Can anyone suggest a transistor or circuit arrangement to produce the following:
2.4Ghz square wave
12V
2 - 4 amps continuous
Any ideas or comments would be good. Obviously anyone with reasons for this not being a good idea for whatever reason, please let me know. I have a project idea but would like to know the limitations in this stage of the design before worrying about the rest.
Thanks again
Without doing any research, my gut reaction is that this is extremely unusual and so it will be a very rare engineer who has the knowledge to give you useful circuit or component suggestions. While there is a vast supply of 2.4GHz transistors and modules out there, they are all tailored to the communications market, and for relatively narrowband operation. Your desire is for an extremely wideband switch, something with a bandwidth to above 15 GHz, in order to get even a sloppy rectangular waveform. This is a huge bandwidth and such a bandwidth at the power levels you want may be beyond common technology capabilities. Perhaps the current research into "Ultra Wideband" communications and imaging devices may unearth some devices capable of fast switching of these high power levels.
It would be an awful lot easier if you didn't need a square wave. Can you use a rounded waveform rather than a square one? This would improve feasibility enormously. If your goal is simply to generate power at 2.4GHz, a rounded waveform such as a sine wave would be almost simple by comparison.
When dealing with large power levels at these high frequencies, a designer will need to know what the nature of the load is. So you are not likely to get a comprehensive answer to your question without divulging some electrical properties of the load into which you will be pushing these 2 to 4 amps.
Of course there will be no regulatory issues if the energy is completely contained inside your apparatus. This could meet the existing Part 15 rules for an unintentional radiator. However, if you plan is to radiate into the open air then I fear it may be an impossible goal.
edit: on further reflection, one way to approach this problem might be to use a magnetron to generate a large amount of power in a sinusoidal form and then pass that power through a diode limiter to change the waveform. The initial high power generation might be at, say, 400 watts, and after clipping the tops off the sine wave, you might end up with somewhere around 50 to 100 watts as the clipping and matching process will have very poor efficiency. At least the magnetron is common, they are used in microwave ovens. Not sure if the diode clipper is feasible though. And of course it would be quite challenging to maintain the waveform shape as power is delivered to the load.