HP 8662A problem

[Garnet]

We have an HP8662A and when we go from 118.0000000 Mhz to 117.9999999
the output has some discontinuity during the change. Other frequencies
seem alright within at least 5 Mhz of this value. Does anyone know if
this is a common characteristic of these units or is our unit faulty ?
Thanks, Garnet.

 

[John Miles]

In article <1111653538.362371.141880@l41g2000cwc.googlegroups.com>,
garnethoyes@hotmail.com says...

We have an HP8662A and when we go from 118.0000000 Mhz to 117.9999999
the output has some discontinuity during the change. Other frequencies
seem alright within at least 5 Mhz of this value. Does anyone know if
this is a common characteristic of these units or is our unit faulty ?
Thanks, Garnet.

There's a whole lot of switching going on inside that box. The
reference range is switched in octave bands from 640 MHz down, and there
are several VCO tank circuits to keep the noise low at any given
frequency. How long is the glitch?

I've got a .PDF of the 1981 HP Journal issue that was dedicated to the
8662, and I don't see any switching logic in their block diagrams that
looks like it happens at 118 MHz. (http://www.speakeasy.net/
~jmiles1/8662hpj.pdf -- 12MB, I'll leave it up for a couple days then
yank it to save disk space).

If you're just seeing the signal drop out for a few milliseconds, I
suspect it's nothing to worry about.

-- jm

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

 

[John Miles]

In article <1111724875.096889.253570@o13g2000cwo.googlegroups.com>,
garnethoyes@hotmail.com says...

Thanks for the HP Journal - very interesting. Would be nice if the
colored text parts had scanned in as well though.

They seem pretty legible here... are you using a reasonably up-to-date
Acrobat Reader version on a 16/24/32-bit desktop display?

In our application if
the signal drops out for more than 270ns we loose everything and have
to start over so its a bit inconvenient. Guess we might be looking at a
replacement generator sometime soon ......

Yep, I think you're going to have to use a conventional sweep generator
for that application. The 8662 will definitely hiccup when it switches
VCO or reference ranges, and probably in a lot more places than that. I
think their time-to-frequency spec is a few milliseconds at the very
least, and the transition will be far from continuous.

-- jm

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

 

[Garnet]

Thanks for the HP Journal - very interesting. Would be nice if the
colored text parts had scanned in as well though. In our application if
the signal drops out for more than 270ns we loose everything and have
to start over so its a bit inconvenient. Guess we might be looking at a
replacement generator sometime soon ......