ฝัง
- เผยแพร่เมื่อ 17 พ.ค. 2024
- A remotely controlled, precise analog wall clock like that one from IKEA is a cool thing. Today we will build our own remote control for such analog and digital clocks. All we need is a Raspberry Pi and a few electronic components. And we will learn something about resonance frequencies, capacitors, and ferrite antennas. Let’s hack!
I am a proud Patreon of GreatScott!, Electroboom, Electronoobs, eevblog and others.
You can do the same using an ESP32: github.com/aknik/ESP32/tree/m...
In this video, we will:
- Learn how radio-controlled clocks work
- We will build a ferrite antenna for receiving and transmitting time signals
- We will test and optimize it
- We will install software on our Raspberry Pi to create signals for DCF77, WWVB, MSF, and JJY
In the end, I should have a precise time in my lab. Even if it is in the basement
Links:
Software: github.com/hzeller/txtempus
DCF77 Clocks: s.click.aliexpress.com/e/Or4Og0Gs
WWVB Clocks: s.click.aliexpress.com/e/rynh4QLQ or bit.ly/2V3f1Dz
Ferrite Rods: s.click.aliexpress.com/e/E9kPH6ku
Copper wire: s.click.aliexpress.com/e/kjXSDLuM
Instructions: www.sensorsiot.org/remote-cont...
Supporting Material and Blog Page: www.sensorsiot.org
Github: www.github.com/sensorsiot
My Patreon Page: / andreasspiess
My Bitcoin address: 19FSmqbBzb5zsYB1d8Bq4KbxVmezToDNTV
If you want to support the channel, please use the links below to start your shopping. No additional charges for you, but I get a commission (of your purchases the next 24 hours) to buy new stuff for the channel
For Banggood bit.ly/2jAQEf4
For AliExpress: bit.ly/2B0yTLL
For ebay.com: ebay.to/2DuYXBp
profile.php?...
/ spiessa
www.instructables.com/member/...
Please do not try to Email me or invite me on LinkedIn. These communication channels are reserved for my primary job
#IKEA_Hack #DCF77 #WWVB #JJY - วิทยาศาสตร์และเทคโนโลยี
![ใจดำกับฉันก่อน - วงทัพห้า [Official MV]](http://i.ytimg.com/vi/I48wa4qurAU/mqdefault.jpg)
I love it when you give us the "little finger" while sitting in the "front row" 🤤 Always a intertaining and educational experience. Much appreciated 👍😁
Did you see: It is white now like my hands ;-)
Thank you Andreas for your informative small projekts that ordinary people can understand and enjoy. Also, I like your sens of humor! Greetings from Finland, regards Ken
Thank you for your nice words!
Thanks Andreas, love the diverse topics you cover.
You are welcome. It seems I also have diverse needs in my lab ;-)
i like how even your hand pointer is wearing a glove
Newly printed ;-)
Andreas, as usual, I thoroughly enjoyed this episode.
Thank you for creating it.
You are welcome!
Thank you Mr. Spiess. I don't have an atomic clock but now I want one. Great video I learn a lot with your channel. 😊
You are welcome!
Your radio related videos are super interesting! Keep sharing!
:-)
This is an interesting project.
"Usually the range is bigger at night". I've noticed that. Lidl was selling DCF77 alarm clocks and weather stations at one point and I bought few of them.
The weather station gets the time automatically during the night. Two alarm clocks haven't gotten the time even once during two months. The weather station only started receiving the signal when days got shorter and nights got darker.
We live 1500-2000km from Frankfurt.
That is correct. LF & MF (Long wave & Middle Wave) have a longer range during winter time as the sun is not as active as compared to the summer time. This is because the ionized layers of the ionosphere change according to the angle of the sun, thanks for correcting Professor Frink.
Mai Mariarti the angular velocity on the Equator is the same as on every point of the Earth, 15 degrees/hour. But the linear velocity is the highest, dropping as one moves towards the poles.
If the transition "zone" between light and dark is X miles wide, it will happen the fastest at the Equator due to the linear velocity, as your location moves through the zone.
Add to that that the "zone" is wider at higher latitudes due to the angled incoming light and scattering.
Good map illustrating twilight zones: www.timeanddate.com/worldclock/sunearth.html
Yep, currently, the Earth is closest to the Sun during Norther winter.
But this changes over time, together with other factors such as the amount of eccentricity of Earth's orbit, the amount of tilt to the rotational axis etc., all with their own periods, combining to higher or lower maxima/minima as the factors combine or counter eachother.
Просто интересно: 2000 км сигнал запаздывает на сколько мксек? Что показываю по ЦТВ задержка 4~5 секунд, но с этим можно смирится если она ежедневно и ежегодно стабильная, правда в высокосные года, как? Да и Земля, ☀️ не совсем круговая орбита, значит должно быть ускорение, или отставание. Корректировка фазы синхронизации с эталоном в в США, Лондоне, Токио, производиться раз в месяц, год?☕🙋
I just finished building this project using an old Raspberry Pi 3 B and it works great. Thank You for posting this project.
Cool!
Again a wonderful video, as allways you pick a unique topic. thanks for the video.
You are welcome!
Thanks Andreas for this video. I live in Australia where we don’t have a radio time signal. The kitchen clock never seems right so this seemed like a great project. 3 weeks later and I have perfect time in the kitchen and will never have an excuse for being late! I bought the inexpensive frequency generator you showed in your video. That in itself was a little fun project. The clock I got was a MSF analogue. I wound about 350 turns on a 9mm ferrite rod, single layer. I didn’t have a way of measuring impedance, so I used an online calculator for working out the windings. After a bit of experimenting I used a 680 pF capacitor with a 100pF trimmer capacitor to tune the circuit. I am using a raspberry pi zero w. I had an issue with the frequency it was generating (161O Hz rather than 60KHz) Fortunately this issue was addressed in the Issues section of hzeller git repository. I would rather not be transmitting continuously so plan to create a cron job to run for a few mins every night. I still have to case it all up but the project has been a lot of fun. One question, any suggestion on a way of working out the transmission power, so I can make sure my transmission is not likely to cause any havoc? Many thanks for your great videos. I enjoy them a lot. Gerard
First: I am glad the Australians now also have accurate clocks ;-) Tuning an antenna by experimenting is fun, but, as you describe, probably a little slower.
Concerning your question: I would put your clock a few meters away till it stops to synchronize after reset. Then you have the reach of your antenna. These clocks are very sensitive and if this range is just a few meters I would assume your transmitter is not stronger than many electrical motors or other noise sources. Just much more narrow-band.
Thanks Andreas spiess, always great & very useful videos
Thank you!
So many school and university clocks are going to be 'mysteriously' changed in the near future 😁
"Timeblaster" the equivalent to irblaster for clocks now available online. :-)
Simon, we want do this after our exam ;)
These clocks here in Germany are syncronised to the mains frequency, so this won't work here :D
@criticalDamange it is possible but most of the time i don't see this in the schematic because old Clocks with sync motors will work but now one with stepper and a cpu it is easyier to generate the frequence from the cpu freqence. But dcf transmit more then the sync impuls.
@@criticaldamage9374 Germany - "The train from Hamburg is 1 minute late. We apologise fot the inconvenience."
Africa - "The train from Cape Town has been cancelled till further notice. Find alternative transport."
Southern Africa's grid's frequency varies by up to 5%. Can you imagine all those hundreds of antique steam turbine alternators suddenly slowing down by 5% as the grid comes under load! 😱😱😱👇
Great project 👍
Thanks for sharing 👍😀
You are welcome!
My dad complained that his DCF77 clock wasn't syncing for a while. "Turn off that LED bulb", I said... and the clock synced immediately. Switch mode power supplies could become a challenge for these LF radio clocks.
Yes we had the same in the UK too, when the original clocks also used DCF77. The signal is a bit weaker here than on the continent. We now have our own 60 kHz which gives a stronger signal and has been adopted for clocks available in the UK. The compact fluorescent bulbs were even worse than the LEDs, I recall. Also bad are things like induction hobs and some mobile phones. I suppose that the secret is to place these devices and clocks in different parts of the room ! Switch mode power supplies yes if made very cheaply. I have a properly made Bosch that charges up my electric bicycle battery, and despite being next to my radio clock, doesn't affect the latter. The usual story I suppose ! great that you solved your dad's problem !
Excellent catch!!!
@Michael, this might be a reason my Ikea clock does not sync in my lab ;-)
I recently had to replace the light bulb in a garage door opener, and couldn't find a rough service incandescent one meant for that. Instead, I discovered they now make special LED bulbs for garage openers, with circuitry that won't interfere with the opener's radio control receiver. Good to be aware of that, and therefore just how many ordinary LED lamps and such could interfere with radio signals.
@Bob Blum - As far as I know, the low-RFI LED lamps you mention are currently sold only by Genie in the United States, apparently commissioned by them. Since they are a sole source, competing garage door opener manufacturers recommend their use in situations where interference from the lamps is hampering reliable reception of the the radio remote control signal.
Very interesting video. I can't wait to experiment! Cute pointer, too.
Enjoy experimenting!
Excellent video. I love your accent. Great work keep it up🕗🕜😂😁
Thank you!
Thanks, this is a useful instructional video how to get txtempus running with higher reach transmission. In my case, I wrote txtempus to get a German watch running in California, but basement clocks with poor reception will of course benefit greatly as well.
The txtempus instructions with the 'wire turn antenna" are also left deliberately simple to make sure the default built will stay within very low radio emissions. But a tuned circuit that Andreas presents here and possibly impedance matching of the antenna will greatly increase reach.
Ah, and daylight saving time switch worked nicely last time it came around :)
Hi Hennes, nice to see you here :-) Thank you for your code! As you can see in the comments the „basement problem“ seems to be quite common... And thank you for your update about the winter time! BTW: Some viewer asked for an ESP32 version. Just that you know ;-)
You always seem to choose a topic for your videos that only a few days earlier I've been thinking about!
Here it was a longtime issue ;-)
I was thinking the same thing. Our timelines for building LoRa receivers were shockingly close.
Time ago I saw a way to test filters with an oscilloscope and a sweeping generator. Synchronizing the sweep start wit x axis gives you a Bode plot of the filter frequency response on the scope. It requires some manual adjustment on generator setup and a general idea of your frequency range, but it should work even in a low cost lab, especially for these frequencies. Cheers!
Interesting idea. Thanks!
Great vid and awesome tip on the txtempus software.
Glad it was helpful!
I waited for this video for a really long time.
:-)
To really easily test the LC resonant frequency, put a scope over the circuit and spark a small voltage over the circuit, the ringing that happens will have the frequency of the resonance.
Good Idea. Never tried it.
I disagree ....:-)... a grid dipper would be most likely an easier way to do so ... sorry just couldn't resist to say
Reminds me of the school friend who built a ttl 60 bit shift register and diskrete analog logic for his own clock (including segment drivers and leds for each bit position) on 4 euro boards. Worked perfectly once we got a stronger powerfully for all the LEDs.
Those were the days... But I see some "retro" stuff coming back. Maybe also such clocks...
Excellent video! Thank you!
:-)
Great idea! I have several of the "atomic clocks" but sometimes they do not set right the time from daylight savings time back to "normal time".
I think, when I'm done with the MakeLincoln - this Saturday, September 21st - 10AM - 4PM in Trago Park - I'll have to make one!
Thanks again Andreas!
Should not be too complicated, if you get the material.
@@AndreasSpiess I hope so! Sometimes I get stuck on a small error, that becomes a mountain! ha ha ha Thank you again Andreas!
The cat is what makes this video so good!
:-)
Extremely interesting, thanks, Andreas!
Was pondering for quite some time if it is not possible to do something like this (or a repeater) as we do not get any or super only weak reception for the Fort Collins signal here in Canada, BC. This solution, if run once a week for a short time of the day/night, should not disturb anyone at all. Perfect!
It should not bother anybody because they will be happy if their clock works now ;-). But the chance of reaching somebody’s clock outside your room is small.
Danke Andreas. Hab es gerade ausprobiert, funktioniert toll. Hab alle meine Funkuhren aus Deutschland (pre 2006) und die paar die ich hier in Canada gekauft habe erfolgreich gesetzt, und das mit nur breadboard connections. Werde das demnaechst richtig schoen verloeten und eine ferrit antenne verwenden.
Danke, danke, das loest ein altes problem was ich hier in BC hatte mit einem Schlag! :)
Mmm... Will try to wake up my old digitals Junghans (and a Brighton weather station). In this days with a lot of switching power supplys around us is very difficult to recevie the 77.5 Khz, even on nigth, even in winter... I love (electronic) clocks so let's do it!
Thanks Andreas for the compilation, explanation and fun.
I did not think on the power supplies. Very good point!
@@AndreasSpiess First step done! Three clocks in a row! Now waiting for ferrite rods and drawing the 50 Kw transmitter... ;-) That was a good pointer to Henner.
I use an app for ios (clock wave) who do something similar: Encode the ntp signal in wwvb (or dcf77) format and play it on the speaker of the phone at 20khz. An harmonic of the signal is broadcast by the speaker coil and sync a watch placed near it. I amplified the audio signal with an audio amp, created a loop antenna, and I can receive the signal few meters around. Anyway I Will try your set-up because I think it's a more "elegant" way to do it ;)
That is a very creative way of using a loudspeaker! I just bought the app and I am curious to use it when I have time.
Hi. This ESP stuff is very new to me. So, thank you for your tutorials. I also use TUNIOT from Adel Kassah.
Kind regards
You ae welcome. Adel does a great job!
Always: Thank you.
You are welcome.
Casio make various watches called "Wave Ceptor" which have a built-in receiver for all the various time signal transmitters around the world. The watch sets automatically once a day.
I recently bought one for about EUR 110. Really happy with it. Remains accurate to less than a second.
I did not know that Casio also produces such watches. Thank you!
@@AndreasSpiess The one I got is model WVA-M640TD-1AER and best price on Amazon de as below. Apparently very similar watches at prices up to EUR300+ and I don't understand why they vary so much.
One curiousity is that many of these watches do not have an all metal case. Mine is Titanium mostly, but includes some plastic as well. Presume that needed else the case would shield out all the RF.
Titanium case, black dial, solar powered, multi-band
www.amazon.de/Casio-Wave-Ceptor-Herren-Armbanduhr-M640TD/dp/B00EDAT9DM/
That is great ,i've never heard of that. Here in Greece we still use batteries !
It's not energy harvesting. It's for syncing the time.
So dcf77 clocķs use batteries also
@@jone1jone145 ahh i see
Hilarious dry humour. I really enjoy your videos and builds Andreas!
Thank you for you nice words!
very smart bro, congrats
Thank you
Спасибо! Направленная ферритовая антенна с усилителем = уверенный прием!
You are welcome!
Andreas, Great video as always. I continue to be amazed at the quality of your videos. I know you are retired but still youtube must take a lot of time nonetheless. Great job. A comment on your schematics: I would use a 560 Ohm resistor instead of 50 Ohm. Best Regards.
Thank you! A small correction. We get pension at the age of 65 and I am only 62. so I have to work for the next years...
Nice! Useful.
Thank you!
May I add an additional comment please ? In his introductory text ( above ) Andreas very kindly includes every possible link for further information - thanks to him for that ! His first link is to "Software" on Github. Just looking at this, there is an expanded version of the information that Andreas provides, including additional information which is a great help to some of us. Firstly the software HAS been tested on the Pi Zero WH ( I asked below ), which is cheaper to buy than bigger models and possibly easier to use in some respects. Secondly, for those of us in the MSF reception area, there is a suggested variation upon the resistors used in the external components. Thirdly, there are different ideas suggested specifically for wrist watches. So as well as taking in everything that Andreas tells us above, also take a look at the Github project page for the additional notes. Many thanks !
Thank you for summarizing the content of Henner‘s GitHub!
My favorite wayy to determine the resonance frequency of a RLC is to use XY mode of oscilloscope. Once you reach resonance, voltage from source and accross the resistor will come in phase. Then you get a line instead of of an ellipse displayed on the XY mode. This is pretty accurate
Good idea!
I have recently developed a probably unhealthy interest in these LF time code transmissions ...
However, I highly recommend other interested folk to look for a TH-cam video called "MSF Radio Clock Data Monitor start-up" which shows, apparently a "hardware" implementation of the decoding of the time code. I find this entirely captivating, which is probably why I describe my interest as "unhealthy"! It is also a good illustration of how these time codes actually work. The DCF77 and MSF codes are similar but slightly different (I am fortunate in that I can easily receive both services where I live)
Thank you Andreas for producing a video on this subject from your own (Swiss) perspective, and I always appreciate those little humorous touches you seem to sneak into most, if not all of your videos!
After all, apparently the Swiss are famous for their punctuality, and the thought of a Swiss guy "interfering" with time is a perfect example of your (mostly) subtle humour! 😂
I do not know how they implemented it, but it is certainly helpful for understanding the code. You could do something like that with a Neopixel ring, maybe even in color ;-) . And I am glad you like my content, of course.
Andreas you are very fortunate being a Swiss national when it comes to discussing time and time keeping instruments (horology). Horology is a hobby of mine so I am very familiar with the swiss heritage in making accurate clocks and watches (both automatic and quartz movements). I also believe I understand the Swiss Horology mindset.
I know you like to make interesting TH-cam videos, especially using off-the-shelf embedded systems (Raspberry Pi), but in my opinion the Swiss Horology solution for your IKEA clock is not to use an external radio timing reference (even one you make in your lab) or a satellite signal or any other signal external to your clock, but to have the timing accuracy inherent in the clock movement itself.
Your IKEA wall clock has a Chinese made radio controlled movement, which by itself should make you being Swiss cringe. Just look at your IKEA clock seconds hand, you can see how it is bouncing on the gear lash, meaning it does not even have a basic pressure backlash feature on the second hand gearing. Your clock should have a smooth moving second hand with no perceivable stop or start. This would also translate to both the minute and hour hands through the gear train.
I think you being Swiss should set yourself a goal of making your own movement with an accuracy of +/- 4 seconds per year. Which is not hard to do if you have to manually set your clock (calibrate) every 6 months or so when you come off and on daylight saving time. So, we are looking at +/- 2 seconds every 6 months. Which I think is doable with some newer OCXO (SiT5711) on the market without a phase lock correction source or on-going calibration. The clock is indoors so you don't have to worry about temperature errors, and you can provide a hyper stable supply voltage and a OCXO consistent load, so your only real obstacle is aging.
Using a OCXO will mean you will not be running off of a battery but AC power, and if you lose power you will have to reset your clock, unless of course you have a backup battery power source (Preferably rechargeable). You can use discrete logic to divide down your 10MHz clock source to give you the pulse rate you need, or perhaps use a programmable logic IC, which would then drive the stepper motor power amplifier.
You can hack the movement in your clock and keep all of the basic gearing and just power the stepper coil. You can even make your PCB the same size as the original PCB in the clock and just notch the case for a connector (perhaps USB-C). If you want a smooth running seconds hand you can hack a similar sized Seiko wall clock movement with the same hand mounting shaft diameters, or use another set of hands that fits the posts (many options are available) where the only difference is in how the gearing is done relative to the stepper motor pulses, from one (your IKEA clock) to eight steps per second. Here again you can just take out the original PCB and put in your own improved version. If you want even smoother movement you can add your own micro-stepping stepper motor and upgraded micro stepping driver. Micro-stepping is perfect for a clock movement because you are not challenge by RPM speed or torque. But, if I where Swiss I would hack an indoor swiss made MOBATIME indoor wall clock movement and toss the IKEA clock which has obvious problems.
You are Swiss, you are self reliant, do not rely on outside foreign signals to accurately run your clock. "Unus pro Omnibus, Omnes pro Uno"
Thank you for your ideas. I made videos about precise clocks where it matters (satellite communication on 10GHz). Search for QO-100 if you are interested. But as an engineer, I live the slogan "good enough". This crappy IKEA clock is "good enough" for the purpose of a lab clock as long as it does not have a trend. In addition, I am not very good at mechanics :-( So somebody else has to do this hack.
Great video
Thank you!
Can use the FFT function on your scope to characterise Filters, Anntennas, etc. Obviously alongside an ARB/Frequency generator.
Not as good as a decent Spectrum Analyzer, but more than good enough to get the job done in these simple cases.
I tried with a noise source and the FFT. But did not see a dip :-( I do not know what went wrong...
Well explained as usual. Just one thing to add, if you buy one of the simple analogue ones, those may not allow for an offset. So for example, a DCF77 clock will show he wrong time in the UK as we one hour behind Frankfurt. I have one of those (longer story) and it works fine in the North West of the UK, but the only way i could adjust it to the correct hour was taking it apart and slip the hour pointer an hour backwards.
With regards to transmitting on random frequencies, if i remember that correct, anything above -10dBm ERP is supposed to be an illegal transmission (mileage may vary, depending on country), obviously except you got a licence for what you doing.
Because you create your own time signal here you can send UK time. Concerning ERP: Well possible that such a maximum value exists also in other countries. This device then should be well below this value.
To check your LC tank circuit resonant frequency you can use a Grid Dip Meter.
Good Idea. Thank you!
Microsemi sells a Chip Scale Atomic Clock (CSAC) for about $2000. Also it might be interesting to use GPS time to set the clock on the Raspberry Pi and then transmit that to your clocks.
Interesting. Yes, you could use GPS. But NCP is exact enough for me. And simpler. Especially in the basement. But you might see a GPS for another purpose on this channel in the future...
Super 👍 I like it
Thank you!
I have got me a little ESP that gets the time over ntp and simulats the dcf77 signal and puts ist directly in the clock works great never had problems
Do you have a link to the code you use?
@@AndreasSpiessI have modiefied this code for my timezone ntp and the esp8266. It's from some guy in the internet I can't find anymore
Link: drive.google.com/open?id=1htHZdS5-wic5BXFYRgLzpwAN1tLwk3-3
Thank you for the link. In the meantime another viewer pointed me to a ESP32 project ;-)
Always interesting videos, and this one is particularly useful since I found an atomic clock among my late Father's things and had no idea what the slider switch on the back were for: setting the signal to synchronize with! Now I'll set it to the closest source and wait for it to adjust itself. Any insights on how they adjust for time zones?
By calculation. It really is that simple. It transmits the native information based on the clock origin. The clock itself is then programmed with known DST regions.
These clocks work fine in areas with the same "time principles" As Doug pointed out they include "summertime" or "daylight saving time". And in Europe, most of the countries in reach are in the same time zone. So often, these clocks do not need an adjustment for time zones.
Beautiful cat!
Thank you!
Nice video, let's try to change the clock of our classroom :-D
Slower to learn more?
@@AndreasSpiess Of course ;-)
Terrific!!!
Thank you!
Thanks a lot Andreas, I'll not need to travel thousands kilometers just to keep my Radio Controlled wristwatch synchronized anymore. 😁
Good for the environment ;-)
@Andreas Spiess - I watched the video several times, but it is still not clear to me what you are using as a stable time reference to generate the synchronising signal for your radio-controlled clocks. Is the Raspberry Pi connected to the Internet and thus being synchronised to a time server?
Yes
Great project Andreas! I begun to recreate the project with an orange pi zero, but and it don't have GPIO 4 or 17 on this Pi. So I think that I can change thoses Gpios on the program for one's that Orange Pi have. Do you know where I can modify the corresponding lines ? I found the instruction for the GPIO 4 (gpio.cc line 184) but I don't find the gpio17 location. (I'm a real newbie ;) Any help ?
I do not know and I am also a newbee in many cases, This is why I only use Raspberries ;-)
A couple of extra things.... ferrites are very directional.... also lots of noise at 60kHz due to computer equipment (switch mode PSUs, monitors etc)... nice video.
I agree on both statements!
Hi again Andreas
Your circuit works perfectly well, but I note that in most RF schematics, the coil is parallel to the capacitor. Yours are in series. Is there an advantage to do this ?
You can try both ways and see which one is better for you. I did not test.
Interesting. One thing I I wondered , having been recently bitten by it, is if it would be best to specify a COG or NPO capacitor for the resonant circuit. Both of these are class 1 temperature stable whereas some of the others have very strong temperature dependence, a class 2 x7r is not as good but still a lot better than the class 3 & lower spec etc. If the resonant circuit is always at the same temperature this would not matter & may not matter anyway in this application but it made me wonder having been troubled by it in something I was designing that can likely see temperature fluctuations from cold winter, say minus several centigrade up to high summer of say 30 centigrade. Thanks for sharing!
You are right. As you wrote, in this application temperature coifficient is not as important as the temperature does not change too much in the lab and also the application is not too sensitive.
Hi Andreas, I have a couple of DCF77 receivers for arduino. Can I cut the wires of the antenna of one of those receivers and use that as antenna for the transmitter without the need to purchase a ferrite bar and wind the wires as in the video ?
Should be possible
Would it be possible to add or remove winding of an antenna to match the frequency perfectly? Since it is hard to get a capacitor with a precise desired value, maybe tune the coil?
Yes. For me it was easier to select a capacitor.
@@AndreasSpiess Good to have a collection! Thank you for the reply and for the very interesting video! Previously I thought only GPS broadcasts their atomic clocks in the air.
Now you only need to add some sort of monitoring to the transmitter, otherwise there is a risk that it will break without anyone noticing, and time will drift again
You are right!
to see the signal, do you connect the second ferrite rod with a capacitor of 1nf ? do you also use 200 turns of wire?
You can use the same antenna also for receiving. But you have to test it as I showed. It might be that your parts value are different.
@@AndreasSpiess it worked!
I could see the signal with sdr#. Lot's of harmonics though..
Verry interesting video.
When you have a safe with a time lock i hope the time does not sync with this radio time signals.
No need for a safe here. Poor guy ;-)
Hi Andreas, thank you for these clear explanation. I successfully bought an emitter from your schematic for the JJY 40kHz frequency using a loop antenna.
A question : why is there a 1 nH coil in your schematic ? Shouldn't it be 4.4 mH ? An 1nH antenna and 1nF capacitor should give a 159 Mhz frequency.
Yes, as shown in the video, 4.4mH is correct. The drawing you refer was only to show the method.
Great video, I'm having the same problem with my clock in Canada. If it can't connect for several days it loses a couple of seconds. My next project is to follow in building one for my clock.
I hope you will be successful!
Another great video as always. I guess I would have gone with an air core multi turn loop antenna for the low power 77.5, 60, or 40 kHz transmitter. The ferrous core is fine for the receiver where saturation isn't an issue. Either way, the radiation efficiency will be awful. As an alternative to winding the antenna, I've used the magnet wire right on its original reel. This is nice because you don't have to do the winding (boring), and you can try with or without the ferrous core by sliding it in or not...
I thought about your proposal. Bot was not able to locate the other end of the wire without unwinding the whole coil :-(
@@AndreasSpiess In some cases, the manufacturer makes it really easy to access the inside end of the wire on the spool. Here is an example: www.alliedelec.com/product/rs-pro/357772/70825771/ In other cases it is almost impossible. I frequently use small diameter multi-loop antennas, and I am way too lazy to rewind, so I restrict myself to the products that make it easy. The ability to easily slide amidon rod cores through the reel for testing is another plus.
Good to know. I purchased these wires without having a particular purpose in mind...
Great Work Andreas! Interesting work! I would be great if it could be do with a esp8266 or esp32. but in Vienna the dcf Signal are not realy working. Most of the clock don'T syncronize...
I was envy of northern europe people because my DCF77 clocks sync very very difficult, and I'm in southern italy!
You could try to port Henner‘s C code to an ESP. I run it on my node-red server which is always on
same thought. Really looks like Raspberry is an overkill for such an easy task
Other viewers posted links to an ESP8266 and an ESP32 project for DCF77
I do also like the Bodet clocks in the hospitals which are in a network all in sync. Are these also in sync with DCF77? scavange those FM/ AM old radios for Ferrite core and wire coil . Ahhh Qucs :P ..or Qucs-s? Nice tutorial
I do not know about the hospital clocks. And ferrite antennas of an old radio is a good idea!
Good morning Andreas Spiess. Even tough I live in Switzerland, I am asking my question in English, since some visitors of your project may encounter the same issue.
My question is about the radio controlled Citizen Pro Master Land watch, which officially syncs with JJY signals only.
Since many TH-camrs claimed, that this watch can be sync. easily outside Japan using radio sync apps like Clock Wave or Radio Wave Sync (from Apple store) with an iPhone I bought the watch.
But with neither of these apps I am able to sync the CPML with my iPhone (OS 12.5.7) to set the lokal time here in Switzerland. I am following the instructions and the apps generate the sync sound.
Do you maybe know an app (for iPhone or iPad), which I could use or a «go around» to sync this watch on the fly?
Thank you for your help.
No, I do not know another app. Anyway, these apps only work "by chance" because they use the coil of the loudspeaker to create an electromagnetic field. This is only very small and will not bridge big distances... I assume they differ from model to model, too.
My work in Alaska, USA purchased an atomic clock and we discovered it doesn’t work up here. Is it because it is out of range or some other reason? Can we make an antenna that would work? Thank you for the video.
I think it is out of range. No antenna can help then. But this project could help
Thanks!
Thank you for your support!
I'm currently working on a radio project of this type as a student. Do you know what type of antenna is inside those clocks? Ferrite rods maybe?
Maybe I will do a teardown of the IKEA clock. So far I do not know.
I have a digital wall clock bought from aldi... The great advantage in this application is that it will run for many years on a set of AA batteries. The date and time is transmitted at one bit per second so it's a very low processing load. I don't think GPS or NTP or any other exotic system can be received and decoded using as little energy... though I'm happy to be correted if anyone knows better.
You are right. This is why I like the concept.
Hi Andreas Spiess and thanks for a outstanding project.
I already did the basic experiment (with the 3 resistors only and a few centimeters of coiled electric wire) with one of my RPI and I can adjust the clocks by almost touching the clock's DCF antenna to the power cord. I already bought a 77.5kHz Fine Tuned Loop Stick Antenna for DCF77 Receiver at eBay, from Canada and while I'm waiting for it to arrive, I take this opportunity to ask the following questions:
What could be the coverage of the DCF signal, using the basic circuit, but with the DCF77 antenna (the length of this antenna is 60mm, with a core diameter of 10mm)?
What could be the coverage of the DCF signal, using your modified circuit with the PNP transistor (3.3Volt) and the DCF77 antenna?
What could be the coverage of the DCF signal, using your modified circuit with the PNP transistor (5.0Volt) and the DCF77 antenna?
Many thanks in advance and best regards.
Rui, CT1AIC
I can only share my setup with a transistor and a ferrite antenna. Its reach is maybe 2 meters in my lab (which has lots of electronic noise).
@@AndreasSpiess Hi and thanks for the kind answer. What I would need is coverage of 10 to 15 meters with some walls between, where I have some DCF77 clocks. At Portugal, I only can get the signal of Frankfourt from time to time and at night, at a window, directed to Germany. You can imagine...
I then would forsee more voltage. And it will probably be illegal, if you care.
Thinking can use an upconverter with an RTL-SDR and Spektrum with a noise generator for a more cost effective tuning method? Maybe first verifying the source signal to adjust for frequency offset once the RTL-SDR has stabilized. Cool build.
Did you try a noise source with an SDR? I tried it without success. Maybe I made a mistake.
@@AndreasSpiess I still have to try... I've only read about and watched videos regarding. www.rtl-sdr.com/rtl-sdr-tutorial-measuring-filter-characteristics-and-antenna-vswr-with-an-rtl-sdr-and-noise-source/
General search with a few references that might help:
www.rtl-sdr.com/tag/swr/
Do you think since low frequency measuring was the issue... I was wondering if direct sampling RTL-SDR alone will work?
www.rtl-sdr.com/rtl-sdr-direct-sampling-mode/
In the meantime, I tried again with the noise source of my oscilloscope and it worked. Hat to get the right FFT settings of the oscilloscope.
@@AndreasSpiess Awesome! Please share. This RTL-SDR or other SDR method appears to be the most cost effective I've found and is on my list of amazingly cost effective procedures to develop this fall/winter as I'm currently focused now on other projects tasks.
While on the topic of cost effective uses for the RTL-SDR, there is an oscilloscope plugin for SDR# also (albeit I'm not sure what the operating accuracy specifications will be) on the web that I've downloaded and haven't played with much from IN3IEX:
www.amfone.net/Amforum/index.php?topic=34858
qsl.net/e/ea1kr//sdr/rtlsdr/
I also just found this reference to SDRSharper doing a Google search to find the above links and will try when I get the time:
www.qsl.net/s/sdr/
On another note, I just found this reference to use an oscilloscope as an SDR which is something I've wanted to try with my TDS-520 test bed frankenscope either directly or with a pan adapter:
www.rtl-sdr.com/gr-oscilloscope-using-an-oscilloscope-as-a-software-defined-radio/
I will not do too much RF on this channel. I cover many other topics, too. Concerning using the ADC of the oscilloscope: Maybe an interesting experiment, but probably not very useful for everyday's operation :-(
I'm planning on using a RPi zero W and a ferrite core antenna taken from a WWVB clock to transmit the JJY 60 signal. I can solder and follow instructions but I have very basic electronics knowledge! Why can't I just put my ferrite core antenna in place of the wire in Henner Zeller's setup? I don't understand the circuit at 10:00 in the video (I'm only assuming it's isolating the antenna from the pi and allowing the signal to be boosted), but my goal is to transmit with enough power to cover a small room (I am in a country where this is allowed of course ;-p). Is there somewhere you could point me to where I could learn the function of your circuit at 10:00 or give me any tips/instructions? Thanks!
Coils can create high voltages and Raspbery pins do not like that. This is the reason for my comment. The function of this circuit is best understood if you look for a basics transistor tutorial. One of the pins creates the high frequency signal and the second "overrides" this signal to switch it off in the right moment. You probably need an oscilloscope to play around with this circuit and learn how it works.
How about building a simple amplified repeater from the cheap receiver module ?
The problem is that you have to receive and transmit on the same frequency.
To you manage that you can cross out simple.
@@ZZtop-gg3lu i meant a repeater the tx'down a coax to his basement
Thanks for the great video! I just built this DCF77 transmitter. I have one problem with it which is the low field strength: my clock can receive the radio signal only from a few centimeters. I need to have it in my whole flat: cca 10 meters. I have cut off the antenna from a ready-made DCF77 receiver then I attached it to my wemos d1 r2 mini with a 1nF capacitor and a BC547A transistor. Unfortunately I cannot measure the induction of the ferrite antenna I have. Can I suppose it have the same induction as in your case (145mH)? I tried to increase the voltage on the collector of the transistor up to 10 V but didn't help. I have no spectrum analyzer or oscilloscope. Any idea? Thanks!
These antennas are directional. Maybe turning it in the right direction helps.
@@AndreasSpiess I tried to rotate both horizontally and vertically but nothing changed unfortunately.
Now I am hopefully on the right way. I bought a measurement device to measure my ferrite antenna and calculated the right capacity. Now it covers a much bigger area but still not as much as I need so I need to improve a bit further :)
thank you for the cheap method of the measurement. btw. can the antenna be smaller?
I think so. But I did not try to cut this brittle metal.
@@AndreasSpiess no no. different type. there are hand watches. what kind of antenna do they have?
I assume a similar one, but I do not know.
Great video, but one part I don't understand. In your circuit you have a 1nF capacitor in series with the coil. I tried googling it but all the articles talk about receivers, not transmitters. I'm assuming it has something to do with tuning the circuit. Can you explain and/or give me some keywords to more properly search? It isn't on the original author's schematic. I'm interested because I'm moving to Alaska and WWVB cannot be received there according to their signal maps. Thanks!
I forgot to mention, I bought a decoder circuit off AliExpress and plan to use that antenna. It was easier than trying to to DIY it (I don't have the enameled wire, etc.) I bought the 60khz version so the antenna should match.
As shown in the video, it has to do with tuning. You can try without capacitor. Maybe it works. Otherwise I showed how you can tune it.
@@AndreasSpiess Ah! Thank you, I didn't understand that part, I'll go back and re-watch. For what it's worth, it did not work with the capacitor installed. I removed it, per your suggestion and assuming the antenna was reasonably closely tuned, and it worked perfectly. I have learned so much from your channel!
To swiss fir me to even consider doing it )))
Very interesting.
I used to play with DCF77 but it's rare to receive it in the Middle East.
What's the name of the IKEA clock?
Its name is Burrig
@@AndreasSpiess Thank you.
Seems to be sold in Europe only.
Couldn't find it on my local IKEA website.
I've known about this technique for updating clocks for some time, its great, but what I really want to know is how those analogue clocks know where the hands are?
Can you do a video where you take one apart please?
They use a Stepper Motor as all electronic clocks. During reset they put all three hands at 12 o clock. From there they count the minutes and seconds, I assume. But I never took one apart.
I have this problem here in the south of Spain. It's a Braun exactly like yours.
I can imagine because you are at the end of the reach...
For viewers in North America, the relatively new La Crosse 404-1235UA model picks-up the newer Phase Modulated signal from WWVB, which provides something like +17dB (x50) better signal. So even in eastern Canada, about 4000 km from WWVB in Colorado, the signal is effectively very 'strong' and the clock just works very reliably. This model works in almost all Time Zones (not limited to the four USA zones), does automatic DST, and even Leap Seconds! Check the reviews for further details before making any decisions..
Thank you for your info. Good to know you can buy a good device BTW: At least DCF77 includes "summertime" as it is called here and all signals should include the leap seconds. This is one of their purposes.
404-1235ua la crosse
An interesting way to test the resonant frequency would be by giving the system an impulse, and observing the resulting oscillation frequency. By giving an impulse I mean using a very low frequency square wave, or in other words, by just applying a voltage to the circuit and wathcing what happens at the output voltage.
Of course you still need an oscilloscope for this, but not a function generator.
Another viewer suggested the same method. But it is not so easy to do this fast. You have to give it energy and then, go high impedance to let it ring. Maybe with an Arduino pin, this would be possible
Regarding the coil and the ferrite rod, I'd first look for a broken small radio-controlled clock, and salvage those items from it. A lot of cheap ones have been imported over the years, particularly by the German discount retailers like ALDI, and on sale for as little as £5. When they get knocked onto the floor, the displays break or the buttons jam. Maybe the battery compartment terminals become broken or corroded. A broken one seems an ideal opportunity to salvage the ferrite and coil for a project such as this. The only thing for us in the UK is to look for a 60 kHz model, as that signal is of course stronger than 77.5 kHz over here. A really interesting project, Andreas, and has made us all think about something a bit different for a change ! If you read this, would a Pi model Zero be as equally good at running the application as the Pi 4 ?
Thanks !
A Pi Zero should work. But I did not check. I used a 3B+
@@AndreasSpiess Many thanks for taking the time to comment. If I try one, I'll let you know !
Now I see that Einstein got so pissed as lowly at Swiss patent examiner that he thought of relativity to put all clocks in doubt.
At least I can say that he studied at the same university as I did ;-)
Is it possible to display the 60 MHz WWVB original carrier on the scope? Thanks
Only if you add an amplifier. And the signal is on 60kHz, not MHz which makes it easier...
I think some low-frequency capable SDR will be the best cheap transmitter tester... :)
Not many are available which also transmit on these frequencies...
But I was talking about testing with SDR, instead of using an Oscilloscope / Signal generator... No need to transmit. With proper signal up converters, it should be possible to do the testing even with a cheap TV stick!
I know of at least SDRPlay, which is less than 150 €.. Of course for that price it is also possible to get an oscilloscope.
The antenna does not work without a signal. You cannot receive anything if you do not put power into the antenna. You need a transmitter with a variable frequency for that test.
rkan2 is responding to the part of the video where you ask for cheaper methods of verifying the antenna is transmitting on the correct frequency. With a $25 RTLSDR and some free frequency analyzer software you can see what the transmitter is doing easily.
I still insist that you have to have a variable frequency transmitter to test and optimize the antenna. Of course, you can test the 77.5kHz signal strength with just an SDR receiver. But this can be done with an oscilloscope, too.
You can also do this with a Arduino for lower Energy consumption
There is a project from Blinkenlight. (blog.blinkenlight.net/experiments/dcf77/dcf77-generator/ ). But the Arduino does not know the time.
I would like to know how you got the cat to stay behind the oscilloscope and not directly in front of it.
She was in front of it till she got enough „attention“ ;-)
@@AndreasSpiess I prefer to believe it was a specially selected Swiss lab cat.
How about the GPS signal from the caesium and rubidium clocks in America. There was 13 or 15 in a room which feed the satellite system. I remember it as we used it in the lab and truck to calibrate time with tge counters and spectrum analyzers.
GPS would be good. But my problem was more the display with this analog clock without wires...
@@AndreasSpiess I can see your point. Have a great week.
9:10 I appreciate the interference by kitty :)
She does what she wants ;-)
@@AndreasSpiess no radio control for her yet?
Actually I'm thinking of a project with an ESP8266 to get NTP time and transmit signal to the radio controlled clock, but I have no idea on how to make the transmitter and implement the corresponding encoding protocol.
You can have a look at Henner´s program. It is in C and might be ported to a ESP.
I am also thinking NTP-to-RadioClock.
Would be a great project @Andreas Spiess!
What about to use an antenna from a dcf77 clock mechanism you can find also on Amazon?
Should be ok.
@@AndreasSpiess i have removed the antenna/inductance from the clock mechanism and it measure only 1.2mH so i need a capacitor of 3.51nF =~ 3.50nF is correct? i have a signal/function generator and since i do not have a single 3.5nF capacitor i have used some series+parallel to achive the correct value, now i have not understand how to see if the circuit is ok usin the generator+oscilloscope, i have generated a signal 1Volt / 77.5Khz (sine ? square? no matter?), what i had to 'see' on the oscilloscope if all is ok and what if not? (i'm using an Hantek 2D42 that output the signal AND also measure it) Thanks !
TNX 4 another great video !!!
73 N8AUM
:-)
Does anyone know if this system could be made to work with Primex clocks to avoid buying their expensive transmitter? I've not yet found anything on the primexinc.com site that details their protocol.
Primex seem to use 72 MHz signals (according FCC info). I assume they use a proprietary protocol.
Isn't your weak radio signal competing with the much stronger one from the DCF77 station?
How does your Ikea clock know which signal to sync to?
Maybe I missed that part, but I am confused as to why he would want a different time in his lab, other than the one from DCF77
His clock does not reliably receive the DCF77 signal in that basement lab. (His clock doesn't sync to it or even syncs wrongly.) So that main signal will be hardly above noise floor in there. His weak makeshift signal will be way stronger than what it receives from DCF77.
@@CarolineC0815 That makes sense, but that also makes it a very specific solution for clocks at isolated locations.
Best not get a radio clock for that location.
Your comment brings up more questions than it answers. :-)
I still enjoy the video, its always educational.
Anyway, thanks for the response
Think repeater radio
@@DougHanchard As if you have read the reply i typed but deleted. :-)
I was going to say:
The basement clock will never be more accurate than the clock source you use for your local transmitter.
(I guess that's good enough for home/office use)
But there will be a delay between receiving the DCF77 and the retransmission.
Also: If the original DCF77 signal can be received in the basement, but not reliably, you will still have interference between the two, unless your local transmitter drowns out the DCF77. (which is of course possible)
I was also wondering if the Raspberry would be powered on 24/7 just to avoid clock to get corrupt time-data.
It seems a bit of overkill. Since he probably has internet in the basement, he could use a clock that syncs with an NTP server once per hour/day.
Or, since Andreas seems to know quite a bit about Radios, wouldn't it be possible to receive the signal outside and re-amplify it in the basement? (without using a raspberry)
(I guess that's what you mean by repeater radio ?)