... Just a range of voltages with trimpots for setting the endpoints should be good enough, especially if I can rig up a marker or two using crystals...
This likely would work, depending on your application. How critical is the frequency accuracy?
If the actual frequencies aren't that important, you could just use sequentially-switched crystals,
such as those from a scanner. I often find boxes of odd frequencies at hamfests. The same
switching signals could be used to choose separate trimmer capacitors in the input tuned circuit.
Or just get a cheap scanner receiver, which is designed to do exactly this.
We still have a lot of users in the 30 - 50 MHz range in this area, and any TV stations still on
the low VHF channels may swamp your preamp. You certainly want to filter out the FM broadcast
band, as that is one of the most saturated portions of the RF spectrum.
So on front ends, is a 40673 about as good as a BF988?
The 40673 and other metal can dual-gate FETs are no longer made, though they
are still around. There were many other similar parts (40822, 3N200, 3N206, etc.)
and, looking though the RCA data book, there is little practical difference among them.
The BF998 is current production, and would be the best choice for future parts availability
if you can work with SMD devices. The operating voltages are different, but at that
frequency range I don't think you'll notice any performance difference, and the same
circuits should work with slight changes to the DC voltage and biasing.
Any insights on suitable broadband antennas? I tried a fan dipole with 17 dipoles (http://ab1jx.webs.com/toys/jove/antennas/fan1/index.html
) and now a T2FD. I might try a really stubby LPDA sitting on the ground pointed straight up. I don't really know much about radio astronomy, I'm just starting and trying to get antennas up before snow.
How much gain do you need? What is your noise floor? How much do you need to
reduce terrestrial pickup? How narrow of a pattern do you want? Point straight up?
The T2FD will have some loss, perhaps 5dB or so when it is more than 1/3 wavelength long,
and more at lower frequencies or shorter antennas. It isn't a bad idea if you want good SWR,
but not as good for high sensitivity or efficiency. But you can get better performance and
reasonably wide bandwidth with a true
No, not the kind with separate wires for each frequency, but the old-fashioned type that
is a derivative of a biconical dipole. In this case, take 10 wires each 11' long. Connect
5 of them to each side of a 4 : 1 balun that covers the frequency range of interest. Space
the ends on each evenly so the furthest ends are about 11 feet apart. Mount the whole
thing about 5' off the ground. (A good reflector plane underneath will help - perhaps a layer
of chicken wire.) That should give you an SWR better than 3 : 1 from 30 to 60 MHz (depending
to some extent on the ground underneath.)
If you want to try a Log Periodic array, I'd suggest looking at the Zig-Zag Log Periodic:http://www.cebik.com/content/lpda/zzlp3.pdf
Once you do the initial math, these are fairly easy to build by winding wire back and forth
between sets of nails on a wood frame. You can get good performance as long as you
use enough elements: in the article W4RNL used 20 elements to cover 50 - 200 MHz.
I built one with 12 elements to cover 200 - 500 MHz, but the rear pattern wasn't well
controlled. That might not be as much of a problem if you are aiming it straight up.