开云体育

First off, thanks to everyone for all the helpful responses! It made me
realize just how loaded my question actually was. It also help unjam the gears
in my head. That said, there's a lot of stuff here to respond to, so hopefully
I don't forget anything I wanted to discuss.

To clarify, I intend to operate at a fixed frequency, no modulation,
preferably somewhere in the 900MHz ISM band, driven by an amp operating at
less than 5W. Each run of the experiment should only last a few minutes at
most.

Probably due to me not knowing any better, I'm leaning towards making my own
amp along with bypass filtering. The amp section would be based on modules
such as:





@Jim Lux & Andy G4KNO
Your references to the FCC requirements and the given examples are very
helpful. It provides a better perspective of what I'm dealing with here. It
also shows me how much I still have to learn.

Regarding the monolithic filters in that band, I suspect they're a lot less
common than they once were. I see reference to a lot of discontinued items.
That said, some can still be found which is plenty for my needs as long as
they provide sufficient performance. The datasheets for some of these leave a
lot to be desired. I didn't hear of minicircuits until you mentioned it. It
looks like they may have a viable solution.

@Roger Need
I actually had both the rf-tools page and that digikey part (along with some
others) already opened in a browser tab. I made the mistake of making a really
sharp 1st order bandpass without realizing how ridiculously tiny either the
capacitor or inductor was. After widening the band enough to allow for more
realistic components, the roll off was much more shallow, in turn requiring
higher order filters. By that point, I became a bit more concerned regarding
the complexity of the filter.

@OneOfEleven & John Gord
Thanks for the hardware suggestions (I'm actually considering a TinySA to test
filters and amps), but since a critical aspect of my experiment is getting
phase angle information, I either have to use the nanoVNA to generate the RF
signal, or I don't use the nano at all and instead buy an old so-called
"vector voltmeter" along with all the other hardware I'd need. I can't help
but to think it should be possible to sufficiently attenuate frequencies
outside the nano's 5th harmonic much easier and cheaper than to purchase a
bunch of additional equipment.



So one thing that's not clear to me, particularly when it comes to
dielectric-type filters, is whether they could be cascaded to increase Q
without using amp inter-stages.