2
\$\begingroup\$

Is it possible to modify radio-controlled clocks (so-called "atomic" clocks, like https://www.amazon.co.uk/Wall-Clocks-Radio-Controlled/s?ie=UTF8&page=1&rh=n%3A376322011%2Cp_n_feature_keywords_browse-bin%3A1249834031) to synchronise to an arbitrary (local) time signal rather than the standard time signal (i.e. MSF in the UK, WWVB in US, or any number of other time signal stations across the world)?

As I see it, this would involve two parts:

a.) Adjusting the radio receiver in the unit to listen on a different radio frequency of a locally-provided time signal rather than the default frequency

b.) Broadcast the desired time signal in the expected format on that frequency.

I've had a look at the signal standard, and I think I'd be able to broadcast a signal as described here http://www.npl.co.uk/upload/pdf/MSF_Time_Date_Code.pdf using a handful of components and an Arduino. However, I'm not really sure how I'd approach adjusting the receiver as described in a.) - or whether there's a better approach I could take? (or other flaws in my plan I haven't considered), and my Google-fu has not brought up any useful resources to suggest others having achieved this in the past. I wondered whether I'd even need to change the frequency, or whether producing a local signal on the same frequency would have sufficient power to "override" the official time signal? Any advice would be gratefully received!

To further explain my purpose; I'm not trying to improve the accuracy - I'm trying to broadcast and display an arbitrary local "stopwatch" time, unrelated to whatever the actual current time might be. (From asker's asker's clarifying comment)

\$\endgroup\$
  • 1
    \$\begingroup\$ The main question is why bother? You can't improve on the accuracy of the existing standard time signals. WWWB can be reliably received over most of the United States. What advantages are there in your scheme. \$\endgroup\$ – Barry Dec 22 '17 at 13:52
  • \$\begingroup\$ 1) link to the product doesn't work 2) it depends on the product if it is "hackable", only the people who designed it can tell you for sure. Maybe it has a (mask) ROM so every setting is "burned" forever into the system. 3) It would be much easier to build your own radio controlled clock and make it receive what you want. 4) you might not be allowed to transmit timing signals even if it is for your own use! \$\endgroup\$ – Bimpelrekkie Dec 22 '17 at 14:03
  • \$\begingroup\$ @Barry - Sorry - I didn't explain my purpose; I'm not trying to improve the accuracy - I'm trying to broadcast and display an arbitrary local "stopwatch" time, unrelated to whatever the actual current time might be. \$\endgroup\$ – Alastair Aitchison Dec 22 '17 at 14:03
  • 1
    \$\begingroup\$ You ain't gonna "adjust" a typical radio clock to receive on 433MHz. You will have to replace the receiver entirely. The clock signals a typically broadcast on a low frequency. Germany uses 77kHz - there is no way you are going to adjust a receiver for that to receive a 433MHz signal. \$\endgroup\$ – JRE Dec 22 '17 at 14:19
  • 2
    \$\begingroup\$ You'd probably have more luck removing the antenna and injecting a very weak signal at the original frequency, probably through a tiny series capacitor. But be sure no part of your generator can radiate to interfere with others. \$\endgroup\$ – Chris Stratton Dec 22 '17 at 14:21
1
\$\begingroup\$

These commercial products are very difficult to adapt for other purposes:

  • Data rates are very slow, often one bit-per-second.

  • Clocks often only "update" periodically, (once an hour? a few times per day?)

  • Frequency of operation is very precise, perhaps crystal-controlled.

  • Highly integrated for low cost.

Am mostly familiar with WWVB, and have programmed a microcontroller to receive its 60 kHz signal. The project is not recommended for novices.

Data transmission via radio over short distances is a common desire served by RF modules...the transmitter accepts digital input, and the receiver outputs a digital replica. Ensure that the operating frequency and power are appropriate for your region.

\$\endgroup\$
  • 1
    \$\begingroup\$ Point 2 is what makes this completely impossible, rather than very difficult \$\endgroup\$ – BeB00 Dec 22 '17 at 14:30
  • \$\begingroup\$ I'm not so sure about point 2 though. It is my understanding that they update essentially every time they receive a complete frame with correct checksums, which happens at most once every minute. \$\endgroup\$ – pipe Dec 22 '17 at 14:40
  • \$\begingroup\$ @pipe Some turn off the radio receiver during the day when noise & propagation is poor to save battery power. \$\endgroup\$ – glen_geek Dec 22 '17 at 14:47
  • \$\begingroup\$ Thanks for the reply and comments. Indeed, point 2 would be a killer I hadn't considered. From the spec sheet I linked to "The first second of the minute begins with a period of 500 ms with the carrier off, to serve as a minute marker. The other 59 (or, exceptionally, 60 or 58) seconds of the minute always begin with at least 100 ms 'off' and end with at least 700 ms of carrier. " gave me the impression that the signal was broadcast every second, and therefore a clock could be made to sync to it at that frequency too? \$\endgroup\$ – Alastair Aitchison Dec 22 '17 at 15:07
  • \$\begingroup\$ @AlastairAitchison Available are receiver-only modules, with ferrite antenna that would seem suitable for your purposes (product promotions are forbidden here). You'd require a microcontroller to decode the pulse widths, and to drive a mechanical clock mechanism. These receivers provide pulse output as long as you apply DC power. \$\endgroup\$ – glen_geek Dec 22 '17 at 15:12
2
\$\begingroup\$

You could transmit on the standard frequency, or hack the clock to receive another.

How far do you want to transmit to this clock?

If only an inch or so, then you may be able transmit at such low power that you can drown out the national signal at the clock, and be so low power that you don't upset the neighbours, or the FCC. Any antenna would be near field at those frequencies, and so a well aligned nearly direct coupling to the clock's ferrite rod would end up broadcasting practically no EM radiation.

If you are trying to transmit metres, then that would be illegal in any jurisdiction. You would need to use a power that will upset your neighbours' timekeeping, and would bring down the wrath of the telecommunication authorities.

As for hacking a clock, it's a case of open one and see. Older clocks might be better, more likely to have a separate RF section with audio modulation out. If I was designing one these days, I might be inclined to do something clever with subsampling the RF directly after a bit of tuned amplification, and so bury the demodulator in silicon. If you can locate the modulation, then it should be a cinch to drive with the output of your 433MHz garage door opener. Test it with your Arduino directly first though.

Do you want the clock to actually be a radio wall clock, or look like one, or just look like a clock? With a display and an Arduino, you could write a clock, and equip it with your garage door opener receiver, and you're done.

\$\endgroup\$
  • \$\begingroup\$ Thanks for the reply. No, I don't want it to actually be a clock that tells the time - I want a clock that can be controlled to display an arbitrary value that can be encoded as HH:MM. To be clear, I'm talking about an "analogue" clock here with hour and minute hands. If I was creating a digital clock display, I'd obviously just wire a 7-segment display block to the Arduino and be done with it :) I thought that it would be easier to modify the electronics of an existing radio-controlled clock unit rather than fiddle around with tiny cogs required to make my own mechanism, but maybe not! \$\endgroup\$ – Alastair Aitchison Dec 22 '17 at 15:03
  • \$\begingroup\$ Yes, sub-sampling 60 kHz analog input (into microcontroller AD) after a bit of gain & selectivity is possible. Pushes a simple 8-bit microcontroller to its limit - it is not an easy project. A 433 MHz pulse-coded (OOK) radio link is far more straightforward. In any case, OP still has to drive a mechanical display mechanism. \$\endgroup\$ – glen_geek Dec 22 '17 at 17:05
0
\$\begingroup\$

To be clear, I'm talking about an "analogue" clock here with hour and minute hands. If I was creating a digital clock display, I'd obviously just wire a 7-segment display block to the Arduino and be done with it :) I thought that it would be easier to modify the electronics of an existing radio-controlled clock unit rather than fiddle around with tiny cogs required to make my own mechanism

With this clarification of goal, it's worth pointing out that most "analog" clocks of an electrical nature actually employ a small single-coil unidirectional variety of stepper motor to drive the gear train. The coil is pulsed at an appropriate interval divided from a quartz crystal (or perhaps in line powered models the mains frequency). This tends to be true everywhere from AA-battery pound shop models up through models you might find in a school which take a wired signal from a master clock.

Because of their digital stepper-motor nature, it's quite possible to drive the electromechanical mechanism from your own pulse circuit rather than the original. Many people have used this to create their own versions of the infamous "Lord Ventinari's Clock" from Terry Prachett's Discworld novels. In that application, the clock is pulsed at an erratic rate intended to discomfort viewers, but the number of pulses is tracked and controlled so that the average time advanced is correct over longer periods.

Such a mechanism would of course have a maximum step rate, but if that is determined and honored and by careful counting of sent pulses, you should be able to walk the hands to arbitrary positions through the use of just about any small MCU. This will be far simpler than trying to produce a false time reference signal to similarly walk a radio-referenced clock's hands through its existing electronics. It's perhaps possible that in order to support self-setting radio-referenced models may have mechanisms supporting higher step rates than ordinary ones, but even in that case, it's going to be easier to replace the drive electronics and take advantage of the more capable mechanism, than to try to drive the mechanism by misleading the electronics with a false reference signal.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.