Just to illustrate that last point:
Consider a perfect 50 Ohm load connected to a source through three 6ft
lengths of 50 Ohm lossless coax with Vf=0.66. The three lengths of coax
are joined with two connectors. Assume the connectors introduce an
impedance discontinuity equivalent to a 1.2" (0.1ft) section of lossless
100 Ohm line with Vf=1.
We find that the connector nearest to the load introduces a
discontinuity which causes the VSWR on the middle 6ft of coax to rise
from 1:1 to 1.03:1. However, the discontinuity introduced by the
connector nearest the source, acting on the transformed line impedance
seen at that point, tends to lower the VSWR again; the VSWR seen by the
source improves from 1.03:1 to 1.004:1. So, in that example, from an
impedance discontinuity perspective two connectors are better than one!
Only if the line section between the two connectors was an exact
electrical half-wavelength (or multiple) would the connector
perturbations be fully cumulative.
As I said, you can't simply sum mismatch losses for each connector like
you can resistive losses.
On 15/01/2012 14:57, Steve Hunt wrote:
> By the way, you can't simply sum mismatch losses for each connector like
> you can resistive losses - the interactions are much too complex for that.
TowerTalk mailing list