On 8/16/21 12:09 PM, Syd via groups.io wrote:
In the manual there is a Transform setup for the bandpass mode and here is the picture of the results attached. I think this setup is strictly for measuring band pass filters, correct me if I'm wrong. From the pix I can see that the time being scanned from the frequency setup is from 0 ns to 43 ns shown on the bottom. From what I am seeing I think ch0 shows the pulse being generated and ch1 is showing the results from the BPF.
not exactly, what you're seeing is the pulse that would be *reflected* back from the UUT for CH0, and the pulse that would be transmitted for CH1.
The 20.0n/-0.14 and 10.0n/-0.02 scaling for ch0 and ch1 at the top I thought was showing the scaling per division and the level, but now I am not sure what it is telling me.
Perhaps the reflection coefficient magnitude?
Then again the white marker notation is showing 4ns 0.4m? How do I interpret that?
The basic measurement is in time, so some assumed velocity factor is being used to turn that into distance.? The reflected pulse is 4ns, so the 1 way time is 2ns, and that probably works out to 0.4m.? (in free space, 2ns would be ~60cm, so with a VF of 66%, that would be 40cm, I think.
Additionally, what is the TDR measurement of a band pass filter giving me information wise that I can't get from scanning the BPF and looking at the logmax output over a span of frequencies?
Not much.? It's not a commonly used way to look at filters.
At least from the logmax method I can see the characteristics of the filter, and I don't see anything like it with this method since it is in the time domain.
Time domain is interesting if you're wondering about group delay and dispersion.? If you have a SAW filter, sometimes you can see "triple transit" effects as a peak at 3x the nominal delay.? It's also interesting if you're concerned about envelope distortion.
syd/wt1v
Roger Need