Richard Fry wrote:
As you well know, the system change you describe will not remove all
reflected power seen at the output of a tx expecting a 50 ohm load. You
have simply moved the source of the reflection back to the plane of the
tx output connector.

Sometimes it helps to understand the situation by mentally
adding one wavelength of lossless feedline to the system.

Roy's suggestion to change to a 300 ohm load only apparently
eliminates the reflections.

XMTR---1/2WL 300 ohm feedline---300 ohm load

Mentally add the one wavelength of lossless 50 ohm line.
Conditions remain the same but the reflections are exposed.

XMTR--1WL 50 ohm feedline--+--1/2WL 300 ohm feedline--300 ohm load

There's a 6:1 SWR on the 1WL 50 ohm feedline and we have reflected
energy incident upon the source, just as you described above.
--
73, Cecil http://www.w5dxp.com