WWV receiver for ntpd time source

[skeezix]

I've been busy with ntpd as of late and decided that I need to have my own stratum 1 servers that sync to WWV (5, 10, and 15 MHz, optionally 2.5, 20, and 25 MHz).

Have two paths:

In one case, have one computer hooked to a receiver on 5 MHz and another computer connected to a receiver on 15MHz.  Both would run ntpd that can decode the audio output. Their success would come & go throughout the day, but they would be complimentary and any client stations would get at least one of them.

Alternatively, have a one computer and one computer controlled receiver that would be commanded to switch freqs depending on the time of day.


Then got to reading about other stratum 1 servers that use GPS, GNSS, and CDMA. Saw an article that I found that had a fascinating tidbit in it:
http://www.hfindustry.com/meetings_presentations/presentation_materials/2009_feb_hfia/Presentations/ShadNygren_HFIA2009.pdf

Specifically, from page 3 of the document:
• Use difference from local CDMA 1PPS UTC to calculate path length from WWV to Datron
• Use path length and great circle distance to calculate Ionosphere Virtual Height

The document it references from 2008:
http://www.hfindustry.com/meetings_presentations/presentation_materials/2008_feb_hfia/presentations/08_shad_nygren_hfia_presentation_feb2008.pdf


Right now, going to start with the stratum 1 server(s), then eventually run that experiment as I get more information about it.

Any recommendations on a WWV radio? I have a Yaesu FT-840 that I'll probably use in the near-term to get the thing going, but don't want to tie that radio up for too long.

[ChrisSmolinski]

Would WWVB work as well / better?

[Fried Chickin']

This is a cool idea. I just got an "atomic" watch a couple weeks ago and was wondering about this sort of thing. Keep us updated!

[skeezix]

WWVB would have to be handled differently. ntpd's WWV/H support is via audio and its specific format. WWVB is just different enough to not do it as is. I'd like to have WWVB as well, but that one will wait until WWV is going and after the data is collecting for the WWV ionosphere measurements.

However, someone from 2003 thinks it wouldn't be that hard:
http://lists.ntp.org/pipermail/questions/2003-November/001357.html

"The signal is now very similar to the WWV/H demodulator/decoder and
 the same algorithms can be used. All you would have to do is change the
 seconds state machine table."

"Here's a way to test the idea without building anything. Several
shortwave receivers today can tune 60 kHz with 1-Hz resolution. Wind an
antenna on a ferrite rode maybe with a preamp. Radios I know about have
atrocious low gain at and below the broadcast band. Set the BFO to
produce 100-Hz note and connect your soundcard. Rip off the WWV/H
driver, toss out the 1000-Hz stuff and synchronize directly to the
subcarrier. Weekend project."

Maybe, maybe not. Perhaps some day I'll look at the WWV/H audio source code in ntpd and see how easy it would be to alter it. Any serious mods would probably beyond what I can do.

One of the problems today with WWVB is that there used to be a little receiver called the CMMR-6 that apparently worked fine. However, its been long since discontinued.

Perhaps getting a WWVB radio that supports IRIG audio. But I have a feeling that means $$$.

For even more fun, ntpd also supports demodulating CHU without any more fuss than WWV/H.

Wonder if there's a way to integrate an SDR into the mix so it can listen on multiple freqs at a time, say, 60 kHz, 2.5, 3.33, 5, 7.85, 10, 15, 20, and 25 MHz. It would monitor all of the time freqs and sync to any that are good enough, meanwhile keeping track of signal strength & demodulation quality. Each freq would be its own time source in ntpd. So if all were sync'd, would have 9 sources.

Obviously need custom code to take care of that and something special in ntpd. This just made a long-term project list. Good thing I bought a DSP programming book last year. Will probably start with an SDR monitoring one freq at a time and when it becomes too poor to sync, then it scans for others.


I've had a few clocks around here for many years that sync to WWVB. Three, I think is the current active count (although I thought there was more). Clock radio, mantle clock, and a wristwatch. Wish my SW receivers with the built-in clocks could sync to WWVB, WWV/H, or CHU.

[redhat]

AFAIK, it's usually easier to use GPS in this application, because the GPS receiver through triangulation can figure out propagation lag and correct for it.  All the studio and CDMA master clock oscillators I've seen are GPS backed.  They have a holdover delay of an hour or so before they fall out of tolerance should the GPS fail.

You can sometimes find used 10 MHz GPS disciplined sources at hamfests and ebay.  They're not cheap, but it might be cool to say you are "the" 6925 KHz reference

+-RH

[skeezix]

You're right, GPS better than HF since it can correct the delays, and right now I do not know what my HF propagation delay is. There's a formula that I can use to estimate it, but that's only an estimation.  The GPS/CDMA source is on the list as I want to measure the propagation delays from there to here instead of estimating it. Then can see how well it coincides with the formula, or more importantly, how much it varies by frequency, time of day & time of year.

Yeah, I agree that a reference standard on 6925 (or thereabouts) would be quite handy. I like that a lot.

Been wanting to use WWV for many, many years and never got around to it until now.

[ChrisSmolinski]

AFAIK, it's usually easier to use GPS in this application, because the GPS receiver through triangulation can figure out propagation lag and correct for it.  All the studio and CDMA master clock oscillators I've seen are GPS backed.  They have a holdover delay of an hour or so before they fall out of tolerance should the GPS fail.

You can sometimes find used 10 MHz GPS disciplined sources at hamfests and ebay.  They're not cheap, but it might be cool to say you are "the" 6925 KHz reference

+-RH

I use a surplus GPS 10 MHz reference with my netSDR. Works great.

[redhat]

I use a surplus GPS 10 MHz reference with my netSDR. Works great.

I know, I read your article

I've been looking for an affordable 10MHz GPS source for a while now.  At least in my case, it would serve two fold, drive the studio clock, and provide a frequency reference for the transmitters   I guess I'll have to get some "stratum 1" liners done

+-RH

[Hacker_of_things]

I'm thinking the cheaper more accurate solution is a Raspberry Pi running NTP compiled for PPS support, unless you've got a good RTC board to maintain time between broadcasts.  NTP can also be compiled for irig/wwvb support as well.  There is a sample irig.c file somewhere in those directories, that when compiled, it can be run to generate irig or wwvb waveforms.  To manually decode either of those waveforms, one could just read the source code, or... run NTP, and make time run slow, make an arduino trigger (like an oscilloscope) only on the sync-word, and then output the waveform through the serial port.  From there, just like decoding Infrared pulses:  What does a binary 1 look like? what does a binary 0 look like?  Is it binary? (of course it is).  How many bits make a base 10 human readable number?  How are the bits arranged to make that base 10 human readable number?  Is the time-string transmitted backwards or forwards, or a mix?  In the waveform, where does the waveform change the fastest => there's your milliseconds, microseconds, or nanoseconds, depending on how far back the decimal goes.

[mysticstar]

If WWV etc. stops transmitting next year, what kind of effect will that have? I know, it's probably a stupid question, but I'd really like to know. I know it will affect us in the ham/shortwave end of things, but people forget that various manufactured items depend on that signal. I would really like to know.

[dad98253]

Would WWVB work as well / better?

Actually, I had my wwvb ("atomic clock") receiver connected to my ntp server back on Red Hat 7 in the early 2000s. I had to tweak the driver of what turned out to be an identical receiver designed to be used with the British equivalent to wwvb (I think it was called MSF). If I recall, it was just 1 or 2 lines of code that needed changing. The code was a bit messy and could have used some cleanup. But, it worked. I don't remember the driver name. You'd have to pull the latest release to see if it was still in there.
Also, the wwvb clock I had was the only model that I could find at the time that had a serial output. I don't know if there is anything available these days with a digital output. You might be able to find a CMMR-6 module (maybe from an old atomic clock?) to roll your own. But, programming a commercially available SDR might be easier.
-----------
update:
I just found a wwvb receiver module on e-bay. Model number WVB-0860N-03A. It might make an interesting project!

[ThaDood]

https://www.radioshackcatalogs.com/flipbook/1985_radioshack_catalog.html?fb3d-page=102   Would that work?

[skeezix]

https://www.radioshackcatalogs.com/flipbook/1985_radioshack_catalog.html?fb3d-page=102   Would that work?

I presume you mean the "Desktop Weather and Time Receiver" (12-149). I have the 12-148 and would need a better antenna plus some way to siphon off the audio into the computer. There is only a speaker, no audio output so some messing around would have to happen. There's a switch for 5, 10, and 15 MHz.

Following up to my original post from 2014... I did have that FT-840 tuned to 10 MHz for around 9 years. Turned it off last winter as the server I had connected to I didn't use much anymore and no point to leave it running. Plus, the backlight on the FT-840 is now burned out. Apparently it can't be left on 24x7 for 9 years. As far as ntpd went, it worked just fine, when 10 MHz was propagating. Once in a while it would sync to WWVH. The other systems at home happily sync'd to that server.