eblc1388
Active Member
How difficult is it to build a 77.5KHz receiver? More like an audio amplifier, I suppose?
Lately I have learnt how to program a PIC to work as a digital clock but I have to set the time up every time I powered the PIC up. So I thought, why not let the DCF77 radio time station do that.
DCF77 receiver modules are available cheap but I want to build one nonetheless because those modules do not provide a 77.5KHz carrier output but only the timing signal pulse. If I build one myself, I can also use the carrier as a frequency reference to cablibrate my frequency counter.
My first attempt is a simple LC tune circuit with transistors RF front end. It kinda work but there are lots of signals at the output, with varying frequencies. Obviously the selectiveity(Q) is not good and all those interference signals from my desk CPF lamp, the printer, my PC and ADSL modem all give out interference at approximate that frequency range, which I turned OFF one by one to confirm. Even my oscilloscope does so when I move the receiver closer to it but I can't turn my CRO OFF. It is very easy to tell by just moving the receiver closer to the noise source and watch the scope amplitute increases. I cannot separate the wanted signal from these interference.
My second attempt is a double-tuned LC stage which has a very high Q. In order to tune it, My frequency generator simply would not do. I need a stable frequency source of 77.5KHz so I built one using a 8-pin PIC and finally got one to output 77.501KHz, which is good enough for all tuning purposes. With this signal source, I finally trimmed the capacitors on the double LC stage to let the receiver circuit resonant at exactly 77.5KHz.
Now I can see a beat on the received signal(between my source 77.501KHz and the real 77.500KHz) on my scope. So far so good.
After several more stages of amplification and AGC, I finally got a clean blinking trace every second on the CRO. The DCF77 carrier output by the radio station falls to 25% every second to signal a ZERO or ONE. Even with AGC, the output amplitude of my circuit changes a little bit with the modulation. If I fed this signal into my frequency counter and select 0.1s time gate, I get a stable 77.50KHz display when there is no modulation but the value changes when the amplitude dips. My AGC is not fast enough to react to the 100%/25% amplitude changes.
These problems can be solved, with additional circuit I'm sure.
However, the nail in the coffin is a strange signal from evening onward which is about 78.125KHz and the amplitude is very stable. Because of AGC, the DCF77 signal is reduced to like noise in the background. I thought my circuit is going into oscillation because of its high Q. It took me sometimes to realise that the signal is TV related, namely the harmonic of 15,625KHz, while my TV is some 30 feet away. The signal reduced by half when I finally switched my TV OFF but it is still there. I can't ask my neighbour to switch their TV OFF.
So in all these are my findings:
a) Interference from nearby source is a very very big problem
b) the receiver circuit must have a very narrow bandwidth
c) usually a crystal of the same frequency is required to filter the signal
d) there is absolutely no point trying to build one
e) I found out that the commerical DCF77 front end receiver has a bandwidth of just +/- 1.5Hz from 77.500KHz !!! :twisted:
Lately I have learnt how to program a PIC to work as a digital clock but I have to set the time up every time I powered the PIC up. So I thought, why not let the DCF77 radio time station do that.
DCF77 receiver modules are available cheap but I want to build one nonetheless because those modules do not provide a 77.5KHz carrier output but only the timing signal pulse. If I build one myself, I can also use the carrier as a frequency reference to cablibrate my frequency counter.
My first attempt is a simple LC tune circuit with transistors RF front end. It kinda work but there are lots of signals at the output, with varying frequencies. Obviously the selectiveity(Q) is not good and all those interference signals from my desk CPF lamp, the printer, my PC and ADSL modem all give out interference at approximate that frequency range, which I turned OFF one by one to confirm. Even my oscilloscope does so when I move the receiver closer to it but I can't turn my CRO OFF. It is very easy to tell by just moving the receiver closer to the noise source and watch the scope amplitute increases. I cannot separate the wanted signal from these interference.
My second attempt is a double-tuned LC stage which has a very high Q. In order to tune it, My frequency generator simply would not do. I need a stable frequency source of 77.5KHz so I built one using a 8-pin PIC and finally got one to output 77.501KHz, which is good enough for all tuning purposes. With this signal source, I finally trimmed the capacitors on the double LC stage to let the receiver circuit resonant at exactly 77.5KHz.
Now I can see a beat on the received signal(between my source 77.501KHz and the real 77.500KHz) on my scope. So far so good.
After several more stages of amplification and AGC, I finally got a clean blinking trace every second on the CRO. The DCF77 carrier output by the radio station falls to 25% every second to signal a ZERO or ONE. Even with AGC, the output amplitude of my circuit changes a little bit with the modulation. If I fed this signal into my frequency counter and select 0.1s time gate, I get a stable 77.50KHz display when there is no modulation but the value changes when the amplitude dips. My AGC is not fast enough to react to the 100%/25% amplitude changes.
These problems can be solved, with additional circuit I'm sure.
However, the nail in the coffin is a strange signal from evening onward which is about 78.125KHz and the amplitude is very stable. Because of AGC, the DCF77 signal is reduced to like noise in the background. I thought my circuit is going into oscillation because of its high Q. It took me sometimes to realise that the signal is TV related, namely the harmonic of 15,625KHz, while my TV is some 30 feet away. The signal reduced by half when I finally switched my TV OFF but it is still there. I can't ask my neighbour to switch their TV OFF.
So in all these are my findings:
a) Interference from nearby source is a very very big problem
b) the receiver circuit must have a very narrow bandwidth
c) usually a crystal of the same frequency is required to filter the signal
d) there is absolutely no point trying to build one
e) I found out that the commerical DCF77 front end receiver has a bandwidth of just +/- 1.5Hz from 77.500KHz !!! :twisted: