Can radio-controlled clocks work on electric railways?

[Basil Jet]

I used to have a DCF-receiving watch and I could never get it to receive the signal from Frankfurt while on an electric (OLE) railway in London. I tried an MSF-receiving clock on Wednesday night at Waltham Cross station but no luck - even taking the clock a significant distance across the adjacent field wouldn't pick up the signal from Cumbria. Thinking that being practically at sea level might be part of the problem, I then tried reception on a station on the Hertford Loop with instant success, but since the trains were bustituted on Wednesday night, the OLE on the Hertford Loop might have been switched off.

The reason it matters is because there is currently a competition to design a clock for the railway, and the competition advice does not specify how the time will be received - if MSF is not receivable on electric railways and internet time sources are intrinsically laggy, how would one set the time on a railway clock?

 

[DelW]

I have several radio controlled clocks and watches, and I've found that they are less reliable at updating since the transmitter was moved from Rugby to Cumbria quite a few years ago. There's quite a variation between individual units, some update anywhere in the house but others need to be placed on an upstairs windowsill. (I'm in south east England).

One answer to your question might be to use GPS signals as AIUI that system works off very accurate timekeeping.

The FM RDS signal also contains time information, though I don't know whether that's planned to continue for the foreseeable future.

 

[stuu]

Weren't the NSE clocks radio controlled? Or was that a myth to show how modern and focussed on punctuality the organisation would be?

 

[Ediswan]

The reason it matters is because there is currently a competition to design a clock for the railway, and the competition advice does not specify how the time will be received - if MSF is not receivable on electric railways and internet time sources are intrinsically laggy, how would one set the time on a railway clock?

For internet time, use Network Time Protocol (NTP). That seeks to compensate for lag and can be expected to deliver accuracy of tens of milliseconds when used over the public internet. There is plenty of detail available online.

How good is your internal clock ? It may not need to sync very often, so even intermittent MSF/DCF77 reception might be sufficient.

Is there a specification to be met ?

 

[The exile]

I used to have a DCF-receiving watch and I could never get it to receive the signal from Frankfurt while on an electric (OLE) railway in London. I tried an MSF-receiving clock on Wednesday night at Waltham Cross station but no luck - even taking the clock a significant distance across the adjacent field wouldn't pick up the signal from Cumbria. Thinking that being practically at sea level might be part of the problem, I then tried reception on a station on the Hertford Loop with instant success, but since the trains were bustituted on Wednesday night, the OLE on the Hertford Loop might have been switched off.

The reason it matters is because there is currently a competition to design a clock for the railway, and the competition advice does not specify how the time will be received - if MSF is not receivable on electric railways and internet time sources are intrinsically laggy, how would one set the time on a railway clock?

My German bought radio-controlled watch received its signal from Braunschweig and is/was right at the range limit - is it possible that the weakness of a signal coming from Frankfurt may not cope with electrical interference, but a stronger (because nearer) signal might?

 

[dosxuk]

internet time sources are intrinsically laggy

Internet time sources are extremely reliable and more than accurate enough for a clock - NTP is specifically designed to counteract any connection latency. Internet time and GPS have pretty much entirely replaced the need for radio sources.

 

[Carlisle]

Weren't the NSE clocks radio controlled? Or was that a myth to show how modern and focussed on punctuality the organisation would be?

Yes they were, they had an aerial usually fitted high up on the station building or canopy somewhere.

 

[Bletchleyite]

I think this competition is more about industrial design, isn't it? That is, to create a clock face as iconic as the SBB one? It's fairly likely it'll often be shown on an LCD screen as LNER stations now do.

 

[Ashley Hill]

My TOC watch is supposedly radio controlled but doesn’t change when the clocks go back/forward.

 

[Basil Jet]

Thanks, all. Doesn't GPS time use a modified version of GMT that has ignored every leap second since Telstar by The Tornadoes was in the hit parade, and doesn't it also give no information about daylight saving time?

 

[stuartl]

My TOC watch is supposedly radio controlled but doesn’t change when the clocks go back/forward.

Is that because the signal is always sent as GMT or possibly UTC (Universal Coordinated Time) ?

 

[Ediswan]

Internet time and GPS have pretty much entirely replaced the need for radio sources.

Radio is still good for low power devices with no need other for internet or GPS. Watches and clocks are the prime example.

Is that because the signal is always sent as GMT or possibly UTC (Universal Coordinated Time) ?

Both MSF and DCF77 adjust their main signal for summer time. They do both flag when summer time is in effect, so a device could choose to compensate and always show UTC.

 

[Basil Jet]

I think this competition is more about industrial design, isn't it? That is, to create a clock face as iconic as the SBB one? It's fairly likely it'll often be shown on an LCD screen as LNER stations now do.

Each shortlisted competitor will receive GBP7000, presumably to help them develop a physical object, and I don't think they'd do that if you were just designing a GIF.

 

[Ediswan]

I think this competition is more about industrial design, isn't it? That is, to create a clock face as iconic as the SBB one? It's fairly likely it'll often be shown on an LCD screen as LNER stations now do.

Assuming this is the RIBA competition, the brief does say:

The Phase 1 submission should outline the overall vision and big idea, the concept for the display and communication of time across the UK National rail network and associated built environment estate.

How the clocks would be set does seem to be in scope.

 

[josh-j]

I read "communication of time" to mean represent the time to the public, i.e. show people what time it is. The competition looks much more designery than techy

 

[Richardr]

My Casio radio controlled watch changes for the start and end of summer time

Don't such watches try to connect in the early hours, and try every hour for three or four hours if they don't connect at first? Given the accuracy of most timepieces, I assume any radio controlled clock would only need to pick up the time signal once every day or so.

 

[josh-j]

My local station's screens seem to pick up the correct time about once in a blue moon...

 

[Ediswan]

I read "communication of time" to mean represent the time to the public, i.e. show people what time it is. The competition looks much more designery than techy

Could be. Gloriously ambiguous. My background is showing. I had time sync 'sorted' for my employer long before Y2K came along.

 

[jfowkes]

Thanks, all. Doesn't GPS time use a modified version of GMT that has ignored every leap second since Telstar by The Tornadoes was in the hit parade, and doesn't it also give no information about daylight saving time?

Yes, any clock using GPS time has to apply local information (timezone, DST, leap seconds) to the GPS time to get local time.

Note this is also true an NTP-based clock. You get UTC time, which includes leap seconds, but you still have to apply local timezone/DST offsets to get true local time.

For a railway clock I'd use NTP because sometimes stations are underground and you can't rely on GPS. Also because one day, maybe, if we're very, very lucky, DST will be consigned to the dustbin of history and with a network connected clock you can check if DST is in effect or not. A pure-GPS one would need to have DST rules hardcoded.

 

[Buzby]

When MSF was in Rugby, reception up here in the Clyde valley was hit and miss. When it moved to Anthorn, I never had any problem. Some older clocks needed to be oriented carefully, whilst newer ones would even work from a biscuit tin. I did find that WiFi routers and LED lighting cause much interference so the issues of ‘electricity’ may not be the real issue. My latest clocks all access NTP so if MSF shut down, I probably wouldn’t miss it…..

 

[tellytype]

MSF operates at 60kHz - its possible that the frequency is covered in hash from the electronic systems of the train, or that you're just inside a very well constructed & grounded faraday cage when you're inside one.

 

[rower40]

Before it transferred to York ROC, the York IECC had an MSF clock supplying an accurate time to ARS and other items in the signalling centre.
When the health of this time signal was presented to the signallers in the form of a colour-coded clock in the corner of the workstation, they noted that it went red (free-running - not synched to the Anthorn transmitter) whenever there was a Voyager in platform 11. As soon as the Voyager pulled out, the clock went green (time signal received) again.

My recollection is that re-aligning the antenna so that it pointed towards Cumbria rather than Rugby fixed this.

But this goes to show it's not just electric trains!

 

[Basil Jet]

MSF operates at 60kHz - its possible that the frequency is covered in hash from the electronic systems of the train, or that you're just inside a very well constructed & grounded faraday cage when you're inside one.

Thanks. IME you can't get the signal sitting on the platform while no trains are near, or even in a nearby field. Trying to set a radio controlled clock in a moving vehicle, or even while you are carrying it down the street on your arm, is hopeless.

 

[tellytype]

Thanks. IME you can't get the signal sitting on the platform while no trains are near, or even in a nearby field. Trying to set a radio controlled clock in a moving vehicle, or even while you are carrying it down the street on your arm, is hopeless.

Interesting - Theres an EMC issue here possibly. That said where are you? Just says London here. The MSF move to Anthorn may well be partly to blame but its been years now! Although it IS run by Babcock so...

Got access to a Spectrum Analyser

 

[PM77]

Before it transferred to York ROC, the York IECC had an MSF clock supplying an accurate time to ARS and other items in the signalling centre.
When the health of this time signal was presented to the signallers in the form of a colour-coded clock in the corner of the workstation, they noted that it went red (free-running - not synched to the Anthorn transmitter) whenever there was a Voyager in platform 11. As soon as the Voyager pulled out, the clock went green (time signal received) again.

My recollection is that re-aligning the antenna so that it pointed towards Cumbria rather than Rugby fixed this.

But this goes to show it's not just electric trains!

This is still an issue at the IECC building whenever an electric train is in 11 and sometimes 10. The MSF clock also supplies accurate time for most the original IECC interlockings which are still situated in the old IECC. The time is used for the SSI logging systems.

 

[edwin_m]

Many years ago at BR Research I tested a Datatrak navigator fitted to the roof of a train (pre-GPS, apparently operating on 70kHz from ground-based beacons). This worked tolerably except when near an overhead wire (either 25kV or even a National Grid line) when it became basically useless. 70kHz is well above the mains frequency and probably well above any frequency produced by electric traction (this was before solid-state drives anyway) with the possible exception of short bursts of interference from pantograph bounce. EMC works in mysterious ways.

 

[Dai Corner]

A directional aerial for a wavelength of about 5km must be quite large?

 

[Gaelan]

I think this competition is more about industrial design, isn't it? That is, to create a clock face as iconic as the SBB one? It's fairly likely it'll often be shown on an LCD screen as LNER stations now do.

A fascinating choice in this day and age - I struggle a bit* to read analog clocks and I think many people my age (early 20s) are the same.

(Presumably they do intend to provide both, but still, it seems like a lot of space for something fewer and fewer people will use.)

* I can do it! But it takes a few seconds to stop and think, whereas I can read a digital clock instantly. Certainly I wouldn’t be able to read one as I was rushing for a train!

 

[Dai Corner]

A fascinating choice in this day and age - I struggle a bit* to read analog clocks and I think many people my age (early 20s) are the same.

(Presumably they do intend to provide both, but still, it seems like a lot of space for something fewer and fewer people will use.)

* I can do it! But it takes a few seconds to stop and think, whereas I can read a digital clock instantly. Certainly I wouldn’t be able to read one as I was rushing for a train!

As an aside, a school examinations officer of my acquaintance tells me that she has to provide digital clocks in the exam rooms as the 11-18 year candidates can't understand analogue ones.

 

[Ediswan]

A directional aerial for a wavelength of about 5km must be quite large?

I built an MSF receiver once, 1970s. It used a short ferrite rod antenna. That was directional. Same as an old-style portable AM radio.