RF Cascade Workbook for Excel
RF & Electronics Symbols for Visio
RF & Electronics Symbols for Office
RF & Electronics Stencils for Visio
RF Workbench
T-Shirts, Mugs, Cups, Ball Caps, Mouse Pads
Espresso Engineering Workbook™
Smith Chart™ for Excel
|
 |
50 ohm filter, DC to light - RF Cafe Forums
|
| rpaulsen
|
|
Post subject: 50 ohm filter, DC to light
Posted: Mon Apr 09, 2012 6:09 pm
|
|
Joined: Mon Apr 09, 2012
6:07 pm
Posts: 1 |
|
I need to design a bandpass filter with 50 ohm input
and output impedance. This would be a simple task
except that I need the input and output impedance
to be 50 ohms in the passband AND in the stopband.
Most of the bandpass filters I've designed
in the past have a very low impedance at frequencies
below the passband and a very high impedance at
frequencies above the passband. Has anyone seen
any design material on matching the stopband of
a filter to 50 ohms? I'm sure this is going to be
very challenging because the impedance swings all
over the Smith Chart. Can anyone help?
Thanks
|
|
| |
|
 |
|
fred47 |
|
Post subject: Re: 50 ohm filter, DC to light
Posted: Fri Apr 13, 2012 5:40 pm
|
|
| |
| General |
 |
Joined: Wed Feb 22, 2006
3:51 pm
Posts: 104 |
|
There's only one way to get what you need: an "absorptive"
filter.
That looks like 2 filters, connected
in parallel at the input, with one going to the
output and the other going to a resistive termination
capable of absorbing all the
applied power in
the stopband. If you need bidirectional termination,
looking like 50 Ohms
at each port, it starts
to look like 3 filters - doable, but rather a mess.
Usually, the textbook examples of these
absorptive filters are for high-pass or low-pass
filters, but the same principles apply.
Good luck!
|
|
| |
|
|
|
|
biff44 |
|
Post subject: Re: 50 ohm filter, DC to light
Posted: Tue Sep 25, 2012 10:50 pm
|
|
Joined: Wed Apr 08, 2009
11:07 am
Posts: 33 |
|
a couple ways to do it:
1) put 3 dB pad in front
of BPF, and you will have at least 6 dB return loss
out of band
2) put broadband circulator in front
of BPF, and you will have at least 15 dB of return
loss over a much bigger bandwidth
3) IF you know
what you are doing, you can design LPF/BPF/HPF (or
some similar variation) triplexer using singly terminated
design. Done properly, you could have DC to 20 Ghz
good return loss. Done poorly and you will have
a tuning nightmare.
|
|
Posted 11/12/2012
|
|
|
|
 |
