ...but that is another story...)
> 
> 
> 
> 1.    I first built up my own discrete power supplies, needing +5VDC;
> +16VAC, and -18VDC using transformer wall warts (no switchers).   I used the
> some of the large blue filter caps from the donor HP units to help DC
> filtering.
> 
> 2.    I took an older HP10811 oscillator and its controller-board from a
> donor unit and fed the 10MHz output to an oscillator card I had from a donor
> 3585A spec analyzer (A21).
> 
> 3.    I wrapped the 10811 oscillator in several layers of cork I had around
> the house; and attached a small digital thermometer sensor to the outside of
> the oscillator, prior to wrapping.  It seems to maintain a constant 100F
> with little effort or regard to room temperature.
> 
> 4.    The standalone 3585A, A21 board takes the 10MHz from the 10811
> oscillator and locks it to its own, discrete, on-board 90MHz oscillator on
> the A21 board.  Then it nicely divides this 90MHz down several times
> (frequencies the 3585A spec analyzer originally needed).
> 
> 5.    I get outputs from the A21 board:  90MHz, 10MHz, and100KHz.  (lots of
> ECL on this little A21 board)
> 
> 6.    I then added my own simple divide by 10 to take the 100 KHz to 10 KHz.
> 
> 7.    I used an (eBay) NAVMAN-Motorola GPS unit that has a 10 KHz output to
> feed an XO; this XO feeds a few RC discretes that develop the analog EFC
> back to the HP10811 oscillator.  (Adjusting the 10811 oscillator to such an
> initial frequency that allows the lock EFC voltage to be positive is
> necessary.
> 
> 8.    Letting all this run overnight is showing the HP1011, 10MHz oscillator
> locked to GPS that has a TI of 100ns with all zeros to the right of the
> decimal point, . out to decimal places eight & nine on my good 5370B TIC,
> .(decimal places eight and nine fluctuating around +- 20)
> 
> 
> 
> Btw .I use my first GPSDO (similar, with NAVMAN/Motorola, but no fancy A21
> board .etc) .as the 10 MHz external reference to the 5370B TIC.
> 
> 
> 
> Question, please, how do I get this GPSDO (input) into Lady Heather?   I do
> have an RS232 interface into my computer for the NAVMAN, and can monitor the
> GPS with a simple GPS monitor program.but this is not the locked,
> disciplined frequency from the 10MHz HP10811.
> 
> 
> 
> With my setup .how does one compute the ADEVs?
> 
> 
> 
> -Don Lewis
> 
> Austin, TX (Hyde Park)
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> 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.