20 Pounds And A Gut Feeling Yields A Configurable Rubidium Atomic Clock Source


So you see an image like this and the description “Aircraft stable oscillator” on an eBay listing for twenty pounds (about thirty bucks), what do you do? If you’re [Alecjw] you buy the thing and crack it open to find an atomic clock source inside. But he really went the distance with this one and figured out how to reconfigure the source from the way it was set up in the factory.

First off, the fact that it’s made for the aerospace industry means that the craftsmanship on it is simply fantastic. The enclosure is machined aluminum and all of the components are glued or otherwise attached to the boards to help them stand up to the high-vibrations often experienced on a plane. After quite a bit of disassembly [Alec] gets down to a black box which is labeled “Rubidium Frequency Standard”… jackpot! He had been hoping for a 10 MHz signal to use with his test equipment but when he hooked it up the source was putting out 800 kHz. With a bit more investigation he figured out how to reconfigure the support electronics to get that 10 Mhz source. We think you’re going to love reading about how he used a test crystal during the reconfiguration step.

Once he knew what he had he returned to the eBay seller and cleared out the rest of his stock.

[Thanks DIY DSP]

62 thoughts on “20 Pounds And A Gut Feeling Yields A Configurable Rubidium Atomic Clock Source

  1. Very nice! i have an $150 ebay sourced 10mhz GPS controlled rubidium clock standard i put in a nice fancy case and used a few video distro cards to shove them out on to my equiptment

    tho i dont think you can use them as a standard if it outputs a square wave …

    1. Allow me to explain myself here… I bought a mysterious part with the risk of ending up with something completely useless. I then spent a lot of my time documenting it, adding value to it for anyone who buys one. Hopefully this doesn’t seem too unreasonable! Besides, I’m at college at the moment, so I have an education to pay for!

      1. Are you going to fool us by saying you “worked hard” on this hobby to find a fcking datasheet when anyone with scope and a screwdriver would do the same?

        I have no words. Take a job to pay bills, not hacking opportunities. You do not deserve to be mentionned on any interesting blog.

        Please fellow hackers, do not sponsor college fees for such a mron.
        There are tons of rubidium standards for EUR 60-70 that outputs 10 MHz out of the box, and are much smaller.

        BTW, is M0TEI a ham call?

        1. what the fuck is wrong with you people? all over the comments on the article i see retards like yourself flaming the guy for making a bit of profit off of a project, are you really so fucking stuck up that you just have to make these nasty comments on articles about it?

  2. Good read!

    I wonder if he could check it against the NIST atomic time standard by injecting his clock into a shortwave radio tuned to WWV at 10MHz and looking for heterodyne beats. As I recall, WWV’s frequency is accurate to better than 10^-12

    1. its always nice to have references in your lab … a calibrated voltage, current, resistance exc reference to make sure your stuff is up to cal … a frequency standard is no different

      tho the benefits of a 10mhz reference is most good equipment will take an external 10mhz clock to not just keep all your equipment synchronized but you know quite damn well that the frequency you are measuring is DEAD on accurate

      tho it has its own wacky uses .. put it on a watch and have the worlds most accurate watch, use it on an atmel and have an AVR with the frequency drift of a satellite, use it to heat your room or you can throw it at intruders!

        1. Um, since you want to be an ass, you wrote the abbreviation for millihertz. the correct answer is MHz, not that it really mattered. we all knew what he meant.

          People who live in glass houses shouldn’t be dicks.

  3. If I remember correctly the radioactive decay products poison the accuracy or increase the phase noise of atomic clocks over time. Might want to check how old those units are as that may be the reason why they are for sale so cheap.

    1. But even if that is true is the decay in accuracy such that it would be a big deal? I mean the accuracy the military demands (not needs maybe, but demands) is quite different than what the average hacker would be happy with.

        1. I’m more interested in how they us Rb as a clock source. The common isotope has a halflife of 49 billion years. The decay product isn’t a problem, it’s a stable 87Sr.

          Okay, read up on the way the clock is used. It’s not detected the decay, but like quartz it’s detecting the energy given off by 87Rb when exposed to a known frequency. The long HL means there should be very little decay product that needs to be accounted for, and it probably doesn’t respond at the same frequency. Anyone want to find out how much energy is generated by 87Sr at 6 834 682 610.904 324 Hz?

          And WOW, if I read that right, that source should be accurate up to a awfully high MHz, even 1 or 2 GHz, reference

    2. Apart from this phase noise, remember that the product presented here produces a signal via a DDS followed by a crystal quartz. A DDS is not really a good frequency reference.

      A trimble GPSDO would probably generate a better short term signal with a better long term stability.

  4. It was probably use don conjunction with a GPS. It’s best to use both as GPS accuracy can waiver depending on your location and the time-of-day. That way you can maintain extremely high accuracy without any variance.

        1. I was in the military and we used time standards that used both Rb and GPS. GPS has issues, some listed in other comments, but some were not known at the time of implementation. So, at certain times of day in combination of our position on the planet, our accuracy varied. Not wildly, but, it varied. This is the reason why the DOT installed WAAS which uses two land based DGPS stations (E/W coast) and a satellite that’s right above MO. This is actually more accurate for air travel and other related commerce. Also, the US Coastguard has installed DGPS stations all over the east and west coast. The reaosn is that even with GPS “unlocked” it is still not accurate enought to find someone in the event of an emergency. You throw in DGPS stations and you can locate someone in coastal waters, with a commercial GPS unit, within several meters.

          Some things that prompted the UGS and DOT and WAAS wa sthat originaly the L1 GPS band was “spiked” by the military to reduce accuracy. The military stopped this after DGPS stations were abound as it made no sense anymore. But, the L2 is still an encrypted signal and this provides additional accuracy. But, once again, with my own personel military experience, the accuracy would drift and this is actually important as you can cross into the territorial waters of some other country. So, we had to compensate and be furthur out. Also important for GPS guided bombs, there is a “window” of time where the GPS they use is the most accurate. I’ll leave it at that.

          You can use Google and look up atmospheric scintillation and GPS do find more info on GPS accuracy issues with the expansion and contraction of the atmosphere.

      1. I think if they can compensate for General and Special Relativity (which they do) they can compensate for Doppler if that was even a problem… Which on the speed an airplane is moving at relative to the satellites is negligible. If it is necessary, I’m sure they would add a dash of math and be in business.

        I am rather curious about the ringing on the square wave, seems like an awful lot for a frequency standard. If it is, perhaps this is why it’s retired?

        1. This may have been a post from Reddit on some point, but Mike didn’t get it from there. He got it from me. I sent it to him on Sunday after my friend Kerry sent it to me. You can see he even thanks me in this article for it. I don’t know where Kerry got it from, but I don’t think he reads Reddit.

          How about instead of critiquing Mike, thank the lucky stars for a cool sites like this, or better yet go and make some material of your own! :)

  5. Nice STALO, but I bet as others have said, its probably on ebay for cheap for a reason… I would just hope that it’s for another component failure and not the Rb source itself.

    1. Oh, and the ringing may be due to his hand held scope and probe, while being a nice fluke, is still not a $50k LeCroy or anything. And looking up that NSN brings it back to being part of a “Nimrod MR2 Aircraft XV236” of the RAF. Nimrod, really? Anyway, that is one ugly airplane!

      1. http://en.wikipedia.org/wiki/Nimrod “Nimrod”, via Elgar’s _Enigma Variations_, which is exactly the sort of classical music that 50s/60s chief RAF staff would know.

        http://en.wikipedia.org/wiki/BAE_Systems_Nimrod_MRA4 is of more interest, as it’s several billion pounds of “hacking”; trying to use antique airframes from the dawn of the jet age with modern radar. This didn’t go well and was eventually cancelled with extreme prejudice.

      2. eatith,
        I think the ringing is just because of the probe! I love my Fluke 123 – it’s by far my most reguarly used piece of test equipment since I bought it a few months ago. It came with a shielded probe but I couldn’t find it at the time I was doing this, so i just used a standard multimeter probe.

    1. Thanks, Sherlock.

      Quote from the source:

      “After I found one of these inside the Mysterious Aircraft Part, i decided to buy the seller’s remaining stock (another 6 of them)”

        1. Everyone had their opportunity, its an open marketplace, anyone could have got there before him. He achieved a little something here; not big business, just a chap with skills and who got lucky. You’re just making yourself look like a right jealous so and so and frankly quite stupid.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.