<html>
<head>
<meta http-equiv="content-type" content="text/html; charset=ISO-8859-1">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<img src="cid:part1.05010607.02020101@bartk.us" style="" class="uw
xO"><br>
<br>
The device on top is the HP 8672A and it seems to be fully
functional. It can generate signals from 2GHz - 18.5GHz, with power
levels from -120dBm to +10dBm, and with AM or FM applied. It's
fully GPIB controllable. Here you can see it sending a 0dBm signal
@ 18GHz into the top spectrum analyzer. Anyone up for a 10GHz QSO?
;)<br>
<br>
Speaking of the top specan, the face is now reconstructed and
working well! The main dial knob is not working yet, needs a #683
light bulb replaced. A pack of 10 is in the mail. The display
section of the top specan is also buggered up. Any RES BW < 3kHz
has severe attenuation. I'll have to inspect the filters in there
and figure it out.<br>
<br>
The bottom specan's only remaining problem is a dim/fuzzy display.
There's apparently a fix for this. Last night I also thought the
high range (2-22GHz) of the bottom specan was screwed up since it
was showing signal levels 25dB down from what the calibrated power
sensors were reporting. Turns out nothing was broken, just the YIG
tuned mixer was misaligned and needed to be corrected (there's a
digital correction table inside the instrument). When I pulled the
memory board out of the bottom unit to help troubleshoot the top
unit I cleared all that calibration data.<br>
<br>
All of these need to be calibrated/performance tuned still. I have
a replacement atomic frequency reference on the way since the one
that arrived last week is broken. Took 15 minutes of heating time
before it achieved signal lock. And then it only locked for about
60 seconds, never to return. The other parts missing for
calibration are extension boards (custom bus extender PCBs for these
model specans) and an active probe. I think I have everything else
to perform a good cal.<br>
<br>
Ever onward,<br>
<br>
--Bart<br>
<br>
<br>
</body>
</html>