fast pulse radio transmission

I did a somewhat similar thing a while ago, but at much slower speeds. Essentially, I used two slightly different frequencies at the transmitting end, one for a "high" signal and one for a "low" signal. On the rx end, I fed this into a PLL (4046 in this case - like I said, slower speeds :) ). The loop back of the PLL was then AC-coupled to two half-wave rectifiers (one "positive", one "negative") feeding into Schmitt triggers, which then drove a set-reset flipflop.

The idea was that on a transition from one frequency to another, the feedback loop of the PLL jumps in either the positive or negative direction as the PLL tracks to the new frequency. The AC coupling was needed in my case because the oscillators were shoddy and drifted all over the place making a pure level detection scheme unworkable. Making this work requires a lot of careful hand tweaking of the components, and quite possibly won't scale up the the switching speeds you need. There's probably some off-the-shelf parts that would do the job much more reliably and with less hair loss :)

--
Michael Brown
Add michael@ to emboss.co.nz ---+--- My inbox is always open

Thanks michael for your answer. I would like to know more about your project. I have a few questions like: What frequencie range, did you use? How did you generated those frequencies? So any further information is very much welcome. Best regards. Sergio

You say "without delay". 1Km at the speed of light takes

3.3microseconds. There will be a finite delay in the receivers, filters don't respond instantly. Without knowing what on earth you are doing we can't say if that is signficant. From your other post about very accurate monostables I suspect that jitter may also be an issue.

You appear to be posting from a university in Tenerife (a spanish island) so I'm guessing you are either a student or an acadaemic. Perhaps if you tell us what you are trying to acheive someone will be able to suggest a better approach to the whole project.

Bob

I second what Bob wrote...let us know what you are actually trying to accomplish and we can perhaps suggest other ways, better ways, of doing it.

I would add some more questions: at the receiving end, how accurately do you need to resolve the pulse width? Can the pulse be repeated, or do you have only one chance to properly detect it and decode it?

Many (most??) RF transmitter/receiver ICs these days target specific modulation schemes. It seems like what you want is just wide bandwidth modulation that is capable of just two levels. You could use either amplitude or frequency modulation: on-off keying, or frequency-shift keying. The data rate and pulse width dictate the required modulation bandwidth, which in your case is not extreme.