Limit to square wave output frequency?

A square wave is just like a normal sinusoidal wave except that it has plenty of odd harmonics (each at the appropriate amplitude).

For example, if you're looking to create a square wave of 20MHz, then your circuits need to be able to handle at least 60MHz to produce a waveform that even looks slightly like a square wave. A 'nice' 20 MHz square wave would require harmonics well above 100 MHz.

The issue for you isn't really creating a nice 20 MHz square wave (that's easy) - the issue is creating a nice 20 MHz square wave that is capable of driving your cyclotron (I assume that is what you're up to...).

Which leads to the next question:

Are you sure you want a square wave?

Circular motion and sin waves are one in the same. If you're using the 10-20 MHz wave to drive the cyclotron, then you likely just need a nice easy sin wave (from any HF transmitter).

If you're going to be using much power, then choose an ISM frequency like 13.56 MHz (check the regs).

And unless you're licensed, randomly spitting out RF is illegal

TTFN

Well I have been looking around for square wave generators and it seems there are some that will do what I want and are tunable. Is it possible to just amplify that signal without losing the form of the wave?
Also to IRstuff I'm planning on having a faraday cage around my project, so hopefully the Feds won't show up at my garage.
thanks for the help
-Todd

A faraday cage isn't a perfect device. If you will be generating squarewaves for your experiment remember that the odd harmonics of the fundimental may have substantial energy. One thing to keep in m ind is that Ham operators talk around the world on a few hundred miliwatts and less.

I remain,

The Old Soldering Gunslinger

What does a cyclotron load look like? Capacitive? What are the voltage and curent requirements? As mentioned before, ham gear may be just the ticket, modified for your square wave. Are you sure you want a square wave? If so, what rise time is needed?

If your cyclotron needs a bit of power at 20MHz, as I suspect it will, and you are driving it with anything resembling a square wave then you have a serious shielding issue on your hands. A 20MHz square wave from a fast source will have significant amplitude harmonics well into the hundreds of MHz.

For a faraday cage to be effective it needs to be electrically continuous with all joints soldered/ welded/ sealed with finger-stock or RF gaskets, everywhere and have a filtered power entry. A metal box grounded is not a faraday cage and will pass lots of radiation at the higher frequencies. Eg, a riveted metal box with rivets every 4" is ineffective as a shield above say 500MHz due to the tiny gaps formed by sheet buckling between the rivets.