Richard Clark wrote:
On Tue, 11 Mar 2008 16:00:11 +0000, Ian Jackson
wrote:


Even at 1GHz, I don't believe that you would get as much as 2dB of loss
even if you joined two pieces of coax by stripping the ends with a knife,
twisting the conductors together, and wrapping them in scotch tape.
--
Ian


It is surprisingly easy to get significant loss at 1GHz and above, 2dB would
be very easy to achieve even when using what appears to be a 'good'
connection.

regards
Jeff


Hi All,

Anyone can do it wrong (come up with 2dB loss), but we would have a
great more deal traffic here if doing it wrong was that common.

I have measured attenuation of near everything from DC to 12GHz at the
standards bench and RG58 with BNCs for a short run (2 to 3 meters)
never presented any problems being wildly imagined here. At the
worst, the connectors (note plural) "might" show 0.2dB mismatch loss -
trivial.

On the other hand, if Jeff is so sure of this 2dB figure, it should
reveal itself in blisters to the fingers for even mild power (100W).
Since this unequivocal evidence is so obviously missing from the
records (or the testimony, as the case may be); then it is very
apparently one of those "someone heard from someone else about their
brother's friend's boss who had a customer who made this claim."

73's
Richard Clark, KB7QHC

I've encountered (defective or broken) SMA hardware that has remarkably
high *apparent* loss. Typically, internal cracks or voids. I had a
batch of Pasternack right angle M-F elbows that had losses that were all
over the place at 1 GHz sorts of frequencies. I'm pretty sure that
internally, there was something not connected, or a spring loaded
something that wasn't making contact.

In this case, the connector wasn't actually lossy, but more of an
extreme mismatch, and when hooked up with a 50 ohm signal generator on
one side and a 50 ohm power meter on the other, it looked like loss.
Actually, it was reflecting the power back to the source.

In this scenario, with 100W, your fingers wouldn't have gotten burned
(but the signal generator would probably shut down with reverse power
warnings)