Re: Insertion delay needed for BNC-BNC couplers UG491A/U (M-M) and UG914/U (F-F)

John KN5L

Hi Kurt,

Here is another plot which demonstrates frequency dependency.
Calibration with SDR-Kit Female followed by sweep of a SHORT with less

Above 40MHz, trace is no longer flat.

John KN5L

On 5/23/20 8:28 PM, John KN5L wrote:
Hi Kurt,

I do understand. Though I admit, I have not advanced to using arbitrary
mode calibration.

Most all that I do is below 30MHz. Attached is a SHORT measurement to
100MHz. Flat enough for me. Again, notable is Sound Card sample
frequency, 14.2MHz, and last point in the weeds. All below 30MHz looks
good to me.

Linear trace is only suitable for measurements within narrow frequency
range. Once over a decade or so range, Log frequency is a must. It is
the only way to have an acceptable low frequency resolution, with
limited sample size, and maximum realistic dwell time per point. In this
trace, 100 points, 10 seconds per sweep is acceptable.

For example, plot attached, with 100 sample points. If linear, first two
sample points would be 1MHz and 2MHz. Leaving out all of the detail
between 1 and 2 MHz, 2 and 3 MHz etc.

For experimenting with real devices; transformers, transmission line,
inductors, etc, below 30MHz is good because most everything follows
theory equations.

John KN5L

On 5/23/20 7:14 PM, Kurt Poulsen wrote:
You have still not understood my point about frequency dependency for
the female short

Join to automatically receive all group messages.