• Print
• .

AtomicClocks

07/20/25 (ca, pf)

• continue testing SA.22c
• to save the o-setting, enter:

 t5987717<cr>

• after reboot, frequency should be saved.
• to fine tune frequency use the f command, an absolute value compared to the boot-time frequency
• ??? not sure the f values are saved and becomes the new f=0, need testing

06/04/2025 (pf)

• testing SA22.c frequency selection:

 >> n=[1:30];  
 >> f./n';     
 >> [n' ans]   

 ans =

          1.00   30000000.00
          2.00   15000000.00
          3.00   10000000.00
          4.00    7500000.00
          5.00    6000000.00
          6.00    5000000.00
          7.00    4285714.29
          8.00    3750000.00
          9.00    3333333.33
         10.00    3000000.00
         11.00    2727272.73
         12.00    2500000.00
         13.00    2307692.31
         14.00    2142857.14
         15.00    2000000.00
         16.00    1875000.00
         17.00    1764705.88
         18.00    1666666.67
         19.00    1578947.37
         20.00    1500000.00
         21.00    1428571.43
         22.00    1363636.36
         23.00    1304347.83
         24.00    1250000.00
         25.00    1200000.00
         26.00    1153846.15
         27.00    1111111.11
         28.00    1071428.57
         29.00    1034482.76
         30.00    1000000.00

on-line chats re. SA.22c

https://time-nuts.febo.narkive.com/5iGo52H7/cheap-9-8mhz-sa-22c-s

There is a crystal in the sa.22 that will determin what output frequencies are available for output. On the 9.8MHz and 10MHz versions it is 6X the nominal output frequency (58.x MHz vs 60MHz). The sa.22 can be programmed to output xtal freq/2N for n up 2^16 (IIRC, it's in the manual).

Thus, you can change 15MHz units (found on telecom boards) to 10MHz (n is changed from 2 to 3). I have done this. But, you can't change 9.8MHz (found on different telecom cards) units to 10MHz.

Do you think this is a hardware limitation or because they're running the 6.01 code rather than 6.05 hence no g or y commands? I presume there's no hope of getting the 6.05 code which is assumed by the manual.

There's probably no way to get to 10 MHz. I have two of these that I picked up just to play with. Another purchaser contacted Symmetricom and actually got some information out of them. The oscillator in these runs at 58.9824 MHz. The default output is 9.8304 MHz. The output can be changed, but only in submultiples of the 58.98 MHz. e.g. 58.98 divided by 6 equals 9.8304. You can change the '6' to various integers as documented in the manual.

To move the output to 10 MHz you'd have to change the crystal back to the standard 60 MHz so that you could divide it by 6. But then you'd mess up the multiplier that takes the crystal frequency up to the Rb frequency. You'd have to change the synthesizer that controls that side of the system. There's no way to tell if that would require a completely new firmware load or if, just maybe, there could be an undocumented (maybe protected) command to modify it. You might be able to send it back to Symmetricom to get it changed but, based on my experience, you couldn't afford it.

If you start with your 9.830400MHz Rb and then divide that by 300 you get 32768Hz.

32768Hz is what is used in watches and clocks and PC mainboards for time of day clocks. So the maybe the best use of the Rb is to make a really good wall clock. Use it to replace a cheap watch timing crystal.

The 9.8304 MHz frequency is related to the common RS-232 modem frequencies. It is also related to the common sound card frequency of 24.576 MHz. There it is possible to phase lock it to 10 MHz.

The data sheet is confusing. These do indeed have a 58.9824 MHz oscillator and CANNOT be changed to 10 MHz. They can be changed to divide by N >= 2 to include 29.4912 MHz, 9.8304 MHz, and even 32.768 kHz if you wish, as has been said, to have a very accurate watch or RTC.

They do NOT include the PPS disciplining function. The PPS output can be enabled, but I have not figured out a good way to synchronize it. PPS output does not change when the divide by N is changed.