Gary,
This is wonderful information! So the bottom line is for duplex operation, a
pass/notch is needed to protect the rx and tx from each other, but for
everything else a bandpass-only filter is the way to go.
I like your idea of using the signal generator and receiver to make
measurements. Also think you can generate the offending frequency and use the
FM capture effect principal to know when you've reached the same signal level
(to get at least a ballpark number). I often do baseline measurements (sig gen
to rx), then connect up duplexers and record those final numbers. Converting
the microvolts to dbm and subtracting the two give me an idea of how well my
duplexers are preforming. Of course I had a local shop double check them, and I
was pretty close.
Some day I'll get better measurement equipment. For now simple tools and
process theory are my best learning methods. I can make a crude graph to see
how the overall system is working by using the sig gen and a tunable receiver
(i.e. 10 or 50 Mhz steps from 50 to 500Mhz).
Tony
--- On Mon, 11/23/09, Gary Schafer <gascha...@comcast.net> wrote:
From: Gary Schafer <gascha...@comcast.net>
Subject: RE: [Repeater-Builder] Has anyone measured out-of-band rejection for a
duplexer?
To: Repeater-Builder@yahoogroups.com
Date: Monday, November 23, 2009, 4:03 PM
You can make the measurement with a signal generator and a tunable
receiver
that has some kind of indicator for signal strength. It doesn't even need to
be calibrated. Connect the signal generator to the antenna port and the
receiver to the receiver port of the duplexer. Be sure to disable the
transmitter.
Find a reference level with the signal generator on the operate frequency.
Then tune the signal generator to the interested rejection frequency and
find it with the receiver. Then note the signal generator level and increase
its output to match the receive strength that you noted at the start. The
difference between the two levels is the amount of rejection the filters are
giving you at the frequency of interest.
With a pass/reject duplexer you won't have a lot of off frequency rejection
as there is not much of a pass band on that type of duplexer. There will be
good pass band rejection in the space between tx and rx frequencies due to
the overlap of the filter skirts but outside of either it is not much.
For an added receiver filter, your pass/notch filters again will not do too
much for you as far as pass band rejection goes. If you use them to reject a
specific frequency, each can should give you about 30 db of notch rejection
but you may have some degradation of the wanted frequency if it is far
removed. And you will probably not be able to move the notch far enough such
as the broadcast band.
You may be able to convert the cans to pass cavities by changing the
coupling loops. Then you can do the same measurement as described above to
see how much rejection you will get.
Also look at some pass band curves in the catalogs and you will see about
how much rejection a pass cavity will give you at a given distance away from
where it is tuned.
73
Gary K4FMX
> -----Original Message-----
> From: Repeater-Builder@ yahoogroups. com [mailto:Repeater-
> buil...@yahoogroups .com] On Behalf Of Tony KT9AC
> Sent: Monday, November 23, 2009 1:36 PM
> To: Repeater-Builder@ yahoogroups. com
> Subject: [Repeater-Builder] Has anyone measured out-of-band rejection for
> a duplexer?
>
> Hi Everyone,
> Without the benefit of a spectrum analyzer, I would like to find out
> how much rejection of out-of-band signals can be expected from a typical
> UHF duplexer. Have a MSF5000 on 452 that works fine with the T4084
> duplexers (1500 style), but have a lot of VHF data and FM broadcast hash
> that is trying to make it in (the 45kW FM is about 400 yards away and
> the VHF data is almost 1/3 harmonic).
> Looking at the documentation, I can guess its about 20db per cavity (or
> can), but the graphs don't extend very far. So for a regular four-can
> duplexer I might be providing 40db of protection. I want to increase
> this, and plan on adding one or two more cans on the receive side, and a
> Sinclair preselector in-between the latter two to make up for the
> increased insertion loss.
>
> Just wondering if anyone ever tried/measured this, or had ideas about
> filtering out FM broadcast. Eric mentioned using a single 7" Sinclair
> cavity, but I'd like to see if I can use some spare 1500 bandpass/reject
> cavities first. I don't think a 1/4-wave stub will work with that much
> field strength prying open the MSF's front-end.
>
> Thanks,
> Tony
>
>
> ------------ --------- --------- ------
>
>
>
> Yahoo! Groups Links
>
>
>
