Hi Rick,
Agreed, the numbers surely seemed too good to be true. I also researched
cable delays versus temperature at various manufacturers. I found no data at
the low bands we use. I found just enough information to make me do the test. I
am confident the answers I recorded are correct. I did not look at higher
frequencies with the VNA. Teflon cable is the only one I found that had a
nonlinear curve so rejected that idea as I don’t use Teflon for this.
There are way too many things going on with the cable construction to
estimate in my mind. That’s why I did the test. I also reaffirmed that after
the cold test the cable loss returned to the starting point.
Perhaps someone else would care to look at the cable.
Even if the copper center conductor expands, the way the cable connectors
are made allows the copper to simply push in the connector past the connection
point maintaining its physical length.
Again way to many things and so I did the tests. I think I will put the
cable in the freezer again and do the tests again which seems like the right
thing to do. Am confident the VNA is accurate. I have some short pieces of
cable and I will add some cable to the DUT to verify the VNA sees the
difference.
I will report on this again soon.
Lee K7TJR.
These numbers seem too good to be true.
The tempco works out to less than 7 PPM/°C.
Consider that the coefficient of thermal expansion of copper is +17PPM/°C.
Refer to Figure 7 of this:
https://www.timesmicrowave.com/DataSheets/Literature/Current%20innovations%20in%20phase%20stable%20coaxial%20cable.pdf
The graph on the right shows non-linear tempco curves.
When you have non linearity like this,
it is possible to pick two points on the curve such that a line drawn through
them has a slope of zero.
Maybe you ran into that.
The numbers you reported for the loss seem roughly consistent with what the
thermal coefficient of resistance of copper predicts.
Rick N6RK
_________________
Searchable Archives: http://www.contesting.com/_topband - Topband Reflector
|