Author Topic: Symmetricom X72 (and other rubidium oscillator) signal breakout boards  (Read 329 times)

0 Members and 1 Guest are viewing this topic.

Online texaspyro

  • Frequent Contributor
  • **
  • Posts: 529
The Symmetricom X72 is a very small rubidium oscillator.   It outputs 10 MHz,  60 MHz,  1PPS, and has a programmable ACMOS level output signal.  It is usually available for a decent price.  The main problem with them is it requires an obsolete, rather unobtainium Molex connector to connect to it.  Symmetricom did offer a small board that had an edge tab that could plug into the Molex connector and broke out the signals to a 0.1" dual row header. 

A while back a bunch of the X72s with the connector board were available on Ebay, but it only broke out a few signals.  Those available now don't seem to even have that board.  Sooo.. what's a boy to do?  Well, build a better mousetrap, of course.

I built a two board solution.  The first (interface) board has the edge tab and breaks out all 26 pins of the Molex to a 0.1" header.  It also has the 10 MHz impedance matching R-C network.

The second (Breakout) board connects to the the interface board and breaks out the signals to 5 BNC connectors and has a 5V regulator and MAX232A RS232 level converter and DE9 connector.  The RS-232 connector can connect a selectable polarity RS-232 level 1PPS signal to pin 1 of the DE9, but most X72's output a 400 ns PPS pulse which might be too short to be useful unless stretched.  You can also route either the DC input or 5V regulated to pin 9 of the DE9.

The regulator has enough capacity and the header has a place for 8 more pins in order to be able to work with a future interface board for the Symmetricom SA22.c rubidium. 

Each board has a 2.1mm DC jack, but only use one of them (preferably the interface board one... it' has the shortest  connection to the X72)  The power caps are rated at 25V... should be OK to run up to 19V without changing them.  I run mine at 12V.

I have just finished an interface board for the Spectratime SRO100 rubidium and will be doing one for the LPFRS.  The SA22.c one is going to be a pain.  It uses a 2mm pitch connector mounted in the middle of the baseplate where you want to mount the unit to a heatsink... not sure of the best way to handle that.

I have shipped the first X72 batch out to those on time-nuts.  PM me if you need an X72 interface.  Price (assembled) is $27 shipped in the US,  $35 international.

« Last Edit: December 08, 2017, 08:05:52 AM by texaspyro »
 
The following users thanked this post: SoundTech-LG, Tom45

Online texaspyro

  • Frequent Contributor
  • **
  • Posts: 529
Re: Symmetricom X72 (and other rubidium oscillator) signal breakout boards
« Reply #1 on: December 09, 2017, 01:44:00 PM »
I got in the first cut of my Spectratime SRO100 adapters for the X72 BNC board from OSHPARK.   They seem to work well... except I did the boards before the SRO100 came in and they don't invert the serial data lines (the SRO100 serial lines are inverted from what a normal UART outputs).  I removed the MAX232A chip from the X72 board and jumpered across it to test the board.

I have the next rev off to DIRTYPCBs.  They have the serial data inverters and also a place for an EFC pot (driven by the SRO100 Vref line) and a pair of EFC adjustment range limiting resistors.
 

Offline cncjerry

  • Supporter
  • ****
  • Posts: 698
My boards arrived today and they work perfectly now that I have the USB serial adapter sorted out..  I was able to set the output frequency by successive tuning to 10.000 000 000 00Mhz using my GPSDO as a reference.  I will be doing further testing when my Cesium Standard gets here after Christmas.  This s a really nice, small, package for precision offline timing.  It has a very clean output signal and is very flexible.

Thanks again.

Jerry
 

Online texaspyro

  • Frequent Contributor
  • **
  • Posts: 529
It has a very clean output signal and is very flexible.

If you like your output signals a little on the raggedy side try enabling the ACMOS output at say 10 MHz.  That does a bit of a number on the peaks of the sine wave 10 MHz output.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf