[R-390] update: HP8640B RF Sig Gen Leakage

[Randy Zelick]

Hello again,

Thanks to all who made suggestions to remedy the generator leakage 
problem. It looks like the leak is plugged!

Here is the tale....

I made a sniffer out of some double-shielded coax and a miniature antenna 
at the end. Then I tested the generator at 440 MHz where the problem was 
the worst, using a very sensitive 2-way radio as the detector.

There was RF floating around the case at different points, but the worst 
leak by some margin was coming out along the shaft of the freq calibration 
knob. This is the tiny knob just to the right of the frequency display. If 
you turn it, the "uncal" light comes on. Next to that is a little hole for 
frequency calibration.

So I popped the hood to look for problems. This 8640 is quite a nice one, 
just a year out of cal with the stickers still over case screws. Clean as 
can be inside. So rough use and/or corrosion was not a likely issue.

I took the cover off the frequency counter section which is also where the 
calibration control shaft enters. Inside there is a small spring leaf that 
grounds the shaft, which became immediately suspect. It was already clean, 
but I cleaned it more and added a little stabilant. Still RF leaking out.

Then I re-surfaced the edges of the casting which forms the cover of the 
counter section and tightened all the screws down. Still leaking.

Finally, I made my own shaft grouding gizmo from a piece of shim bronze. I 
captured it with the screw that faces front and holds the two halves of 
the counter case at their seam just right of the display. The other end I 
notched so it would rub against the shaft. Unlike the HP grounding shim, 
mine is mounted on the outside of the counter case. No leak!!

I sniffed all around the 8640 and discovered that in addition the top
cover must be very well secured, otherwise more minor RF leaks present in
the interior get out.  These leaks seem to be coming from several sources,
which I did not track down. But with the outer covers in place, I realized
satisfactory behavior.

The bottom line? I am a bit perplexed. Here is a generator of the highest
quality and certainly appears as though attention was paid to every detail
of construction. Solid castings held with many screws, gold plated circuit
boards, best coaxial interconnects, and so forth. My specimen is extremely
clean, unmodified, unabused and yet *barely* can keep RF from leaking out.  
Only with every little detail of grounding perfect does it not radiate. No
overdesign there!

I might speculate that the counter is the problem. The 8640 started life
with a slide-rule frequency display (the A model) and likely was
retrofitted with the digital counter to make the B model. Perhaps it was
just too early in the days of RF counter design to seal all the RF leaks.  
Just for fun I ran the generator with the cast lid completely off the
counter module. GOBS of RF pour out. C'mon HP!

As I final note I want to plug the perverse obsession I have of saving
little mechanical parts from cannibalized equipment. There is an old
military transit case in the corner of my shop where this variety of
schmutz accumulates. I'm sure it will be the first thing my kids pitch in
the dumpster when I go to the big test bench in the sky, but it took but a
few minutes of rooting through the six inches of levers, pawls, shafts,
couplings, washers and (very) etc to come up with the magic grounding
shim.

So there you have it.

Cheers,

=Randy=

-- 
R. Zelick				email: zelickr@pdx.edu
Department of Biology			voice: 503-725-3086
Portland State University		fax:   503-725-3888

mailing:
P.O. Box 751
Portland, OR 97207

shipping:
1719 SW 10th Ave, Room 246
Portland, OR 97201