Long term aging control of a FRK
Add Reply
ew via time-nuts
2018-11-20 10:46:03 UTC
     Attached is a picture of my original FRK late eighties. First with heat sink, upgraded to fan control and finally a 12 bit Phibrick D/A controlled by a 3 MHz Collins TCXO. All the time monitored by a Tracor 599 modified for 60 KHz.  I was the computer calculating the aging and by that time the 10 turn was changing the 3 MHz TCXO. Result was 3.4 E-13 with my first Austron Loran C.


We are experimenting extensively with options on the best possible reference frequency resolution. We have Cs. Recent work has resulted in significant improvement with FRK best performing with 0.1 sec. loop time. See attached. The FRK had not settled jet. Needs a month. We also have aging of M 100 over years and are confident that we have better than 2 E-12 per month.

We have also worked with Corby on pressure compensation and have a board with two temperature and one pressure sensor. Also included is a LTC1655 for GPSDO work covering 24 hour filter time.

Now I want to go back twenty plus ears and repeat aging compensation the old fashion way. What I need is help with one or two PICs. One will include a 16 bit down counter and SPI interface to the 1655. Able to set the 16 bit to roughly 90 % with approximate 15 updates a day it will cover 10 Years. Maybe a second setting for adjusting the analog part.To get 15 updates we need to divide 3 MHz with a 34 bit binary divider.

Goal is 1 E-12 frequency.

Part of the project is also a dedicated counter with Ublox so no test equipment is tied up. Like in the old days look at the data once a month and fine tune the TCXO. It replaces the Tracor 599H with its mechanical counter. Had they not changed the WWVB modulation I still would be using it.                It will take month till we have the right analog steps out of the DAC.

If some one is willing to help with the PIC's, please contact me off list.
Bert Kehren