The 8642A or B are great generators until they break. I have one, and
it was even slightly cleaner in the close-in phase noise department than
my 8662As until it started having intermittent noise bursts. The
modules were intended to be exchange only with HP, and I doubt there is
even service data on taking them apart. They are individual cast
aluminum boxes that interconnect. It was not a great economic success
for HP for some reason, but it has very low phase noise in the HF range,
like the 8662A or 8663A. I don't like the LCD display compared to the
LEDs on the 8662A, but it is a great box if you can keep it working.
All the calibration is done digitally, with a module that is stored in
the rear in a little compartment.
Icom users a pair of 8642s to test the IC-7800, plus a set of isolation
amps, pads and low pass filters, plus the hybrid combiner. Isolation is
the issue here.
I have never used a 3000/3001, but I don't think the phase noise is
adequate for dynamic range testing. Few synthesizers are adequate for
third-order dynamic range testing of modern rigs with dynamic ranges in
the 95 dB or better range. Too much phase noise in the typical lower
end generator. Strangely, to my knowledge, there is no current Agilent
generator made that is as clean in the phase noise department as an 8642
or an 8662/8663. Likewise there was no replacement for a generator with
a very high accuracy attenuator like in the HP 3335. On the same lines,
there was nothing to replace the HP 3585A/B spectrum analyzer that could
do as well as that fantastic box, within its limited range of audio to
40 MHz, including tracking generator.
73, Rob, NC0B
John T. M. Lyles wrote:
>The Hp 8640B is a superior instrument to the Wavetek 3000/3001 etc,
>but as someone pointed out, caveat is that the 8640B is painful to
>repair. It has a tunable cavity for the RF generator, explaining the
>spectral purity. The prices of these things have dropped over the
>years, used ones were about $1500 in the mid 1990s, and now they can
>be had for $300 if you are lucky.
>
>Things to watch out for are noisy output attenuator (or burnt ranges)
>and a bad pot for the variable attenuator. This assembly can be
>removed as a unit, if you get underneath the unit. Depending on which
>model, sometimes the oscillator stops (on mine) at the ends of the
>freq bands, but it always restarts if I just move around the crank
>and come back.
>
>I have the militarized black face version in the heavy yellow box.
>They are very easy to find since DOD has quit using them. It lacks
>the frequency lock feature that locks the osc to the counter
>timebase. It does have the freq counter, though, as if i leave it on
>for a long time, it is damn stable. The 8640A had a slide rule dial
>and lacked the counter altogether. Avoid that one.
>
>Later, HP developed the synthesized 8642A and B, which had similar
>performance to the 8640 in most respects, but was much more complex
>and used microprocessors. It is a very heavy and deep unit, and last
>time I checked, was still supported, due to the DOD users. It has
>built in fault diagnostics and self test features.
>
>73
>John
>K5PRO
>_______________________________________________
>Amps mailing list
>Amps@contesting.com
>http://lists.contesting.com/mailman/listinfo/amps
>
>
>
>
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|