wilkinson power combiner
 

question now bout these things!

if i built a 5 port device, what would be the insertion loss (so i can
calculate the power rating of the resistors)? would it be 3dB or 7dB (the
signals aren't in phase or anything, in fact they're different frequencies).

thanks again!


dr. x


---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.799 / Virus Database: 543 - Release Date: 22/11/2004

To my knowledge, you can't use a Wilkinson divider or combiner for
signals at different frequencies or phases. It'll do something, for
sure, but it certainly won't function as a Wilkinson divider is intended
to. You won't have port-to-port isolation, and the loss will be much
greater than you'd get with simple dividing. You'll likely end up with
more power going into the resistors than out the ports, so I'd size the
resistors to handle the sum of all signals going in, unless you can
model it with SPICE or a similar program.

Roy Lewallen, W7EL

James Bond wrote:

question now bout these things!

if i built a 5 port device, what would be the insertion loss (so i can
calculate the power rating of the resistors)? would it be 3dB or 7dB (the
signals aren't in phase or anything, in fact they're different frequencies).

thanks again!


dr. x

"James Bond" wrote in message ...

if i built a 5 port device, what would be the insertion loss (so i can
calculate the power rating of the resistors)? would it be 3dB or 7dB (the
signals aren't in phase or anything, in fact they're different frequencies).

My guess would be 7 dB. But I would do like Roy says and model it
with Spice or whatever. Multi-way Wilknsons are just too bizarre for
my brain, though I have seen the occasional paper on the subject !

I did model a 2-way Wilkinson combiner (the microwave kind with real
transmission lines) with a single input signal and much to my (and
probably Roy's) surprise it was perfectly matched and had a 3 dB loss
from one of the two inputs to the common port.

73,
Steve VE3SMA

"James Bond" wrote in message ...
!

if i built a 5 port device, what would be the insertion loss (so i can
calculate the power rating of the resistors)? would it be 3dB or 7dB (the
signals aren't in phase or anything, in fact they're different frequencies).

My guess would be 7 dB. But multi-way Wilkinsons are too bizarre for
my brain ! I would do what Roy says and model it with SPICE or
whatever.

I did model a normal 2-way Wilkinson combiner with only a single input
(the microwave kind with actual transmission lines) and to my surprise
(and probably Roy's) it was perfectly matched at one frequency, had
perfect isolation at one frequency and gave a 3 dB loss from one of
the two input ports to the common port (all at the design frequency
only, of course).

73,
Steve VE3SMA