WWVB receiver advice

[ledtester]

I'd like to build a WWVB receiver. I've found two designs:

- one by Russel Kincaid published in Nuts and Volts, May 2006 (see attachment)
- another by Hans Summers - http://www.hanssummers.com/images/stories/radio/report.htm

I have a couple of "atomic" clocks, so I figure their loop stick-cap assemblies could be helpful in building my own receiver since they are most likely already tuned to 60 KHz.

My questions:

The Kincaid design calls for 3x 680uH inductors. Any good suggestions for how to go about making these? Since 680uH is a standard inductance, will something like this RF choke work (Mouser P/N 434-17-681J)? http://www.mouser.com/ProductDetail/Fastron/07P-681J-50/?qs=%2fha2pyFadugl%2fvul%2fEqHt6RvekLGgR1gz%2f98kulmaXo%3d

The Summers's design calls for two 3.5mH inductors which are described as:

"Coil L1 and L2 were wound from about 95 turns of swg 30 wire on Maplin Ferrite core type 2, LA4345."

However, I can't find any source for "Maplin Ferrite core type 2, LA4345". Any ideas of an alternative?

Thanks!

[cdev]

Here is an antenna that likely works well. You could probably make it smaller if you used a larger capacitance.

http://www.febo.com/time-freq/wwvb/antenna/index.html

[JimRemington]

I have a couple of "atomic" clocks, so I figure their loop stick-cap assemblies could be helpful in building my own receiver since they are most likely already tuned to 60 KHz.

In many of the clocks, the receiver is an entire separate module, which works fine as it is. You probably can't do any better building one yourself, so take one apart and see how well it works (I have done so).

They are crystal controlled direct conversion receivers that typically run on 1.5 - 3 V and output a low voltage pulse that corresponds to the AM modulation of the WWVB carrier, to be decoded elsewhere. Ed Nisley posted Arduino code in the April 2010 issue of Circuit Cellar, which decodes the signal and does extensive error and sanity checking. The article was entitled something like "A Totally Featureless Clock".

[cdev]

I've seen a few stories that might be relevant.  Sometimes people find hardware that contains WWVB receiving capabilities, very very cheaply, it doesnt always work though.

There is a WWVB refclock for NTP and so its possible that you could build a very cheap way of keeping a computer time server in sync that would be separate and distinct from GPS, which IMHO would be useful in terms of redundancy.

With the appropriate antenna, and software, you might be able to receive the 60 KHz WWVB signal perhaps using some PC sound cards - For low frequencies that might work quite well. You could perhaps interface it to NTP using the NTP WWVB refclock- Not important to most people but I know that with HF (3-30 MHz) tiny variations in signal propagation makes received frequencies including WWV jitter - but by very small amounts. I don't know how much lower frequency propagation jitters, it might not jitter as much.

[MrSlack]

Used to be able to get antennas for this ready made in the UK that were good. Ferrite stick ones. Managed to build a simple receiver in the distant past for Rugby MSF but never got to  decoding the signal before I got bored. Was a simple TRF receiver.

[cdev]

This looks interesting, the NIST's publication on the three time and frequency stations..

http://tf.nist.gov/timefreq/general/pdf/1969.pdf


Probably most or all AM radio ferrite rods could be made to resonate by trial and error, if you add additional wire and/or add more capacitance.  Some of the long wave radio clocks have a very small ferrite rod in them so I suspect that its simply a matter of getting it to resonate at the desired frequency, which is a lot lower than the AM broadcast band, so you may need to use a much longer length of very thin wire.

[uncle_bob]

Hi

The easy way is to dig up one of the sub $20 boards that do this all and just power it up. They generally have an antenna on them and everything. If that's not exciting enough then yes you will need an antenna. Since you are down at 60 KHz, the material normally used is different than what an AM broadcast band antenna uses (frequency is 10X lower). Finding the material is a bit of a chore. The last batch I got was from a Russian source. The other alternative is to buy a full blown antenna for ~ $100 or so.

After that you have two choices:

Either decode the "old" WWVB AM modulation format or do the "new" phase modulation format. The phase mod *can* be done with a bunch of this and that. I'd do it with one of the ARM DSP radio boards. The older format is a bit easier to do. Either one done with parts rather than code will involve tuned circuits / bandpass filters and tuning stuff up with test gear.

Of course there is another choice ... buy a cheap WWVB clock and take a hammer to it. There *is* a radio in that $20 clock.

Bob

[MrSlack]

You can buy ones with serial ports if you put a few more $$ in as well from Galleon. We use them on sites with no NTP+Internet connection as a reference clock except MSF here in the UK. Bit of software on the computer keeps the clock synced with it and then everything slaves off that box with NTP.

[cdev]

What about sound cards? I biught a USB sound card the other day that samples up to 96 KHz and I paid around $10 but in the past I saw it for around $2. Its quiet, the SN ratio is over 100 db. It should be able to receive 60 KHz *with an antenna*...

There is something fun to try..  http://www.lowfer.us/k0lr/wolf/wolf4beginners.htm 

Do people need a license to transmit VLF? Its not that much higher than an audio system.. really.

A lot of people seem to have Very low data rate beacons..

[johnkenyon]

Do people need a license to transmit VLF? Its not that much higher than an audio system.. really.

You need a very long antenna (or equivalent), and a lot of power to excite that antenna.

To get things back on topic, if you are only interested in the ASK/OOK signal from WWVB, it should be noted that WWVB (US) and MSF/Time from NPL(UK) receivers are identical.
The only difference between the two is the encoding.

[uncle_bob]

Do people need a license to transmit VLF? Its not that much higher than an audio system.. really.

A lot of people seem to have Very low data rate beacons..

Hi

Yes you need a license and you can only transmit in bands (at power levels) that your license permits.

Bob

[Jeroen3]

In Europe it's called DCF77, https://en.wikipedia.org/wiki/DCF77 and it's located in Germany.
Receiving it is easy, decoding is horror (or a challenge) since you will never receive 1 full minute without noise.

[deephaven]

Here's a UK source of radio modules http://www.pvelectronics.co.uk/index.php?main_page=index&cPath=9

[MrSlack]

Thanks for the link - will definitely grab one of them for a tenner!

[uncle_bob]

Hi

As noted earlier MSF and WWVB are identical from a "cheap receiver" standpoint. There is no obvious reason to pay more for one labeled MSF than for one labeled WWVB. If the receiver has a decoder in it and spits out some sort of converted data (say via a serial port) you then do need to get the right one for your area.

Bob

[cdev]

You need a very long antenna (or equivalent), and a lot of power to excite that antenna.

Thats what I thought too for quite a while, but I just read this.

https://en.wikipedia.org/wiki/LowFER

[uncle_bob]

Hi

....And then take a look at what it takes to *receive* those low power beacon transmissions. The stuff you can pick up with your little receiver modules is transmitted with 60KW and up (as in up to megawatts) transmitters. They run into antenna systems of the "size of a subdivision" or larger category.

Bob

[aa7ae]

I'm also interested in building a WWVB receiver to ultimately use with a nixie tube clock.  I found a schematic for a 60 Khz receiver at http://www.burningimage.net/clock/ however, when started putting it on the breadboard, I'm getting nothing but noise in the oscilloscope and no indication it's picking anything up at all.  I'm even using a commercially available WWVB loopstick antenna from pvelectronics in the UK.  I'm not sure what I could be doing wrong with that schematic.
Link to the schematic: http://www.burningimage.net/clock/wp-content/picclockparttwo.jpg

There is also a fairly new article in QEX November/December 2015 on building a more advanced WWVB receiver for a frequency standard you can use in your lab.  It does use some relatively expensive instrumentation op amps to do the job rather than discreet components like I'm trying to do.  That article is here: http://www.arrl.org/files/file/QEX_Next_Issue/2015/Nov-Dec_2015/Magliacane.pdf 

[deephaven]

That circuit doesn't look very selective, it's just relying on the ferrite rod antenna for tuning. That is why you see a crystal in commercial designs as it's used to provide a very narrow bandwidth. Also 60 KHz is very prone to interference from various sources like monitors, TVs, switch mode power supplies etc. You might get an improvement if you earth the circuit with a real earthing rod stuck in the ground.

[uncle_bob]

I'm also interested in building a WWVB receiver to ultimately use with a nixie tube clock.  I found a schematic for a 60 Khz receiver at http://www.burningimage.net/clock/ however, when started putting it on the breadboard, I'm getting nothing but noise in the oscilloscope and no indication it's picking anything up at all.  I'm even using a commercially available WWVB loopstick antenna from pvelectronics in the UK.  I'm not sure what I could be doing wrong with that schematic.
Link to the schematic: http://www.burningimage.net/clock/wp-content/picclockparttwo.jpg

There is also a fairly new article in QEX November/December 2015 on building a more advanced WWVB receiver for a frequency standard you can use in your lab.  It does use some relatively expensive instrumentation op amps to do the job rather than discreet components like I'm trying to do.  That article is here: http://www.arrl.org/files/file/QEX_Next_Issue/2015/Nov-Dec_2015/Magliacane.pdf

Hi

The QEX circuit does indeed work.

WWVB is "interesting" in terms of picking it up. If you happen to live in Denver, it's not much of a problem. If you live in Bangor Maine, it's a really big deal. In most of the country, midnight is a good time to try. In New England, MSF interference makes that problematic.

The next issue is ... switchers ...For whatever reason, 60 KHz is  a common frequency for low end switching power supplies. They often to not have much filtering. Some can be nasty enough that a neighbor can take out WWVB. Hunting that kind of thing down can be a pain.

Even very small antennas can be directional. Rotating a tiny rod 90 degrees can change the signal quite a bit. If you are out in free space (as in 1/4 wave at 60 KHz) you can predict the way to point the antenna. With "junk" in the near field, you can get some really odd field distortions.

All that said, it's just signal to noise. None of these receivers are very fancy. If I try to get my watch to sync at sunset (worst case for propagation) there is no chance of it locking up. Wait a few hours and bang it gets the signal. The noise hasn't changed a bit. The signal has improved a lot ....

Bob

[aa7ae]

Hi

The QEX circuit does indeed work.

WWVB is "interesting" in terms of picking it up. If you happen to live in Denver, it's not much of a problem. If you live in Bangor Maine, it's a really big deal. In most of the country, midnight is a good time to try. In New England, MSF interference makes that problematic.

The next issue is ... switchers ...For whatever reason, 60 KHz is  a common frequency for low end switching power supplies. They often to not have much filtering. Some can be nasty enough that a neighbor can take out WWVB. Hunting that kind of thing down can be a pain.

Even very small antennas can be directional. Rotating a tiny rod 90 degrees can change the signal quite a bit. If you are out in free space (as in 1/4 wave at 60 KHz) you can predict the way to point the antenna. With "junk" in the near field, you can get some really odd field distortions.

All that said, it's just signal to noise. None of these receivers are very fancy. If I try to get my watch to sync at sunset (worst case for propagation) there is no chance of it locking up. Wait a few hours and bang it gets the signal. The noise hasn't changed a bit. The signal has improved a lot ....

Bob

Here is a design to make a very narrow crystal filter for picking up WWVB or MSF that might be worth trying.  Although these op-amps are kind of expensive, but may be found for less on EBAY.  I happened to purchase a bag of 60 Khz crystals on EBAY a while ago that I might try in this experiment.

Link to article: http://www.creative-science.org.uk/MSF_crystal_filter.html

Schematic:

[uncle_bob]

Hi

The QEX circuit does indeed work.

WWVB is "interesting" in terms of picking it up. If you happen to live in Denver, it's not much of a problem. If you live in Bangor Maine, it's a really big deal. In most of the country, midnight is a good time to try. In New England, MSF interference makes that problematic.

The next issue is ... switchers ...For whatever reason, 60 KHz is  a common frequency for low end switching power supplies. They often to not have much filtering. Some can be nasty enough that a neighbor can take out WWVB. Hunting that kind of thing down can be a pain.

Even very small antennas can be directional. Rotating a tiny rod 90 degrees can change the signal quite a bit. If you are out in free space (as in 1/4 wave at 60 KHz) you can predict the way to point the antenna. With "junk" in the near field, you can get some really odd field distortions.

All that said, it's just signal to noise. None of these receivers are very fancy. If I try to get my watch to sync at sunset (worst case for propagation) there is no chance of it locking up. Wait a few hours and bang it gets the signal. The noise hasn't changed a bit. The signal has improved a lot ....

Bob

Here is a design to make a very narrow crystal filter for picking up WWVB or MSF that might be worth trying.  Although these op-amps are kind of expensive, but may be found for less on EBAY.  I happened to purchase a bag of 60 Khz crystals on EBAY a while ago that I might try in this experiment.

Link to article: http://www.creative-science.org.uk/MSF_crystal_filter.html

Schematic:

Hi

The gotcha with that circuit is that there is no way to set the crystal on frequency. The trimmer is used to null out the C0 of the crystal. If the crystal you have does not happen to match the load that circuit provides .... not going to work. It's a fine way to go in mass production. You call up the crystal guy and order 500,000 crystals (for $50,000) and tell him they have to work in that circuit. He says thank you very much for the order ... they will be there in a month. You are happy, because the crystals work. He's happy because they would have sold for half that price without the special requirements. 

Bob

[XFDDesign]

On the subject of inductors, you will find that many of the large-valued inductors from Mouser/Digikey/etc have awful Q.

For hand-winding, that is definitely a better way to go (for small builds). I source toroids from www.amidoncorp.com -- for large values of 680uH to 5mH, I recommend buying some of the FT50A-J units ( http://www.amidoncorp.com/ft-50a-j/ ). Their Al value is around 3000mH/1000T, so 3.5mH would only be 34 turns (roughly). 680uH would be 15 turns.

Their Q is far better than the off-the-shelf stuff of Mouser, but not as good as a powdered iron toroid.

[Jay_Diddy_B]

Hi,

Regarding the inductors:

The Summers's design calls for two 3.5mH inductors which are described as:

"Coil L1 and L2 were wound from about 95 turns of swg 30 wire on Maplin Ferrite core type 2, LA4345."

However, I can't find any source for "Maplin Ferrite core type 2, LA4345". Any ideas of an alternative?


Maplin, is an electronics retailer, in the U.K.

The LA4345 would have been manufactured by Mullard (Philips) which is now Ferroxcube.

They are 'pot cores'

They would look like this:



To get the target inductance 3.5mH with 95 turns, they would need to have an Al value of 400 nH per turn squared.

Hope this helps.

Regards,

Jay_Diddy_B

[MrSlack]

Seems like a good time to drop this link: http://toroids.info/

[uncle_bob]

Seems like a good time to drop this link: http://toroids.info/

Hi

Look very carefully at the core materials. Some are designed for power applications where tight stability specs are not at all common. For filters, you need a material that has good temperature stability *and* adequate Q *and* can be wound without going nuts.

In some cases, wonderful stuff like Litz wire comes into the mix as well.

Bob

[NivagSwerdna]

Have you seen this

Nice receiver

[uncle_bob]

Have you seen this

Nice receiver

Hi

WWV != WWVB

60 KHz != 10 MHz

Bob

[NivagSwerdna]

WWV != WWVB
60 KHz != 10 MHz

Ah! Apologies.

Here in the UK we have similar and the easiest solution is an off the shelf module...

e.g.

http://www.pvelectronics.co.uk/index.php?main_page=product_info&products_id=7

but actually the MSF60 signal here is difficult to receive especially around computer equipment, the DCF77 signal from Germany is better.

However, if you have line of sight to the sky GPS is far easier.