Budget 10MHz Frequency References

[Dragon88]

I recently started a personal blog to document various projects I'm working on. Today I wrote my second post, which is a comparison of two low-cost eBay frequency references. For the TCXO, I used the "$3 reference" posted by forum user paulie. I figured that might be interesting to some people here that are interested in his project.

Hopefully I can find time to keep this blog going. Any feedback is welcome:

http://syncchannel.blogspot.com/2015/07/budget-10mhz-references.html


The "$3 reference" thread:
https://www.eevblog.com/forum/projects/3-dollar-precision-frequency-standard/


Thanks.

[German_EE]

The thing about frequency references is that you can use anything you want (even the cheap stuff) providing it's calibrated before use. I have a 10 MHz reference oscillator that's built using quality components and double-ovened yet it still drifts one or two hertz a day so, before use, I calibrate it to DCF77 at 77,5 KHz

If the waveform drifts one cycle on the scope in one second then my error is 100 ppb

If the waveform drifts one cycle on the scope in ten seconds then my error is 10 ppb

If the waveform drifts one cycle on the scope in a hundred seconds then my error is 1 ppb, close enough, anything else is a bonus.

[Dragon88]

That's one way to do it, assuming you have access to a reference at home and are comfortable with the oscillator's short term stability immediately after an adjustment.

I think a lot of hobbyists would prefer to buy a cheap 10MHz reference for occasional use, and know it's close enough for their needs when they pull it out. Perhaps a checkup now and then will be performed when they have access to someone else's disciplined reference. Basically a middle ground between having no reference at all and spending $200+ for a disciplined setup.

[German_EE]

Checking the 10 MHz reference against DCF takes me all of ten seconds to connect two cables to the scope then look at the display. The secret here is to keep your reference powered up 24/7 and that way thermal drift is not a problem.

[AutomationGuy]

DCF is a german radio signal to distribute a common time. Probably not known outside of germany.

[blackbird]

DCF is a german radio signal to distribute a common time. Probably not known outside of germany.

DCF has a minimal range of 1500 km so it can be received across the mayor part of Europe. In The Netherlands dcf77 controlled clocks are very common. 

[cs.dk]

DCF77 clocks are also very common here in Denmark.

[Pjotr]

DCF is by far not as accurate as GPS for a 10 MHz reference. It does not account for atmospheric disturbances (and phase wandering it causes). GPS can be 10 nS accurate and you can lock to that within ppb (parts per billion) in a few hours.

[German_EE]

The mention of atmospheric fading is correct although I am in the fortunate position of living 30 Km away from the DCF77 antennas. With a small antenna I see about 30mV of signal on the spectrum analyzer so in this instance fading is not a problem. What IS a problem is a ban on external antennas which means that GPS is not an option for me unless I want to attempt reception inside (which may not be good enough for time measurement).

For those outside the reception area of DCF77 you have a choice of:

WWV and WWVH on 2,5 MHz 5 MHz 10 MHz 15 MHz 20 MHz and 25 MHz
MSF on 60 KHz (UK)
JJY on 40 KHz (Japan)
RTZ on 50 KHz (Irkutsk/Russia)
RBU on 66,66 KHz (Moscow)

All of these signals are very accurate and good enough to calibrate a 10 MHz reference.

[G0HZU]

There's also BBC Droitwich at 198kHz in the UK. This is just 20 miles away from where I live.

Most people really don't 'need' an accurate 10MHz reference and anything that can get within about 2Hz of 10MHz and hold this accuracy for a year or so is going to be good enough for the vast majority of people IMO. I suspect that 20Hz accuracy at 10MHz is probably good enough for most people doing typical hobby electronics at home.

So this means a fairly decent OCXO is going to be overkill for most users. Rb and GPSDO is way OTT. Of course. there will be a few people who can claim/justify that they need higher performance than the basic specs listed in the paragraph above.

[KJDS]

Most people really don't 'need' an accurate 10MHz reference and anything that can get within about 2Hz of 10MHz and hold this accuracy for a year or so is going to be good enough for the vast majority of people IMO. I suspect that 20Hz accuracy at 10MHz is probably good enough for most people doing typical hobby electronics at home.

I agree with that. I've also just bought a rubidium standard for the first time. Don't know if it works, but if it does and I'll do a long term check against R4LW on 198kHz then I'll be able to make sure signal generators, counters etc that I sell leave me having been tweaked back on frequency for the first time in a decade or two.

[Pjotr]

What IS a problem is a ban on external antennas which means that GPS is not an option for me unless I want to attempt reception inside (which may not be good enough for time measurement).

It is the opposite. Finding an accurate position solution is often not possible indoors. For  timekeeping 2 - 3 satellites suffice as long as there are some weak signals available of them. I have no problems with a dedicated receiver optimised for timekeeping and a patch antenna 5m away from a window. For collecting almanac, time, ephemeris, and position at start up, the antenna needs to be closer to the window.

[richard.cs]

One situation when high frequency precision (and in particular low drift) are needed is when working at microwave. Take the ham radio operator working ssb in the 47 GHz band, needing to stay within a few tens of Hertz at carrier to stay intelligable. 50 Hz would be 1 ppb, a 10 milliHz drift in a 10 MHz reference.

That's tricky but doable (short term anyway) with a good crystal oscillator at constant temperature.

[ivan747]

That's one way to do it, assuming you have access to a reference at home and are comfortable with the oscillator's short term stability immediately after an adjustment.

With GPS and an antenna, everyone does