In article et, rae4
@ix.netcom.com says...
Is anyone building or thinking about building the above very interesting
project?

http://users.adelphia.net/~n2pk/index.html


If so, I have two printed circuit boards going spare at cost (about $22)....

Dan
ac6ao g3ncr


I may build one to play with it. Paul's documentation is awesome -- I
can well imagine how much work he put into it.

At a glance, though, there are a couple of things I would have (and may
still) do differently if/when I get around to building a VNA:

1) I don't know that I'd use DDS synthesizers for an application like
this. There's no need for serious close-in spectral purity, but you'd
certainly like to have a higher frequency-coverate limit. Also, to
achieve truly-meaningful measurements with an untuned receiver section,
suppression of harmonic content in the test signal might be more
critical than the existing design allows. (Imagine trying to
characterize a high-pass filter with an untuned receiver and DDS
harmonics coming through at -30 dBc.) The latter may not be a concern
if the AD9851's harmonics (and those generated in any subsequent output
buffer stage) are very low in amplitude.

2) Hand-building couplers and bridges doesn't make much sense when
surplus HP8505-era test sets good to 1 GHz+ are as cheap as they are.
If I were only concerned with HF, though, Paul's approach looks fine.

3) There's no excuse for not releasing the source code to the software,
IMHO. A proprietary closed-source DOS app doesn't exactly lend itself
to customization and improvement. In fact, we'll be lucky if this code
even runs on the next generation of Windows. I'd urge Paul to
reconsider his software strategy (assuming I'm not mistaken about it
being closed-source in the first place).

Pretty darned cool project, all in all.

-- jm

------------------------------------------------------
http://www.qsl.net/ke5fx
Note: My E-mail address has been altered to avoid spam
------------------------------------------------------