Experimental Modern and Quantum Physics for Do-It-Yourself Science Enthusiasts 

Facebook Flickr LinkedIn YouTube RSS
Home Instrumentation Precision Clocks and Timers Atomic Clock d.i.y. 10 MHz Atomic Clock Frequency Standard Using Surplus Rubidium Oscillator

d.i.y. 10 MHz Atomic Clock Frequency Standard Using Surplus Rubidium Oscillator

diy Rubidium Atomic Clock 10 Mhz Frequency Reference by David Prutchi Ph.D. www.diyPhysics.com

Efratom Model M-100 Rubidium Frequency Standard (RFS) oscillators are widely available in the surplus market.  Units on eBay commonly sell in the $150 to $200 range.  Despite their low surplus price, they were originally very expensive components, with superb performance.  The M100 was designed to be used by the military as a master oscillator in high-performance communication systems, frequency standard equipment, advanced navigation equipment, and all other systems which require extremely precise frequencies and time intervals.

With the proper input power provided and suitable cooling provisions, you can turn a surplus M-100 into a free-standing 10 MHz +/-5×10-11  (+/-5 x 10 ^-11 in case that your web browser doesn’t display the superscript font) frequency standard for frequency counters, as well as a precise calibration source.  I use mine to keep precise track of frequency when working on Earth-Moon-Earth (EME) communications, where even tiny errors in tuning can make the difference between success and failure to receive weak echoes.

The Efratom M100, Part Number (P/N)70502-1, is a sub-compact, lightweight, atomic resonance-controlled oscillator. The unit provides a pure and stable 10 MHz sinusoidal signal from a 10 MHz voltage-controlled crystal oscillator (VCXO), which is referenced and locked to the hyperfine transition of Rubidium 87.  The reference element is an optically-pumped integrated rubidium 87Rb vapor cell, contained within the “physics package.”

The long-term stability of the Efratom M100 is better than  6×10-11/month (< 3.6×10-10 for first year) improving to < 2×10-10/yr starting with the second year.

Block diagram of rubidium 87 frequency standard for diy atomic clock oscillator by David Prutchi, Ph.D.

The figure above shows a simplified block diagram of the rubidium standard.  A 10 MHz quartz crystal oscillator is “disciplined” to the rubidium hyperfine transition of 6 834 682 610.904 324 Hz. The amount of light from a rubidium discharge lamp that reaches a photodetector through a resonance cell will drop by about 0.1% when the rubidium vapor in the resonance cell is exposed to microwave power near the transition frequency. The crystal oscillator is stabilized to the rubidium transition by detecting the light dip while sweeping an RF synthesizer (referenced to the crystal) through the transition frequency.  When lock is achieved, the crystal oscillator’s frequency is exactly 10 MHz.


Efratom M100 d.i.y. rubidium atomic clock frequency standard by David Prutchi Ph.D.

The M100 needs to be mounted on a large heat sink.  I used one that I had in my junkbox.  I monitor temperature at the M100-heatsink interface with a Lascar model EMT-1900 thermometer LCD display.

Power and control signals to/from the M100 are all available at the connector marked J2.   This is a Winchester SGMC20 connector, and a suitable mating connector by Positronic is a 28748 type.  Pins are ordered as follows:




In my setup, 25 VDC are supplied to the M100 via J2-Lfrom an external 3 A lab power supply.  The power supply line is monitored by a Lascar model EMV-1200 4-25V range, signal-powered LCD voltmeter.  Peak current during warm-up is up to 2.2 A at 25°C with 26 VDC input.

The M100 takes up to 10 minutes of warming-up to reach 10 MHz +/- 2×10-10 at 25 °C.  PLL lock is indicated with a low-current (2 mA) LED connected between the positive supply voltage and the lock indicator pin via a 15 kΩ resistor.

The M100’s 10 MHz output signal is available at SMA connector J1.  The output characteristics are:

  • Frequency: 10 MHz Sine Wave, (+/-5×10-11)
  • Amplitude: 0.5 vrms (-10%+30%) into 50 ohm load
  • Phase Noise (SSB 1 Hz BW): >120 dB at 100 Hz from carrier
  • (Signal-to-Noise): >130 dB at 1000 Hz from carrier
  • Harmonic Distortion: >-30 dBc
  • Non-Harmonic Distortion: >-80 dBc

Complete technical information for the module can be found in the m100 rubidium oscillator manual

Finally, the following picture shows the output of  my Efratom M100-based atomic frequency standard being compared to the output signal from a d.i.y. 10 MHz GPS-disciplined oscillator:

d.i.y. rubidium atomic clock frequency standard by David Prutchi Ph.D. measured against GPS-disciplined clock

Please visit www.diyPhysics.com for other cutting-edge d.i.y. projects, and remember to check out our new d.i.y. Quantum Physics book:

 Share on Facebook Share on Twitter Share on Reddit Share on LinkedIn
15 Comments  comments 

15 Responses

  1. […] week I posted detailed construction information for my rubidium atomic clock frequency reference.  Besides that unit, I also built a GPS-disciplined 10 MHz oscillator to serve as a secondary […]

  2. Yes, but you can also do it for $40 with an FEI FE-5680A Rb oscillator from eBay. Giant heat sink not required (I mounted mine on a flat metal surface in an old rack mount enclosure). In any case you’ll be left with a pretty darn stable reference that needs to be calibrated and out of the box probably will not be dead on 10MHz. If you’re just starting out, I highly recommend picking up a surplus Trimble “Thunderbolt” GPS Disciplined Oscillator (GPSDO) from eBay. You can get a kit for about $200 which includes everything you need. The GPSDO, power supply, and antenna. With this an an oscilloscope you can than cal the Rb units.

  3. ChrisW

    The TV station I worked at had a rubidium standard for our sync generators. At some point it lost lock and was operating on the crystal reference only. One day NASA called and asked if we were having problems. It turned out that NASA was using our broadcast signal as a quick reference to check their frequency counters. When we told them we didn’t need the rubidium standard anymore and weren’t going to fix it. So they sent us an HP cesium reference so they could keep using our signal!

    • ChrisW

      The point I forgot to make was that you need to monitor the resonance lock signal to ensure you are running closed-loop.

      • admin

        Thanks Chris,
        Nice story about NASA. Thanks for sharing.
        Regarding resonance lock, that’s a good point, and maybe I should have emphasized the importance of checking the front-panel LED that indicates lock. If it’s not on, then the output is actually worse than just the open-loop crystal oscillator.

  4. […] week, David has done it again with this excellent guide to building your own atomic clock from a used rubidium oscillator: Efratom Model M-100 Rubidium […]

  5. Chris Albertson

    Nice write up but I fail to see why this is the DIY Rubidium clock. Yes it’s Rubidium but all the user does is connect wires and turn it on. By that standard I have DIY TV set in my living room because I connected some wires myself. Still these are god to know about and they are now cheap on eBay, got one for $40.

    You DO have to calibrate them which means you need another frequency standard in your lab

    • admin

      Thanks Chris,

      It always surprises me what catches blogger’s attention. My way-more sophisticated builds (which I hope you have found at least some of those in my pages) don’t make a blip on re-blogger’s radars. Yet the very simple ones – like this one – attract enormous attention.

      We have fun building projects at very different levels, so I hope that you’ll find something that interests you in one of our 3 sites: http://www.diyPhysics.com, http://www.prutchi.com, and http://www.implantable-device.com.



  6. […] with detailed advice and other projects. Had you asked me a month ago, I would have said that an atomic clock is surely beyond the capacity of even the most dedicated hobbyist. Now I know […]

  7. The heatsink, while sexy, is not only unnecessary but makes the heater work harder to keep it warm. The components inside are designed to handle the heat and the rubidium lamp needs to be hot in order to function.

  8. Simon Franklin

    The single biggest point of failure on the Rubidium F-Standards is component failure.

    Not the bulb or the heater wire but the surrounding electronics.

    After repairing more than ten I can say the DDS board is the weakest link in the system.

    The RB is designed to be mounted to an infinite backplane such as a metal cabinet or steel enclosure.

    Regards to all.