Radios, Radio Antennae, Reception, Transmission

[mearvk]

So couple questions.

My assumption is that a radio antenna gets bombarded with all frequencies at the same time. So there has to be a discriminating circuit with a radio to act as a frequency filter.

Well could I do this naively with a low pass filter and a high pass filter together? In other words, if I made a tuned, static (no ability to change it) band-pass filter (say at 100MHz +/- 20Khz) and added another low pass filter at 20 kHz and then hooked that into an amplifier of some sort would I be able to listen to a radio station of my choosing? Adding variable resistors would allow me to tune it I guess. I guess the demodulation might be more complicated than that.

In the age of ICs how is the frequency generation for FM band transmission typically accomplished? There are frequency crystals which have some minor ability to vary, there are MCUs which can use PLLs but it isn't clear that a MCU w/ PLL @ 96 MHz can vary enough to really handle the FM spectrum. I'm really unclear on that. I don't want to mess with tank circuits since they sort of beg the question and can be unstable. So what does that leave?

Edit: Basic, basic question. How does one wire an antenna into one's circuit?

Thanks.

 

[nsaspook]

You might want to look at SDR radio for direct conversion circuits without traditional tuned circuits. The actual hardware can be very simple but you will need a large about of CPU power for the DSP demodulation functions.
http://www.vk6fh.com/vk6fh/SDRradio.htm

Most of the systems today are using cheap USB sticks for receiver front-ends.
http://sdrsharp.com/
http://osmocom.rtlsdr.org//trac/wiki/rtl-sdr

 

[davenn]

So couple questions.

My assumption is that a radio antenna gets bombarded with all frequencies at the same time. So there has to be a discriminating circuit with a radio to act as a frequency filter.

yes, its called a tuner, a tuned circuit. Also having an antenna cut to resonate at the frequency of interest also helps to filter out other frequencies to a smaller extent

Well could I do this naively with a low pass filter and a high pass filter together? In other words, if I made a tuned, static (no ability to change it) band-pass filter (say at 100MHz +/- 20Khz) and added another low pass filter at 20 kHz and then hooked that into an amplifier of some sort would I be able to listen to a radio station of my choosing? Adding variable resistors would allow me to tune it I guess. I guess the demodulation might be more complicated than that.

thats right its more complicated than that ... you need a demodulator circuit, this may be as simple as a detector diode for an AM signal, to much more complex circuits to recover a digitial signal

In the age of ICs how is the frequency generation for FM band transmission typically accomplished? There are frequency crystals which have some minor ability to vary, there are MCUs which can use PLLs but it isn't clear that a MCU w/ PLL @ 96 MHz can vary enough to really handle the FM spectrum. I'm really unclear on that. I don't want to mess with tank circuits since they sort of beg the question and can be unstable. So what does that leave?

All crystal oscillators have the ability to be varied in frequency a little and even with a PLL the same can be done. Producing an FM signal is pretty simple

here's an example...

In the above FM transmitter circuit ( this is just the Colpitts oscillator stage) the modulation is applied to D1 which is a varicap diode. The modulating voltage causes the varicap diode to vary in capacitance slightly in proportion to the modulating voltage. This variation in capacitance allows the otherwise fixed freq of the crystal oscillator to change.
Again, the freq varies proportionally to the variation in capacitance which is proportional to the modulating voltage.

Edit: Basic, basic question. How does one wire an antenna into one's circuit?
Thanks.

the antenna gets connected to the input of the receiver ... that is, the initial tuned circuit


Dave

 

[davenn]

for a PLL system here's an example that a fellow amateur and I built and experimented with some years back on 1296MHz (23cm Amateur band)

the PLL consists of a VCO, upper left, a prescaler IC ( MC12022A) to bring the freq down to what the PLL IC (MC145152P2) could handle. The PLL chip produces phase difference voltages out of pins 7 and 8 and these are summed up in the Op-Amp (TL081) to produce an error correcting voltage that is fed back to the VCO module.
You can see that it at that point, between the Op-Amp and the VCO that we insert the audio modulation ( in the lower left corner of the diagram)

Dave

 

[mearvk]

Thanks to you both for replying.

If I understand you correctly my question would be why would we need two pins and an op-amp to feed a voltage back to the VCO when presumably it could all be done in the PLL chip.

I find the whole thing a bit disheartening TBH. It seems to me that, for instance, you could take a 20 MHz crystal, feed it into a voltage controlled (1v = 1x, 5v = 5x, up to some limit) PLL chip, use a voltage divider with a potentiometer, drop a 1/10th frequency output pin on the PLL for low-end oscilloscopes and be done with the synthesis side. Instead we have 28 pin DIPs and datasheets whose target audience might as well be 15th century Mongolians.

Edit: it occurred to me as I was watching Hell's Kitchen that if you have a VCO locked at 1296 MHz, why do you need a PLL? I thought PLLs were mainly used for frequency multiplication.

 

[davenn]

Thanks to you both for replying.

If I understand you correctly my question would be why would we need two pins and an op-amp to feed a voltage back to the VCO when presumably it could all be done in the PLL chip.

some may well have that stage inside them, some dont, this particular one doesnt...
no big deal

I find the whole thing a bit disheartening TBH. It seems to me that, for instance, you could take a 20 MHz crystal, feed it into a voltage controlled (1v = 1x, 5v = 5x, up to some limit) PLL chip, use a voltage divider with a potentiometer, drop a 1/10th frequency output pin on the PLL for low-end oscilloscopes and be done with the synthesis side. Instead we have 28 pin DIPs and datasheets whose target audience might as well be 15th century Mongolians.

That didnt really make any sense to me... I have no idea what you were getting at ?

Edit: it occurred to me as I was watching Hell's Kitchen that if you have a VCO locked at 1296 MHz, why do you need a PLL? I thought PLLs were mainly used for frequency multiplication.

Do you not understand the purpose of a PLL/synthesiser system ?
PLL = Phase Locked Loop ... the loop in the above circuit is the one from the VCO to the prescaler to the PLL/Synth through the Op-Amp and back to the VCO

The VCO is only locked to a freq because of the action of the loop.
If it wasnt for that, the VCO would not stay on a fixed freq with any sort of accuracy. The PLL/synth chip, in this case the MC145152, is used to lock the VCO to a desired freq and to also be able to control freq (channel) changing

A lot of the newer LMX... series PLL/synth chips, by National Semi., are capable of much higher freq operation on their own as they have built in prescalers and will operate happily up to several GHz

cheers
Dave

 

[sophiecentaur]

Thanks to you both for replying.

If I understand you correctly my question would be why would we need two pins and an op-amp to feed a voltage back to the VCO when presumably it could all be done in the PLL chip.

I find the whole thing a bit disheartening TBH. It seems to me that, for instance, you could take a 20 MHz crystal, feed it into a voltage controlled (1v = 1x, 5v = 5x, up to some limit) PLL chip, use a voltage divider with a potentiometer, drop a 1/10th frequency output pin on the PLL for low-end oscilloscopes and be done with the synthesis side. Instead we have 28 pin DIPs and datasheets whose target audience might as well be 15th century Mongolians.

Edit: it occurred to me as I was watching Hell's Kitchen that if you have a VCO locked at 1296 MHz, why do you need a PLL? I thought PLLs were mainly used for frequency multiplication.

HAHA
I think you may be trying to jump in far too deep here. You need, either to study the very simplest radio circuits (stone age 'Mongolian' electronics) or to treat the ICs as just black boxes and believe that they function as the data sheet tells you. (Believe it or not, those data sheets are very useful for the "target audience" - of which you are not one! )
The reason that such circuits are used is that they actually do much more than simple circuits, they are much cheaper to incorporate into equipment - reducing or eliminating 'wound components' which are very expensive parts of the construction process in mass production.
If you really want fast results from this, you are doomed to disappointment. It's a vast subject and mustn't be underestimated. Basics first, I'm afraid.