TDR, VNA and FFT...

[Fred Bartoli]

Seems simple at first sight.

Apply a sine wave to the DUT through R=Zo
Multiply voltage by 1/(j w)
Take inverse FFT.

The outcome is... uhhh

I must have overlooked something, but what?

....going to bed. Maybe I'll do better tomorrow.

 

[Bob]

"Fred Bartoli"

Seems simple at first sight.

Apply a sine wave to the DUT through R=Zo
Multiply voltage by 1/(j w)
Take inverse FFT.

The outcome is... uhhh

I must have overlooked something, but what?

...going to bed. Maybe I'll do better tomorrow.

What are you trying to do?

There is something called frequency domain reflectometry but, with that, you
need to sweep the DUT through a range of frequencies.

Bob

 

[Joerg]

What are you trying to do?

There is something called frequency domain reflectometry but, with that, you
need to sweep the DUT through a range of frequencies.

Yeah, but what's wrong with good old TDR? It's simple and it works.

 

[Fred Bartoli]

Joerg a écrit :

Yeah, but what's wrong with good old TDR? It's simple and it works.

Yeah, simple, but try to obtain precision on a 5km line while still
having the communication link working with big data signal on the TDR
side, and also decent aquisition time.

Other "neat" equipment monitoring features provided almost for free by
the VNA.

 

[Bob]

Yeah, but what's wrong with good old TDR? It's simple and it works.

TDR has its limitations when the transmission line gets long (lossy) and in
the presence of noise. The advantage of frequency domain reflectometry (FDR)
is that, when applying a given frequency, you can apply a sharp bandpass
filter and thus reject noise. It is then relatively easier to "pick out" the
effect of any reflections.

Yes, FDR is more complex than TDR so it will depend on your application
whether one technique is better than the other.

Bob

 

[Joerg]

Joerg a écrit :


Yeah, simple, but try to obtain precision on a 5km line while still
having the communication link working with big data signal on the TDR
side, and also decent aquisition time.

Yeah, but you didn't say that in you post

Other "neat" equipment monitoring features provided almost for free by
the VNA.

If the channel has to keep running TDR might indeed not be such a cool
option. Except if you could use syncs or certain other features in the
signal as a "TDR pulse".