The problem with a straight splice near to an amp output usually comes with having a coax stinger cut slightly too long. When this happens the coax stinger bottoms out in the splice, but the radiation shelid does not bottom out in the cored out coax. That leaves a small amount of cable with no insulation. In an extrememly long stinger, there will be a radiation leak, but this is very rare. This is the culprit when there is a reflection affecting the forward signal.
So conversely if there is a splice in the near proximity of the forward-feed input {which is where the return would launch from} of an amp that is return-capable, the reflection from a mis-cut stinger would affect the return but leave the forward signal alone.
Someone once asked me why a long stinger in a standard connector does not affect the signal. The reason as I see it would be the fact that a splice is not constructed exactly like a pin fitting, it is more similar to a plain old wire butt splice. However when dealing with RF signal, with the proximity of two mis-cut stingers in a straight splice the reflection grows exponentially. So if you're using those wonderful "magic splices" to mate dissimilar or even like cables, you could still run into reflective issues a close proximity to amps.
> Aside from seeing return signatures from splices, where can I link to find there losses? What are there losses?
![(confused)](/smiley/original/confused.gif)