John,

"John Smith" wrote in message
...
You missed the point... what I propose is simpler... it is a radio--not a
computer... and still can be done by those who ignore the nay sayers, no
matter how loudly these nay sayers attempt to shout down progress...

Have you ever heard the saying to the effect that the beginner sees only one
option, which the experienced designer sees many? It really is something of a
curse. :-)

By all means do keep pursuing your interests, but by the same token you might
want to start learning more about RF design and understand where some of the
other posters are coming from. Addressing your original idea, there ARE
"modular" radios out there -- I know I've seen some guy's web site where he
takes this approach -- but the idea that a modular radio can somehow offer the
same performance as a more integrated one is about the same as claiming that
you can build a CPU with the same performance and price of a 3GHz Pentium by
using discrete modules for the ALU, memory controller, cache controller,
instruction decoder, etc. -- it just isn't going to happen. On the other
hand, you certainly COULD build some "many MHz" sort of microcontroller with
this approach, and the sames of microcontrollers today swamps that of Pentiums
anyway. Hence, I think there would be a market for your modular radio
design -- especially within the amateur radio community -- but I doubt you'll
be getting calls from Nokia any time soon.

Software defined radios accomplish a significant amount of the
"reconfigurability" that I think you're looking for, and with ever-increasing
ADC/DAC speeds and DSP horsepower, it probably won't be too long before most
radios digitize directly at RF or IF and the rest is software (oftentimes
highly non-trivial software, however). Even so, you'll always need someone
who understands traditional RF engineering to get the signal from the antenna
to the DAC while preserving the best SNR possible.

---Joel