How can I receive the DCF77 time signal?

[AveragePhysicsEnjoyer]

So I need some help with a decision for a project I am working on, its a Nixie Clock and I want it to synchronize with the German DCF77 time signal. I do not want to use an Of-The-Shelf module (I like it the hard way), so there are two options left and I might need some help choosing one.

Salvaging:
I might be able to salvage the receiver circuit from an old commercial clock.
I actually already tried that but only the ferrite antenna was salvageable, the receiver circuit is made out of blobs which are useless to me. Unfortunately this seems to be common for these receivers.
So I could continue to look for old clocks until I find one that contains a salvageable receiver circuit that I can transfer to my project. This I consider the easy way.

Roll my own receiver:
I have very little experience with RF circuits but I am open to try my hands on designing an DCF77 RF receiver from scratch. Now I have already tried to copy some amplifier circuits from the internet but they don't work, the signal is probably to weak (I live about 400km from the transmitter) and needs really good filtering to be detectable. I am planing on using that small ferrite antenna. How complicated would such a 77.5Khz AM receiver be? Is it just a sharp bandpass filter and amplifier? Or do I need a mixer or something? Is it doable with some opamp's and bjt's?   

I think rolling my own receiver would be a good learning opportunity but it might lead me deep into the RF rabbit hole. What do you think? Is it doable? Should I just stick to salvaging? How would you go about this project? Maybe someone could give me an idea of how such an receiver circuit would work.
Or should I just give up and buy a module?

Greetings from Germany

[Roehrenonkel]

Hi,
 
i've seen them cheaper, but for looks:
https://www.pollin.de/p/dcf-empfangsmodul-dcf1-810054

[kripton2035]

I've had difficulties to have a lasting good reception with these modules a long time ago.
it seems our modern devices all over our homes, and also our concrete walls prevent a good reception from these modules
sometimes it works, sometimes not, and it depends deeply of the location in the house.
I switched to internet ntp servers using esp8266 long ago and it works all the time. getting the time on power up takes less than 2 secs.

[SiliconWizard]

DCF-77 has become unreliable to use due to the amount of EMI around that we have these days. I have implemented a device using DCF-77 about 10 years ago and reception was still ok, but for a couple years this has become unusable in many areas. So, you've been warned. DCF-77 is AM so it's easily drown in noise. Now if you are based in Germany, much closer to the emitter, that may be less of a problem. You'll have to experiment.

I wouldn't recommend rolling your own receiver unless you really want to learn RF - off-the-shelf receivers are dirt cheap. But, if you want to as an exercise, why not. It's AM on a 77.5 kHz carrier. The usual approach is a superheterodyne receiver. There is an article here that could get you started: https://onmenwhostareongraphs.wordpress.com/tag/superheterodyne/

Given the low carrier frequency, if you want to experiment designing it, you could alternatively consider SDR (software-defined radio): a relatively low sampling rate should be enough for this, which you could probably achieve with a modest MCU. Just an idea to explore - I didn't try it myself.

[Benta]

For a chip, only the (obsolete) Temic U4221 and U4224 come to mind. Also possibly T4224 after Atmel bought the line from Temic.
If you're lucky, NOS is around somewhere.

[coppercone2]

DCF-77 has become unreliable to use due to the amount of EMI around that we have these days. I have implemented a device using DCF-77 about 10 years ago and reception was still ok, but for a couple years this has become unusable in many areas. So, you've been warned. DCF-77 is AM so it's easily drown in noise. Now if you are based in Germany, much closer to the emitter, that may be less of a problem. You'll have to experiment.

I wouldn't recommend rolling your own receiver unless you really want to learn RF - off-the-shelf receivers are dirt cheap. But, if you want to as an exercise, why not. It's AM on a 77.5 kHz carrier. The usual approach is a superheterodyne receiver. There is an article here that could get you started: https://onmenwhostareongraphs.wordpress.com/tag/superheterodyne/

Given the low carrier frequency, if you want to experiment designing it, you could alternatively consider SDR (software-defined radio): a relatively low sampling rate should be enough for this, which you could probably achieve with a modest MCU. Just an idea to explore - I didn't try it myself.

Could a tight preselector fix this?

I found this beast and I was thinking what it could be used for
http://www.m0wwa.co.uk/page/M0WWA_MW_band_preselector.html

That is broad band, but I wonder if you went with a similar quality on a single frequency could it retrieve the signal


That guy looks like he had enough interference while trying to pick up MW

[David Hess]

How complicated would such a 77.5Khz AM receiver be? Is it just a sharp bandpass filter and amplifier? Or do I need a mixer or something? Is it doable with some opamp's and bjt's?   

Maybe someone could give me an idea of how such an receiver circuit would work.

Are you interested in recovering the exact carrier frequency?  If so then a tuned receiver is required which has no mixers and no local oscillators.

[eliocor]

I think no one can beat the Chinese prices:
https://www.aliexpress.com/item/1005006439934453.html
some other info (modules):
https://blog.blinkenlight.net/experiments/dcf77/dcf77-receiver-modules/
see also the other posts on DCF77 on the same blog:
https://blog.blinkenlight.net/blog/
and
https://github.com/udoklein/dcf77
otherwise, if you want to follow the SDR route, take a look at:
http://www.i2phd.org/armradio/index.html

[Solder_Junkie]

The head amp part of this project can easily be moved to 77.5 KHz:
http://g4jnt.com/MSFPart1.pdf

As others mentioned, local electrical noise can be a problem these days.

SJ

[coppercone2]

interesting information about LF receivers, mostly about a instrument but I thought it was a real page turner.
http://www.radioblvd.com/Hammarlund_SP600VLF_Receiver.htm

[RoGeorge]

- my first bet is that the blob IC, near the antenna shown in the first picture, will output at one of its pins the demodulated DCF77 signal.
- second bet is the blob IC near the 32768Hz crystal will output a 1Hz at one of its pins

The entire module with blob ICs can be used as a time-base for the rest of the Nixie display.  If not, there are plenty of DCF77 decoders for Arduino, so search for DCF77 Arduino receiver, and a part of these will be for a 77kHz receiver (an opamp), not for the digital ready-made Chinese module.


As for the reliability of the reception, it doesn't need to be 24/7.  Usually a DCF77 clock only tries to adjusts itself once a day or so, main ticking is based on its internal 32kHz tuning-fork crystal.  I live in Bucharest, Romania, at the edge of the DCF77 signal range, and one of the DCF77 clocks I have is able to sync when the antenna is properly aligned relative to the receiver.  The other DCF clock is on a wall and it is not able to synchronize (because the ferrite axis, can not reorient the house walls relative to the DCF77 transmitter ).

Even if the reception is not possible at all, there are Arduino libraries to fake a DCF77-like signal, and transmit that at low power, for your own house only.  The Arduino library takes the time from Internet, either by LAN or by a WiFi module.  For example, https://blog.blinkenlight.net/experiments/dcf77/dcf77-generator/


Some text notes I've took while I was tinkering with DCF77 a year ago

Code: [Select]

2023-04-07 19:45, Fri
=====================
- best DCF77 overview so far, from the PTB itself, at 50 years DCF77 anniversary

        - http://www.marvellconsultants.com/DCF/PTBM_50a_DCF77.pdf
     
    - DCF77 is a call sign, D for Germany, C for long waves, F for Frankfurt, 77 a number (here hints the frequency)
    - 77.5kHz carrier
    - amplitude is reduced from 100% to 25% during ticks
    - each second tick encodes one bit
        - 25% amplitude reduction for 100ms of the carrier means 0
        - 25% amplitude reduction for 200ms of the carrier means 1
        
    - the 59th second mark is omitted, to announce the next tick will mark the start of a new minute
    - during each minute, the following are transmitted BCD encoded:
        - the number of the minute
        - the hour
        - the day
        - the day of the week
        - the month
        - the year (2 digits)
    - all encoded numbers are for the next minute
    - see page 12 of 27 in PTBM_50a_DCF77.pdf
   
    - a pseudorandom phase shift keying is modulating the same carrier (usually the phase modulation ignored by most DCF77 receivers), in addition to the AM modulation

And a receiver that can make use of the carrier's PSK:
http://www.marvellconsultants.com/DCF/

[shabaz]

Another option (maybe you have already considered it) if you don't decide on a network connection, could be to use a GPS module. They are quite low cost (at least, for a one-off project) and work well even indoors in many residential environments. Admittedly it's overkill to use just for the time for a clock, but it's also fun exploring the data received if you're interested in that.

[eliocor]

do not forget the wonderful DCF77 / JJY / WWVB emulators for mobile phones:
https://play.google.com/store/apps/developer?id=mokyuu
they really works!
Even better if coupled with the following antenna (to be connected to the earplug connector):
https://www.aliexpress.com/item/1005004884103037.html

[RoGeorge]

About DCF77 transmitters to sync a clock, probably the funniest way to do that is with a sound card and a speaker. 
https://bastianborn.de/radio-clock-hack
https://github.com/gkowal/dfc77-sync

[coppercone2]

Romania has the lowest frequency radio station for music too, roughly double the frequency.

    Radio Antena Satelor

I tried to find it on some world SDR and its pretty hit or miss in European ones. I wondered if its possible to pick it up overseas. I think the problem with tuning is pretty similar to this DCF signal. 200 KW transmitter.

Also, messing with a clock at work that picks up a LW broadcast by using your phone is a interesting idea.

Wonder if you can improve the front end of that cheap module by connecting it to the front end of a modified old radio parts capable of LW reception.

Getting a atomic clock out of one of those might be the best use anyone has found for it.

[RoGeorge]

Didn't know about such record, but there is indeed an AM radio broadcasting band in Europe (longwave, roughly between 150kHz-300kHz https://en.wikipedia.org/wiki/Longwave ), the lowest band for broadcasting here.  Probably nobody still make receivers for that band nowadays, but old analog AM radios used to have that band.  See for example the numbers painted on the scale of this Romanian receiver "Gloria" (probably from the 70's or so), https://images.okr.ro/auctions.v3/2000_2000/2011/11/30/e/f/674878966433803049741704-3341152-2000_2000.jpg first number is 150kHz, last is 255kHz.

[coppercone2]

it might not be hard to modify it to a lower frequency, its not too far off. I think the ferrites or whatever should be compatible

I think if it can do 150 it can do 70

a simple tecsun can pick up longwave, they are common

I think it will offer better selectivity then that  cheap board

[RoGeorge]

That could be an option, too.  I remember seeing something like that on a ham website, not sure if it was for DCF77 or for VLF https://en.wikipedia.org/wiki/Very_low_frequency (which is even lower than DCF77, VLF goes down to audio frequency, but with radiowaves  ).

[Kleinstein]

An RF receiver for the relatively low frequency is possible and can be a learning project. One can still use a mixer and similar, even if one is intereseted in the exact frequency. A superhet type receiver can actually be a bit easier than a simple 1 frequency receiver, as there is less gain at 1 frequency and less porblems with coupling back to the antenna that can make a receiver oscillate. A large part of the filtering should be in the analog way, but the final part could well be in the digital domain, as some SDR project with low bandwidth.

Some of the EMI can be so strong that one will not get a good result at all times. So one has to make the receiver part somewhat error tolerant. After all the signal has quite some redundancy and error correction over more than 1 minute is possible, though not the standard and rarely used. One may not need it inside Germany, but would be a thing for long distance, beyond the original planed range (e.g. in Spain).
There is now more EMI from SMPS, but less from CRT based TV (the 5x harmonic from the 15.625 kHz horizontal frequency was a nasty interference).

For the antenna one could use ferrite rods or a relatively large copper only frame antenna - both ways work, though the ferrite is more common and more compact.

The signal is no longer pure AM. There is also a phase modulation part than can be used and has some extra data.

[eliocor]

as I previously posted, at: www.i2phd.org/armradio/index.html you will find a complete SDR receiver (8÷900kHz) which costs no more than 35÷40€, front end included (see documentation).
There is also a sample of DCF77 reception.
Not needing a display (fixed frequency) the code can be easily ported to less expensive chips like STM32F407: https://www.aliexpress.com/item/1005001658111159.html which has the same 3 ADC structure like its bigger sibling (STM32F429).

sample of DCF77 reception: https://www.i2phd.org/code/dcf-77.mp3 (from Northern Italy)

[Doctorandus_P]

Here in the Netherlands DCF reception quality varies. Often there are many hours where the signal can't be received, but at other times, I get a valid time code every minute for hours on end. At night reception is much better then during the day. Maybe less interference when people are sleeping and factories are closed, but probably more due to atmospheric related issues.

I scavenged some very cheap commercial clocks 20 or so years ago. They also have blobs (I think), but the "raw" DCF signal was quite easy to find on the PCB. I did have to adjust my code a bit because pulse durations deviated quite a lot. I also wrote an algorithm to keep track of the pulse widths and use it as an estimation of reception quality. Reception of these cheap commercial things is quite mediocre. I guess a big part of the problems described above is simply due to low quality hardware.

[coppercone2]

That could be an option, too.  I remember seeing something like that on a ham website, not sure if it was for DCF77 or for VLF https://en.wikipedia.org/wiki/Very_low_frequency (which is even lower than DCF77, VLF goes down to audio frequency, but with radiowaves  ).

http://www.vlf.it/

[coppercone2]

Here in the Netherlands DCF reception quality varies. Often there are many hours where the signal can't be received, but at other times, I get a valid time code every minute for hours on end. At night reception is much better then during the day. Maybe less interference when people are sleeping and factories are closed, but probably more due to atmospheric related issues.

I scavenged some very cheap commercial clocks 20 or so years ago. They also have blobs (I think), but the "raw" DCF signal was quite easy to find on the PCB. I did have to adjust my code a bit because pulse durations deviated quite a lot. I also wrote an algorithm to keep track of the pulse widths and use it as an estimation of reception quality. Reception of these cheap commercial things is quite mediocre. I guess a big part of the problems described above is simply due to low quality hardware.

I think the main thing is
-low quality tank circuit

[coppercone2]

And the unique attraction now to LW is the fact that you can use simulated inductors or possibly capacitance multiplier circuits that lets you go to new places for "RF design" without a semiconductor MMIC lab

That is, a active inductor, or possibly something like making a giant capacitor from a small stable one, with amplifiers. (I often wondered if something like a small stable C0G ceramic can be 'multiplied' to some useful value, but putting work on the inductor is the rational choice.)

https://www.aavso.org/improved-gyrator-tuned-vlf-receiver


However, I suspect that just getting a regular ferrite tank circuit working will be more then rewarding

and that it works for a sound card, unfortunately the DCF77 is too high frequency for this  :'(


I am still rofling at the idea of some communist radio spitting out a german atomic clock signal

[Kleinstein]

The frequency of DCF77 is still high enough that one can use real inductors with a ferrite core. One would want a closed core to reduce coupling to the outside. The capacitors should not be that critical and normal film types like PP or polyester should be OK. If really neded one could combine the 2 to reduce the temperature sensitivity, but I don't think this is really needed.

There are a few old projects on building DCF77 receivers. They usually have the antenna with a first amplifier (e.g. source follower) seprate as a kind of active antenna that is separate from the main receiver circuit. I have build such a circuit from an old Elector plan, but it used an now long obsolete mixer chip. AFAIR there are only 2 resonant elements the the 77 KHz, the antenna and one more LC filter. As one would need some µC or similar for the decoding one could as well do some of the final filtering an demodulation in the digital domain, likely as a super-het receiver with a ZF more in the audio range.

A nice thing with the low frequency is that probing is not that critical, normal transistors and if wanted OP-amps can be used and parasitics are not that important.