How to create multiple simultaneous signal sources from a single reference source?

I wanted to build a circuit which would produce a set of signals of different frequencies in different channels simultaneously using a single reference source. I also want the frequencies to be changeable. I am looking to generate frequencies in kHz and MHz range. On searching on the web I came across PLL. I wanted to know if I can do it using PLL or any other methods. It should be fully analog.

PLLs would certainly do the job. They rely on digital systems internally, but usually use a Voltage controlled oscillator to generate the output, I assume you need a sine wave. This is their biggest problem, depending on the band width needed, it is difficult to get VCOs to operate over a wide frequency range with good performance.The other possibility is  direct digital synthesis(DDS), this is an all digital method, the sine waves are commonly  synthesized using a lookup table and digital to analogue converter. They are used in many modern signal generators up to 100s of MHz. Analog devices have some four channel parts  (AD9106 for example), they are all small smd parts, but they have a demo board. A few more details would help, cheers,Richard

But if I wanted to generate different frequencies f1, f2, f3,..., fn simultaneously then should I use n different PLL circuits? Even if I am using it will there be any common elements that I could avoid repeating in all the circuits?

You don't say if you need sine waves, or square waves, although I assume sine since you talk about an analogue solution. PLLs are in fact largely digital. What frequencies are f2, f3 etc, are they integer multiples of f1, for example. Different phase or frequency detectors in PLLs will allow some interesting possibilities, but you have not supplied sufficient information.Normally you would need different VCOs, and therefore individual PLLs unless the frequencies are related in integer multiples of the fundamental for example.The only common element would be the microprocessor (or PC) needed to program the PLLs and the  clock generator.Getting VCOs to work is not trivial, and some sort of leveling circuit (AGC) will be needed if you need constant amplitude.There are many advantages in the DDS method if you can see your way clear to using them, what sort of spectral purity are you needing?
Have a look at http://geoffg.net/SignalGenerator.htmlcheers,Richard

Although I prefer Richard's selection of DDS, and even though you ask for a completely analog method, perhaps you cold consider an hybrid approach. You could take an oscillator (even a crystal if you want stability) at or higher than the highest frequency you need. Then use different dividers to generate the frequency that you want. Of course thise dividers could be programmable by DIP switches, thumbwheels and even a host micro. If you have obscurish values you may need to find a crystal with the highest common factor, depending on the accuracy that you need. Then take the digital wave and convert it to a sine wave using op-amps although the op-amps in the MHz may prove a little pricey.

FWIW we used to use a precursor in a suitable technology for the higher frequency, and then work with the CD4059 that can divide from 3 to 15999 (pin programmable). I see there is a 74HC4059 that will handle up to 18MHz input worst case at 5V.
There are also rate multiplers that will output N*Fin/16 (16 for binary ones), also pin programmable like the CD4089.

I had same question some months ago. I agree with Aubrey, at least it worked with me.---https://www.adslyoigo.net/

John for some days now I have been trying (without success) to post a solution that may be useful for you.  It consists of a shift register that alternately fills with ones and zeroes.  If you take equal value resistors from the outputs to a common point then the common point will output a triangular waveform.  If you adjust the resistor values you can get a stepped sinewave.  If you then filter that, even with a simple RC filter, you will get a pretty pure sinewave.
This circuit (which is all I was able to post) is from Elektor magazine around 40 years ago (reproduced with permission). I have the full article of this, and a 32-step version (the above is a 16-step) and also some good PLL circuits that may give you some inspiration.  If you'd like me to send them to you, please send me your email address:- click on my name at the top of this post, this will take you to my profile- click on Contact at the right underneath the blue profile box- You can then send me a private msg.You said you wanted to go totally analog but unless you have a good reason for this, think about using a microcontroller chip.  This would give you accurate timing and the possibility to generate different frequencies on many outputs(which can be fed into circuits like the above to give sine outputs) with relative ease.Cheers // David

To Ron's very useful circuit, I would also add a resistor (100k is ok) from IC2's pin 3 to either ground or Vdd. Pin 4 is already at Vdd so that would be very convenient.Such that the pin is not left floating in case S1's very first switch is selected.

Any suggestion? liteblue

Would like to know a suggestion
liteblue

Getmyoffer.capitalone.com is specially designed for the customers who are having splendid credit and their age is 18 or above , checkout getmyoffer.capitalone.com reservation number