Radio controlled clocks

[LADmachiningParticipant @ladmachining]

Posted by John Haine on 16/06/2021 11:27:18:

Going back to my first reply to Tony, as all the clocks I make are likely to have Arduino or similar controllers, it struck me that if one could "broadcast" a near-60 kHz signal (or even near-77) around the house, for example on the mains, with the same modulation as used by these clocks, one could display ones "own right time" anywhere you want, on multiple dials, using simple receivers that are battery powered. I don't think the radio regulators would like it much!

There are a number of projects that do exactly this, using either an Arduino or Raspberry Pi. Most show a simple loop antenna, which only allows transmission distances of a few inches to avoid interfering with any other equipment or fall foul of any unathorised broadcast laws.

Two of probably many examples:

RasPi TxTempus (can simulate multiple time signal standards, such as DCF77, MSF etc)

DCF-77 Generator for Arduino (Udo Klein has produced a number of useful libraries and other projects for working with the DCF77 signal)

Anthony

[David CoulingParticipant @davidcouling56621]

Posted by Tony Jeffree on 16/06/2021 10:17:36:

Posted by Neil Wyatt on 15/06/2021 11:13:14:

The biggest challenge is remembering how to tell digital ones that use the Frankfurt signal to subtract an hour after you have changed the battery or dropped them on the floor!

 

I have a digital clock with exactly this issue….is there a fix?

Cheers David

 

Edited By David Couling on 16/06/2021 17:53:38

[duncan webster 1Participant @duncanwebster1]

Posted by John Haine on 16/06/2021 11:27:18:

Going back to my first reply to Tony, as all the clocks I make are likely to have Arduino or similar controllers, it struck me that if one could "broadcast" a near-60 kHz signal (or even near-77) around the house, for example on the mains, with the same modulation as used by these clocks, one could display ones "own right time" anywhere you want, on multiple dials, using simple receivers that are battery powered. I don't think the radio regulators would like it much!

I'd love to do this round the house, but by radio, not all that keen on messing with mains voltage. I've played with the modules available for Arduino and managed to send data, but to do this job I'd need a range of 40ft through at least one brick wall. My potential master clock is in the workshop as SWMBO has decreed it too ugly for the house. My knowledge of radio is slightly less than my knowledge of serial comms, which itself is zero

[Michael GilliganParticipant @michaelgilligan61133]

Probably worth linking ‘the horse’s mouth’ at this point in the discussion: **LINK**

https://www.npl.co.uk/msf-signal

MichaelG.

[Robert Atkinson 2Participant @robertatkinson2]

Posted by LADmachining on 16/06/2021 16:42:03:

Posted by John Haine on 16/06/2021 11:27:18:

Going back to my first reply to Tony, as all the clocks I make are likely to have Arduino or similar controllers, it struck me that if one could "broadcast" a near-60 kHz signal (or even near-77) around the house, for example on the mains, with the same modulation as used by these clocks, one could display ones "own right time" anywhere you want, on multiple dials, using simple receivers that are battery powered. I don't think the radio regulators would like it much!

There are a number of projects that do exactly this, using either an Arduino or Raspberry Pi. Most show a simple loop antenna, which only allows transmission distances of a few inches to avoid interfering with any other equipment or fall foul of any unathorised broadcast laws.

Two of probably many examples:

RasPi TxTempus (can simulate multiple time signal standards, such as DCF77, MSF etc)

DCF-77 Generator for Arduino (Udo Klein has produced a number of useful libraries and other projects for working with the DCF77 signal)

Anthony

Even a "simple loop antenna" can radiate significant energy. While I accept that the probability of causing nterference are low and getting caught even lower, it is illegal and irresponsible to emit any signal on these frequencies. If you really want to do sometning like this use a GPS (with external antenna) for the time and transmit the serial NMEA data using a readymade, approved licence free module using a system like a RF modem, bluetooth serial or zigbee.

Robert G8RPI.

[SillyOldDufferModerator @sillyoldduffer]

Posted by Robert Atkinson 2 on 16/06/2021 19:36:22:

Posted by LADmachining on 16/06/2021 16:42:03:

Posted by John Haine on 16/06/2021 11:27:18:

Going back to my first reply to Tony, as all the clocks I make are likely to have Arduino or similar controllers, it struck me that if one could "broadcast" a near-60 kHz signal (or even near-77) around the house, for example on the mains, with the same modulation as used by these clocks, one could display ones "own right time" anywhere you want, on multiple dials, using simple receivers that are battery powered. I don't think the radio regulators would like it much!

There are a number of projects that do exactly this… Most show a simple loop antenna, which only allows transmission distances of a few inches to avoid interfering with any other equipment …

…

Even a "simple loop antenna" can radiate significant energy…

Robert G8RPI.

Building an efficient antenna for 60kHz is itself a interesting challenge. The easiest simple efficient antenna is a half wave dipole at least half a wavelength above ground. As 60kHz is 5000 metres, we need two towers 2.5km apart and 2.5km high.

A 1/4 wave vertical can be efficient provided it has a good ground connection. This needs a single tower 1.25km high, which is a third taller than the world's highest man made structure. A good earth at this frequency is special too: about 300 copper radials each a 1/4 wavelength long, covering an area of about 4.9 million square metres.

Not practical! We might manage a 50 metre tower, which is 1% of the wavelength and has a radiation resistance of 0.04 ohms. That means massive currents and high losses. A 50 metre tall antenna might be 0.05% efficient. Getting a 60kHz signal to travel far enough to cause the neighbours a problem needs more than a simple loop. Putting 60kHz on the mains could cause trouble though.

For sending a time signal around the house a radio module transmitting in one the public bands makes life easy; efficient antennas are physically small and there are cheap matching receivers. And it's legal!

Dave

 

 

Edited By SillyOldDuffer on 16/06/2021 21:16:22

[John HaineParticipant @johnhaine32865]

The big problem with systems such as Bluetooth etc is the protocol latency, and the power consumption of the modem. RC clock movements work for years on a single battery and the "protocol" is purely synchronous. They are also very sensitive. Quite a number of mains communication systems have used frequencies in this range, for example there was an MK system back in the 80s working at 100 kHz. Sounds reasonable that one could combine them.

[John HaineParticipant @johnhaine32865]

This link is interesting, frequencies for power line communication.. 59 – 61 kHz is protected for standard time transmission (good!), "61 to 90 Available without restriction." – brackets the 77.5 kHz Frankfurt signal. I would guess that applying a small capacitor to a Frankfurt clock in parallel with the coil on the ferrite rod could shift its frequency down a bit then one needs a way to safely couple to the mains wiring. There are quite a few DIY mains comms projects that show ways to do this. In the case of the MK system that I worked on, IIRC it applied the carrier between earth and neutral and left the live well alone.

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by John Haine on 16/06/2021 21:43:08:

The big problem with systems such as Bluetooth etc is the protocol latency, and the power consumption of the modem. RC clock movements work for years on a single battery and the "protocol" is purely synchronous. They are also very sensitive. Quite a number of mains communication systems have used frequencies in this range, for example there was an MK system back in the 80s working at 100 kHz. Sounds reasonable that one could combine them.

I agree that power consumption is a problem, but protocol latency need not be. For example, IEEE Std 802.1AS, a standard I worked on back in the day, defines a time synchronisation protocol that allows such problems to be taken into account – it was originally defined for use in Audio/Video networks so that several nodes in a network could broadcast their media simultaneously, but the engineers at CERN implemented the protocol in such a way that it was possible for them to time-synchronise nodes around the LHC to within a couple of nanoseconds.

[John HaineParticipant @johnhaine32865]

But I bet the nodes didn't run for 2 years on an AA battery, and cost somewhat more than £15?

**LINK**

[Tony JeffreeParticipant @tonyjeffree56510]

Posted by John Haine on 17/06/2021 10:09:52:

But I bet the nodes didn't run for 2 years on an AA battery, and cost somewhat more than £15?

**LINK**

True…and I dread to think of their development cost. However, if you are running an Arduino or similar, there's nothing terribly difficult about the protocol & you should be able to get synchronisation down to within a few milliseconds, especially if you don't have to deal with any other kinds of traffic on the network.

[Tony JeffreeParticipant @tonyjeffree56510]

…and of course a simple approach that doesn't need the receiver to have any additional intelligence is to measure the transmission latency in your software and adjust what and when you transmit accordingly – the latency should be reasonably constant.

[SillyOldDufferModerator @sillyoldduffer]

Disappointed with my own effort, I ordered a Canaduino MSF receiver which arrived last week.

My ultra-basic TRF receiver used an old Long Wave Ferrite Rod antenna capacitively loaded to tune at 60kHz: not ideal because the quality of a tuned circuit depends somewhat on the inductance and capacitance being reasonably matched. Achieving resonance with a big inductor and tiny capacitor, or a big capacitor and tiny inductor, are both compromised. As always the middle way is best.

Although the Anthorn 60kHz MSF time signal was visible on an oscilloscope, my receiver picked up so much electronic muck there was no chance of decoding the signal with an Arduino. I have a couple of communications receivers and they showed the signal to noise ratio in my home from Althorn to be poor : it varies by day and night, but is always marginal due to man-made noise. My home is surrounded by telephone wires and a pole mounted power distribution system, both of which emit high levels of interference from 0Hz to 25MHz, and even beyond. The mains borne noise is generated by almost anything unfiltered my neighbours plug in : switch mode power supplies, electric motors, VFDs etc etc.

The Canadian receiver is a much better bet than my TRF. In addition to a better engineered Ferrite Rod antenna, their receiver comes with a 60.003kHz crystal, which, being sharply tuned, is much better at rejecting noise. (The receiver also has automatic gain control and the electronics needed to create nice sharp pulses for input to a computer.)

Had two misadventures fitting the crystal: opening the well packed plastic box, the tiny crystal flew out Good job I saw it out of the corner of my eye, because the gremlins hid it under a bookcase, Later, soldering the crystal to the board, I thought 'must be careful not to get solder on the chip next door', at which point another gremlin twitched my elbow and deposited a solid blob across 4 pins. Time to solder crystal, 90 seconds, then 20 minutes wasted carefully removing unwanted solder. My eyesight and clumsiness don't go well with Sub-miniature devices.

Anyway, the receiver survived and it works! Alas, not quite good enough. Although the LED flashes more or less correctly, the recovered signal is still unreliable. Next step is to box the receiver up neatly with a battery and take it walkies. Pretty sure it will work correctly when moved away from the overhead street wiring. To fix this problem, I have to buy another house!

Plenty of reports of MSF clocks misbehaving. Having looked closely at the signal received by mine, and read Canaduino's advice, the clocks are vulnerable to man-made interference. I suspect many clocks only get a decodable signal intermittently, and the ordinary crystal movement is corrected randomly, but 'good enough'. Others, may work perfectly well until you, or a neighbour, plugs in a device that f*rts on 60kHz. My consistently high noise problem is probably common too: it's the sum of many mucky devices nearby drowning out the Anthorn signal. May not actually be necessary for me to move house, because I can get a usable signal from a communications receiver, and the Canaduino might work if fed with a well aimed loop antenna on the roof. More trouble than I care for.

Dave

[Alan Wood 4Participant @alanwood4]

I gave up on MSF based sync a long time ago and have now fitted in excess of 30 DCF based modules to clocks for friends and family around the UK and France without any operational problems. The modules have all been sourced from Cousins UK. They are delivered with the 'pin' lock at 12 o'clock. The hands are then set to either 11 o'clock or 12 o'clock before powering up depending on where their operational location.

I made a dual face clock built into a metal tube with Perspex dials and end faces which had two DCF units, one driving each face. I decided to use an angel ring light as a back light inside the tube. The small 12V to 5V SMPU that came fitted to the ring light totally swamped the radio modules leading to a swap out for a linear regulator. That apart the tube housing no matter what orientation seemed to have no impact on the DCF module reception so radio field strength must be good.

In business we used to have a product called Tick Tock used in TV studios that had a GPS based master receiver that distributed the time signal and 'pips' in analogue radio code format to each studio clock. Each studio clock had a standard radio module fitted with a coupling to detect the RF signal which was phantom carried over the power supply feed to the clock.

[Author]

Posts