On 8/26/2012 6:54 AM, Jim Lux wrote:
> when you do the measurements (using some form of VNA, I presume), do you
> cal with a set of standards with UHF connectors, or something else?
I've done calibrations several ways. Fore these products, I'm only
interested in performance up to 6M. I don't have a decent 50 ohm load
with a UHF connector on it, and I'm most interested in VSWR. Also,
whatever cable I connect to them will have a UHF connector on it, so
including the male UHF in the measurement will be representative of
actual performance. So, to answer the question, I'm calibrating with an
SMA load on N to SMA and a through N adapter to approximate the
measurement plane, then replacing the through adapter with a pair of
MIL-spec N to UHF males.
While I can certainly see the effect of the UHF connectors at 6M, the
differences I'm seeing between one box and another, and between one port
and another on the same box, are FAR greater than the effect of UHF
connectors.
One of the major design errors I'm seeing inside these boxes is a
failure to provide a return path that parallels the switched path. That
is, they depend on the chassis of the box as the return path, either by
design, or by accident as the result of a board layout that breaks the
ground plane of a two-sided board under the high side trace. The result
is more stray inductance added to the signal path (in addition to that
caused by the relay contacts). I've improved the 6M VSWR through the
Top Ten A-B boxes by adding braid between the connectors and the board,
and on the board over the breaks, and by adding a short open stub.
73, Jim K9YC
_______________________________________________
_______________________________________________
TowerTalk mailing list
TowerTalk@contesting.com
http://lists.contesting.com/mailman/listinfo/towertalk
|