I thought I'd float this here in case someone else has walked the path...

In several countries around the world there are radio towers that broadcast time-synchronisation signals, which allow suitably equipped devices to update themselves and maintain very accurate time. There are now clocks and even wristwatches that can receive these signals to maintain the accuracy of their time-keeping. I have just acquired a watch that can do this (a quite inexpensive Casio), which is also solar-powered. So not only can it keep itself updated with extremely accurate time data, but it also charges itself from almost any light source. Amazing stuff.

Where it falls down slightly for me is that I am in Australia, and while we have plenty of sun to charge my watch, we do not have any of these time transmission towers here - or not yet anyway. There are many things about the watch that appeal to me so it wasn't a deal-breaker. I doubt that establishing any of these towers here is on anyone's list of priorities, so I started wondering if/how I could go about providing the facility myself. The obvious source of very accurate time is the GPS network, which can be accessed quite easily and inexpensively. I use a Pi 5 as a server for my smart home system, so it would have plenty of capability to run something that (a) talks to a GPS receiver and continuously maintains the current time, and (b) transmits that time on a suitable frequency and in a suitable form for my watch to see and process. It would only need to be very low power/range, I only need my watch to be able to see and sync to it when I'm at home.

I suspect that many (most?) of the Pi tinkerers on this forum live in countries which have these towers and so do not need to be considering this, but I will ask anyway - has anyone here investigated this? The GPS side of things is simple - there are many USB-connected GPS receivers available, and even a few Pi hats that can do it I think. Decoding the GPS data is quite straightforward. But in searches to date I haven't seen a Pi-based radio transmitter module that might be suitable for broadcasting the time info. And of course I will have to find out how to encode the data onto the radio signal, but I suspect that will be out there somewhere.

This is still just a thought bubble for me really, but if anyone has walked this path, it would be good to hear how you went (or how you are going).

Cheers, Steve.

Which model of Casio is it?

Oh no, not again.

My old Casio "Waveceptor" watch could receive the UK MSF (formerly Rugby Clock) or German DCF77 signals. It used to be great, but when the transmitter moved from Rugby to Cumbria the signal seemed to have got weaker and my watch rarely managed to synchronise.

I have a number of clocks around the house that sometimes show incorrect times and dates due to poor reception (and presumably lax error checking) and I had considered building a low power MSF transmitter to make sure they show the correct time. I haven't done so because broadcasting on those frequencies would be illegal in the UK.

One quick warning: You may see some mentions on the web about using a Pi GPIO pin as a transmitter. Don't do that! It produces a very bad waveform and produces messy harmonics that bleed all over the radio spectrum. I only mention this as a warning not to do it that way (get a proper transmitter module instead), and that any posts suggesting or asking about that method get deleted from these forums (the moderators get very upset).

Unreadable squiggle

ame, it's a LCW-M100DE, one of the "Lineage" line and not sold by normal dealers outside of Japan, for some reason. The watch aficionados call them JDM watches, for Japan Domestic Market. But plenty of online sellers. I like it a lot.

rpdom, thanks for the reply, will add all those thoughts to the mix. You mention the possible illegality of transmitting on one of these frequencies - yes it would probably be illegal here too, which is why I'd want it to be very low power.

Having built projects to receive MSF 'Rugby time code' I also built an MSF transmitter for testing. That was a digital signal which I injected into the receiver's demodulated output ... There's no technical reason the output couldn't be taken to a transmitter to make it wire-free and effectively an MSF tower. MSF uses a 60 kHz carrier so can be generated direct from a Pi, Pico or any other microcontroller output pin.

I did that when I was producing 5.3 kHz radio signals to fake heart beat pulses for a heart rate monitoring watch. I took the signal out to a random reel of hook-up wire and that worked with the watch in the middle of it.

Doing the same with 60 kHz may be illegal and produce horrible harmonics but it should be low-power and have extremely limited range. A proper transmitter would be recommended but using the frequency may still be problematic. Putting it and the receiver in a Faraday Cage should make it legal.

The Casio watch specification seems to suggest the watch can receive MSF, DCF77, WWVB, BPC and JJY but I've never used anything other than MSF.

I wouldn't use GPS. A Pico W or Pi which supports Wi-Fi or Ethernet which can fetch NTP time from the internet should be good enough and minimises cost.

My advice would be to not do anything illegal. That may make your idea infeasible.

marshallarts wrote: ↑

Sat Oct 18, 2025 11:01 am

ame, it's a LCW-M100DE, one of the "Lineage" line and not sold by normal dealers outside of Japan, for some reason. The watch aficionados call them JDM watches, for Japan Domestic Market. But plenty of online sellers. I like it a lot.

rpdom, thanks for the reply, will add all those thoughts to the mix. You mention the possible illegality of transmitting on one of these frequencies - yes it would probably be illegal here too, which is why I'd want it to be very low power.

Ok. This guy has some suggestions:
https://www.twentytwoten.com/666/will-a ... australia/

If you're near Brisbane, for example, you might pick up the time signal from Japan.

Oh no, not again.

For DCF77 signals, 77.5kHz there are some proposals in the net about how to create 'fake signals'. E.g. https://blog.compass-security.com/2015/ ... ipulation/ or https://blog.blinkenlight.net/experimen ... generator/

https://github.com/hzeller/txtempus