Alot of TDRs use square waves or needle like pulses for length determination. Why are not other pulse shapes like sine or triangle used?
-
\$\begingroup\$ If you need a good distance resolution, you must generate and detect very short pulses. \$\endgroup\$– jonkCommented Nov 12, 2020 at 7:33
-
\$\begingroup\$ very short pulses can also be sine \$\endgroup\$– electrococukCommented Nov 12, 2020 at 7:37
-
\$\begingroup\$ If it's a pulse, it ain't a sine. \$\endgroup\$– JRECommented Nov 12, 2020 at 7:40
-
\$\begingroup\$ If it's a pulse it ain't a sine but it can be a "raised cosine" or a Gaussian. \$\endgroup\$– user16324Commented Nov 12, 2020 at 14:49
1 Answer
In TDR, the two main parameters are pulse length (or rather shortness), to get bandwidth, and pulse power, to get signal to noise ratio. Once you've struggled to produce those, the pulse shape is whatever the mechanism you've used happens to give you.
It's rather like the dog that says 'sausages'. The wonder is that the dog can talk at all, not that its pronunciation is poor.
TDRs tend to use a step recovery diode into a pulse shaping network, for instance a shorted transmission line. All pulses tend to come out looking a bit Gaussian (if you're lucky). Square pulses tend to have their edges rounded off, delta pulses tend to be spread out a bit. As long as the pulse doesn't have a long tail, or ringing, or a double peak, then most users are generally happy with what they get.