> You can see every connector's impedance bump. Not so with N
You can with some "Brand X" Asian imports...
I recently made a set of precision OSL loads for use with a vector network
analyzer. The 50-ohm load was made using a 0.1% SMD resistor precisely
fitted between the BNC's center pin and shell.
When checking return loss across a sample of five Asian BNCs, the
repeatability was horrible. For example, I could calibrate using one load
and attain 60-90dB as I swept from 500 kHz to 60 MHz. If I switched to an
identical load, return loss degraded to 20 dB and in one case, return loss
would change substantially with the slightest pressure applied to the
connector. These changes are indicative of QC problems. The connectors
look well made to the eye but are in fact an electrical performance
However, when switching to Amphenol brand BNCs for the test load samples,
repeatability was excellent with no more than a couple dB variation. When
you're up in the 60+ dB return loss area, nearly any mechanical change will
be magnified under this test so it's important to maintain top QC.
This one test alone made me a believer in paying a bit more for
quality-controlled connectors, regardless of whether they're
constant-impedance types or not.
TowerTalk mailing list