Some thoughts on building an impulse radar unit (aka wide band radar)
I've wanted to build an impulse radar unit for a long time.
A decade or so ago it was predicted $5 radar chips would be used in everything from heart-beat monitors to water level sensors for washing machines. Low power units appeared in stud finders shortly after this but I haven't seen it used anywhere else.
I got to play with some ground probing radar circuits at CSIRO.
It was a relatively complex circuit which I didn't really understand.
Today (1'st Nov 2007) I found this paper at http://www.hackaday.com/.
I'm sure there is lots of scope for interfacing this to PCs, PDA or micros for some nerdy fun.
I modeled the sampler shown in the above paper and it didn't seem to work.
It wouldn't sample a DC input. It is possible it will capture edges.
I will try modeling my own design.
I also noticed 9Ghz transistors are fairly cheap - it might be possible to build a transistor based sampler.
Googling "diode sampler" will find articles on single, dual and quad diodes samplers.
My design is dual diode with a single negative going gating pulse.
Most two diode samplers seem to use complementary gating pulses.
I've done some modeling of my own sampler.
Spice simulations using ideal diodes do what they should but the real world is likely to be less co-operative.
I'm thinking a reflectometer might be a good stepping-stone towards wide-band radar circuits.
It should be easier to measure the reflection off the end of, say, 10 metres of cable than an short RF reflection.
It would still be slow enough to be debugged with a relatively slow CRO.
Two diode sampler from 1966 http://www.hpl.hp.com/hpjournal/pdfs/IssuePDFs/1966-10.pdf