Depends. If you are patient, you can get one at a better price simply by
sticking around and finding a week when nobody's bidding on them - at
least that *eventually* worked for me, and my cutoff point was a lot
lower than $300, so I let a whole bunch go for more money than I was
willing to put into them. I specifically did not want new and Chinese,
and I'm quite happy with my 3325A, LEDs and all. I have not found old
age to be an evil thing in well-built equipment, and 20 (up to 70,
actually, with some limitations) MHz is not a problem for my current
anticipated uses. I do wonder when I'm ever going to make use of the
0.000001 Hz end of the range; I am happy to have it, however, having run
into a situation with a client's generator where something below its 5
Hz bottom end would have been useful. It does a far better sweep, too.
Plenty of 3325s selling for well below $300. Don't actually see any
recent ones that got bid on (ie, more than one bid) selling for that
much. Couple with high starting prices that did get bought for $325 and
$399 with a single bid or BIN. One a couple of weeks back did get bid to
$301, but plenty selling a bit above and even a bit below $200 (not
really with shipping) in the past week. Also plenty of volume to
indicate that another will be along if you don't get the one you want at
the price you want to pay, if you can stand to wait.
Point being, by the time you've built the controls, power supplies,
output amps, connectors, boxes, etc. for your $30 chip, you'll probably
have spent more money on it. It will be newer, and perhaps it will go
faster, but the old one is a fully built system with a well-fleshed-out
user interface, etc. You can get new China junk to 20Mhz for $200 or
less, if you're willing to go there. Building your own, unless you're
sitting around looking for ways to fill your time, is something to do
for the experience or the bragging rights, or else to design the thing
well enough that you can sell them (built, as kits, board and a parts
list...) and make money at it - otherwise it's not "worthwhile". It need
only be "worthwhile" in one of those areas to be worth doing - it can be
a non-economical experiential job, for instance. And perhaps you can
even make it economical (ignoring time spent), if you can find a
"known/presumed broken" hulk on ePrey and pick it up cheap (inclusive of
shipping, border crud, etc) as a source of parts to build a user
interface, box, power supply etc from.
Win's partner-in-book's lab has this up, if you're interested, though I
think it's a bit out of date at this point in time:
http://seti.harvard.edu/synth/index.html
