Appreciate everyone's thoughts. Not a very popular idea it seems. I guess I'm about the only one who has an OCD over accurate time, and with 4 clocks that used to need setting and updating for DST. Oh well, maybe I'll post this as a Project article. Or maybe I'll take a stab at it anyway and see what happens.

My original thought was to dispense with the atomic clocks completely and just use the GPS as a time source to a home rolled display. Interesting project though. Go for it.

 

What do you actually use those atomic clocks for? Just to know the time? Or there is kind of a more demanding scientific application?

 

What do you actually use those atomic clocks for? Just to know the time? Or there is kind of a more demanding scientific application?

They are just wall clocks or alarm clocks. They are set each night by a radio signal from a US Government owned broadcast in Colorado.

 

A couple ideas for you, based on my experience from building wwvb simulators.

1. Pick a crystal or crystal oscillator on the wwvb transmission frequencies. Build your MCU around that frequency.
2. Configure the MCU to output the frequency.
3. Use one of the MCU pins to modulate the output of clock on that clock output pin.
4. If your MCU support serial output, configure those modules to control the modulation through An interrupt.

The whole thing can be done in less than 10 USD, especially if the GPS allows user configurable output frequency on the wwvb frequencies.

 

A couple ideas for you, based on my experience from building wwvb simulators.

1. Pick a crystal or crystal oscillator on the wwvb transmission frequencies. Build your MCU around that frequency.
2. Configure the MCU to output the frequency.
3. Use one of the MCU pins to modulate the output of clock on that clock output pin.
4. If your MCU support serial output, configure those modules to control the modulation through An interrupt.

The whole thing can be done in less than 10 USD, especially if the GPS allows user configurable output frequency on the wwvb frequencies.

The crystal I used was 15.36MHz. From that I can derive both the baud rate for the GPS and the 60KHz carrier. The MCU does output that frequency, gated on/off per the WWVB format. That was only half the story. The rest was converting the GPS time to WWVB format and determining the 2 DST setting bits in the 57th and 58th second. Without that, the clocks keep good time up until a DST change. But I sure don't see how one could do it for under $10 since that's what the GPS itself cost. You also need a long antenna, so for another $7-8 you can get a 100' roll of 4-conductor phone cable with RJ11s on each end, which makes a pretty large diameter 4-turn coil. A small coil like the antennas used in atomic clocks just doesn't have any range. And a power supply, $5. Way more than $10 total.

 

You may not legally manufacture, import, or offer for sale an intentional radiator in the United States unless it is certified to conform to the laws. The high cost and long wait of testing and certification is a significant barrier to minimally funded start-ups.

 

What do you actually use those atomic clocks for? Just to know the time? Or there is kind of a more demanding scientific application?

There are lots of very demanding uses. The broadcast is not just time, it is time and frequency. The time is used by lots of people for lots of reasons. One use is to control the generator speed at power plants so that the integrated error remains very small (something like two seconds since the last time the system was reset). The frequency signals are used to set all kinds of oscillators to the correct frequency so as to be within allowed tolerances.

 

You may not legally manufacture, import, or offer for sale an intentional radiator in the United States unless it is certified to conform to the laws. The high cost and long wait of testing and certification is a significant barrier to minimally funded start-ups.

FCC Part 15.209 reads:

Emissions from an intentional radiator shall not exceed the field strength levels specified in the following table:

Frequency (MHz) Field strength (microvolts/meter) Measurement distance (meters)
0.009-0.490 2400/F(kHz) 300
0.490-1.705 24000/F(kHz) 30
1.705-30.0 30 30
30-88 100 ** 3
88-216 150 ** 3
216-960 200 ** 3
Above 960 500 3

2400/60 = 40 uV/M at 300M (1000ft)

So, why do I think I'm legal? It's an educated guess. NIST designed the WWVB transmitter to provide a minimum field strength of 100uV/M throughout the US. As you can see by their field strength maps, this is barely attained along the east coast (and some parts don't even make that). So I estimated that what I receive here in NC is about their design goal, 100uV/M, since all my WWVB clocks just barely receive this signal at night (outside or in a window), and that's with the antenna oriented just right. My WWVB transmitter will set my clocks at a range of 200 ft (but not at 250ft), so I took that to be roughly the 100uV/M level. Four times farther away (1000ft) and I'm at 39.99uV/M.

Edit: The format of the above table got fractured in posting, but it can still be read.

 

For starters read 15.201 in the pdf file at the link below.
http://www.gpo.gov/fdsys/pkg/CFR-2009-title47-vol1/pdf/CFR-2009-title47-vol1-part15.pdf

 

For starters read 15.201 in the pdf file at the link below.
http://www.gpo.gov/fdsys/pkg/CFR-2009-title47-vol1/pdf/CFR-2009-title47-vol1-part15.pdf

Dick, you are correct. I called an FCC test center and although I may be "legal" as far as .209 is concerned, I would still need to submit the thing for certification testing before I can sell it. (Cost? About 5 grand.)

So, as president and CEO of this little company, I'm hereby canceling this project and firing the engineer who thought it might fly. No wait... that's me.

 

I am glad that you checked with the real authority rather than merely accepting the opinions of posters on the internet.

 

I would still need to submit the thing for certification testing before I can sell it. (Cost? About 5 grand.)

you would expect them to say that.

whether that's correct or not, it is not entirely clear - I think you should consult a lawyer - the FCC would be a biased party here.

 

Well, if anyone is interested, this is how that project turned out. It's in my attic now, working well. Took forever to get my boards out of China.

 

Very nice looking board!

 

That would make a good Instructable. Sell the boards, and kits of parts. The builder gets the responsibility of making sure there's no interference.

 

Very nice looking board!

Thank you.

 

So I would be interested in buying the schematic and board... yeah I am your target market - OCD about time...

 

So I would be interested in buying the schematic and board... yeah I am your target market - OCD about time...

Tell ya what ... I built 3 of these, and I'm using one and the other 2 will probably just sit around and gather dust. I'm convinced that selling them isn't a viable plan. If you'll send me a box with your address and postage on it, I'll drop one in the box and send it back to you. Won't charge you anything. It's a gift.

 

Tell ya what ... I built 3 of these, and I'm using one and the other 2 will probably just sit around and gather dust. I'm convinced that selling them isn't a viable plan. If you'll send me a box with your address and postage on it, I'll drop one in the box and send it back to you. Won't charge you anything. It's a gift.

Hmm there doesn't seem to be any PM on this site...

can you drop me a line at (email address removed by moderator for your benefit. I will send your email address to Wingsy for you. New members are not permitted to send email until after a certain number of posts. -dc)

 

Inexpensive wireless "Synchronized" consumer level clocks (especially cordless wall-hangable clocks) are so-far a market without a solution IMO. Typically a setup would consist of a single "Master" clock unit that is GPS disciplined that sends a time pulse to the "Slave" clocks and (at a minimum) synchronizes all the clocks to 00:00:00 at local midnight (especially the analog clocks) To save battery power, the clocks "tick" once per second using their own internal crystals (type 32.768KHz). The problems with this approach are the wireless segment (very low-power to allow battery powered slave clocks with good battery life) and analog clocks that can (at-least) move all hands to exactly 00:00:00 at midnight (some time ago I did see some slave analog clocks like this from China, but I was not able to actually source any for tests. Don't ask for links - long lost).

The wireless segment seems like it can be accomplished today using ISM-band "LoRa or LPWN/LPN" wireless modules which are now cheap and readily available from China, even in small quantities. See this as a start:

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

Do a web search for something like 'LoRa Arduino'. Lots of links will pop up.