Decoding phase shifted signal in 198kHz LW radio carrier

There is a Long Wave (LW) radio transmission using a 198kHz carrier. The carrier contains a Phase-Shift encoded digital signal. The phase of the carrier is advanced by +22.5deg and retarded by -22.5deg to encode the digital signal. Each bit of the digital signal takes 40ms to transmit, so 25bits/s are sent.

I would like to use an Arduino to receive the LW signal and continuously decode the digital signal in the carrier.

I have no radio or electronics expertise. My hope is that there is a Radio or DSP shield that could receive the LW transmission and output an analogue signal proportional to the phase shift of the 198kHz carrier. I could then sample the analogue signal say every 5ms and this would allow me to determine the encoded digital bits.

Can anybody point me towards shields suitable for this task? Suggestions of other ways to tackle the problem would also be welcome.

That is the channel on which BBC Radio 4 is broadcast in the UK, but it is analogue. Just out of curiousity, what are you trying to receive? I may be wrong, but it sounds like the signal you are describing uses DPSK specifically. Wikipedia describes a broad method of decoding it, it makes a good read: http://en.wikipedia.org/wiki/Phase-shift_keying#Differential_phase-shift_keying_.28DPSK.29

I know this really wasnt the most useful response, but I hope I have been at least somewhat helpful.

It is BBC Radio 4 on Long Wave upon which you can listen to the "Archers" etc. However superimposed on the 198kHz carrier signal, by phase shifting, is a 25bit/s signal containing digital messages. These messages can be read wherever the radio signal can be received i.e. the whole UK. If I tell you what the messages are about I would have to ...... well you know ;)

I thought decoding this signal would be interesting and a good way to learn to use Arduino. I can only do it though if there is some kind of shield that can detect the phase shift of the carrier. If that exists then the project should not be too difficult otherwise it is way beyond me as I am not a soldering iron person.

It should be a pretty reliable service - it is said that if the UK Trident fleet cannot receive Radio 4 on 198 long wave for three consecutive days, they are to follow their orders. :(

I certainly hope it's reliable! I'd hate to have a power outage start ww3.

I had an idea in mind if you could get the signal into the Arduino (or at least the peaks), but it hinges on detecting a .31 microsecond change in peak timing, just slightly beyond the abilities of the Arduino IDE's micros timing. You could play games with it, tell the IDE that the board is running at 8Mhz and overclock it to 32Mhz that would make the four microsecond minimum tick from microseconds() (or micros(), I forget) a single microsecond, but that doesn't get down to the level of detecting a .3 microsecond shift.

I saw somewhere on here where people were talking about a potential for delayNanoseconds, if that can get down to .1microsecond levels a software decoding scheme might work.

As it stands I think it's a fascinating project, but not one I can help much on.

It appears that the system is used to control the power grid in the UK as well as send out flood warnings and other such information. Also every minute it encodes an accurate time of day packet. This in addition to being an atomic frequency standard and a BBC radio station. Pretty efficient use of spectrum. The cell phone carriers ought to take a look. :D

You need a phase locked loop to do this, google it for details. You can get a complete phase locked loop in a chip one such is the 4046.

Just had to go and get me interested in another cheap part didn't you?

Any idea if the US has a similar broadcast? It'd be fun to make a receiver for the US atomic clock or something. Forget a RTC board, just use the governments!

radman: I could then sample the analogue signal say every 5ms and this would allow me to determine the encoded digital bits.

Yes, and you'll need to synchronise with the bit-stream somehow. Do you have a reference for how the signal is encoded? Sounds like an interesting project -- do let us know how you get on!

AWOL: It should be a pretty reliable service - it is said that if the UK Trident fleet cannot receive Radio 4 on 198 long wave for three consecutive days, they are to follow their orders. :(

I'd have thought that sort of frequency far too high for submerged reception. Perhaps somewhere around 16kHz might be a more realistic figure. Yes I know that's within the audio band but it doesn't preclude wireless transmission - you "just" need a few gizmos and a suitably tuned aerial

Anachrocomputer: Yes, and you'll need to synchronise with the bit-stream somehow. Do you have a reference for how the signal is encoded? Sounds like an interesting project -- do let us know how you get on!

The digital data rate is very slow, 25bits/s , so I will have no problem using an Arduino to synchronise with the bits and to decode the message providing I can get something that gives me an analogue signal showing the phase shift of the carrier.

Full details of the radio signal can be found here http://downloads.bbc.co.uk/rd/pubs/reports/1984-19.pdf, the key information is figure 2 "Phase of data modulated carrier as a function of time". It is quite simple to understand;

  • A ‘1’ is represented by the carrier signal phase changing from a phase advance (+22.5 deg) to a phase retard (-22.5 deg).
  • A ‘0’ is represented by the carrier signal phase changing from a phase retard (-22.5 deg) to a phase advance (+22.5 deg).
  • td is 40ms.

I need off-the-peg hardware that gives the analogue signal shown in the figure.

That is where I was hoping the forum could help. Basically I am looking for a shield that acts as a LW radio receiver and provides an analogue reading of the phase shift of the carrier. I know there are shields that act as radio receivers that can be controlled by an Arduino, allowing bands to be searched and so on, and which provide information such as signal strength – but I have not been able to find any that provide phase shift information.

Grumpy_Mike: You need a phase locked loop to do this, google it for details. You can get a complete phase locked loop in a chip one such is the 4046.

The 4046 might well do the job, but for me to be able to use it would need to be on a shield along with the LW radio receiver circuitry. I am hoping somebody will know of existing hardware that will do the job.

jackrae: I'd have thought that sort of frequency [198kHz] far too high for submerged reception. Perhaps somewhere around 16kHz might be a more realistic figure.

Sorry to spoil the Conspiracy Theory but BBC Radio 4 on 198kHz is not used by Nuclear Subs. They can communicate submerged using Very Low Frequency (VLF) [3-30kz] and Extremely Low Frequency (ELF) radio. So jackrae was right in the middle of theVLF band. This article http://en.wikipedia.org/wiki/Very_low_frequency suggests VLF signals can be received using a PC - but that is a whole other project.

Sorry to spoil the Conspiracy Theory but BBC Radio 4 on 198kHz is not used by Nuclear Subs.

It is only used if the VLF fails.

It is only used if the VLF fails.

And when the skipper fancies a football game? "Haul out the 1500 meter antenna, mates!"

Actually the antenna is a long wire trailed out behind the sub just under the water, I think it is about a quarter of a mile long. This can only be done while the sub is moving. The tuned wavelength is reduced by the action of the water because EM waves travel at about 60% of their vacuum velocity through water, this makes the antenna required a lot shorter that it would need to be on dry land.

Grumpy_Mike: Actually the antenna is a long wire trailed out behind the sub just under the water, I think it is about a quarter of a mile long. This can only be done while the sub is moving. The tuned wavelength is reduced by the action of the water because EM waves travel at about 60% of their vacuum velocity through water, this makes the antenna required a lot shorter that it would need to be on dry land.

Oops! I was going for a full wavelength antenna, but you're right, the speed of light in a vacuum is around 1.3 times what it is in water so my antenna would have to be around 3/4 of a mile long.

The wikipedia article suggested that VLF radio signals could be received using "a [u]coil[/u] of insulated wire". Is that really correct?

Returning to the subject of my post, the most promising component I have found is this Si4753 based shield from Sparkfun http://www.sparkfun.com/products/10342. However reading the datasheet I don't think it provides the carrier phase change information I need (at least not in a way that is readily accessible to me). Can anybody comment/advise on this?

radman: There is a Long Wave (LW) radio transmission using a 198kHz carrier. The carrier contains a Phase-Shift encoded digital signal. The phase of the carrier is advanced by +22.5deg and retarded by -22.5deg to encode the digital signal. Each bit of the digital signal takes 40ms to transmit, so 25bits/s are sent.

The full details are here: http://downloads.bbc.co.uk/rd/pubs/reports/1984-19.pdf

Given the carrier is amplitude modulated as well (for the audio), you need a limiting amplifier that can bring the carrier up to a constant level without introducing phase delays. Fortunately 198kHz is far easier to work with than most radio frequencies, I suspect it can be done with a high-speed opamp (video opamps have good phase linearity) and high-speed comparator to generate a square wave. What you really want is an FM front-end despite it being an AM signal!

Then some sort of phase detector can be employed (the phase shift is 0.6us in the time domain, perhaps this can be done by Arduino interrupt handler - perhaps not)

The wikipedia article suggested that VLF radio signals could be received using "a coil of insulated wire". Is that really correct?

Yes. Probably a large coil of course ;)

However reading the datasheet I don't think it provides the carrier phase change information I need

That's right the FM demodulator is unsuitable for what you want.