All,
I see some great suggestions on how to deal with 902/903 problems. Here in
the NE we are inundated with cellsites that can kill us on 903, not to
mention all the other shared services in that band. FYI, as a retired
Cingular/ATT engineer they do not have the best TX filtering and if you live
near one, you can have extreme overload. I happen to have one 150' from my
ham tower with S50+ signals which tends to destroy my front end on 903. AT&T
is licensed to transmit on 880-894mhz and if you live within a mile, you
will hear them on 902-3. Nextel is somewhere on 901 and they too will kill
us. They, Nextel, has forced AT&T to place notch filters in their TX to help
with overload. I had them do so at my cellsite and it DID NOT HELP.
I found some TX/RX single cans on Ebay for $20 and bought 4 of them. I
carefully did a slope tune of them so 903 was on the lagging edge (not
center tuned) creating a sharp highpass filter. This worked but I had to
except 4db insertion loss on Rcv. I tried a single large can with the same
loss but was worse on rejecting the 880-894mhz cellular RFI.
What a shame all this great equipment is going to the junk yard. If you know
an AT&T tech, buy him/her a beer and let them lead you to the sites where
they leave this stuff outside. Focus on the AIF (antenna interface frame)
for filters, preamps and lots of SMA jumpers.
Steve
K1IIG
> Hi Folks: From time to time I've seen some 5 pole 900MHz interdigital
> filters show up on ebay. They are about 3.5 by 2.5 x 3/4 inch in size and
> are marked 890-905MHz. The bandpass is very flat and the loss is .8~1 db
> as
> close as I can measure with my surplus test gear. One ebay seller
> mentioned
> something about their use in railroad communication systems.
> The response can be narrowed for a 902~903 peak response, it depends how
> much patience you have for tweaking with a spectrun analyzer and sweep
> gen.
> I would not try peaking one of those without instruments tho. The skirts
> are
> pretty sharp below 890 and above 905, so they are usable without tweaking.
>
> Another possibility is looking at the old F-204/U twin cavity filters that
> were originally designed for 374~404 MHz. I cut these down for use on
> 1296MHz ala the WA5VJB pipe cap filter principle. On that band the cavity
> cylinders are cut down to 2 inches tall and the inner guts pulled except
> for
> the rotating tuning cap and I/Os replaced with #12 wire probes about 1.25
> to
> 1.375 inch long for low loss (<.5~1db)/broader bandpass, or 1.0 inch range
> for higher loss (3db) tight bandpass. I use one of these modified filters
> in
> series with a Parabolic interdigital filter on 1296 to further tighten the
> BPF response and have some adjustable tuning control for input matching.
> The
> combined loss on both filters is .8db (12W in 10W out) on 1296. No reason
> the same method could not be used on 900MHz. I'm guestimating I would have
> to go with cavity cylinder heights about 3 inches tall and I/O probes
> around
> 1.8 inches long for a 900MHz version. If anyone is interested I can spend
> an hour or two some evening putting one together for 902/3MHz and report
> back on the results. It's been something I wanted to tinker with anyway.
> 73
>
> Mike wa3tts
>
> One side note: It's possible to use other F-series twin cavities for 23 or
> 33 cm BPFs but the ones with the smaller rotating tuner caps yield lower
> throughput losses. Replacing a larger disc with a penny seems worthwhile,
> or
> perhaps removing it altogether (shaft tuning only) and shortening the
> cavity height to compensate.
>
>
>
> _______________________________________________
> VHFcontesting mailing list
> VHFcontesting@contesting.com
> http://lists.contesting.com/mailman/listinfo/vhfcontesting
_______________________________________________
VHFcontesting mailing list
VHFcontesting@contesting.com
http://lists.contesting.com/mailman/listinfo/vhfcontesting
|