[time-nuts] Problem with HP 83623A 20 GHz sweep generator stepping up/down 100 Hz when not wanted.

Dr. David Kirkby (Kirkby Microwave Ltd) drkirkby at kirkbymicrowave.co.uk
Fri Dec 23 11:57:14 UTC 2016


On 22 December 2016 at 20:54, Bill Byrom <time at radio.sent.com> wrote:

> Read about the synthesizer design of that generator here:
>
> https://ia600407.us.archive.org/10/items/Hewlett-Packard_
> Journal_Vol._42_No._2_1991-04_Hewlett-Packard/Hewlett-
> Packard_Journal_Vol._42_No._2_1991-04_Hewlett-Packard.pdf
>

Cheers, I will do that.


>
> Dave, looking at the graph you posted it appears to me that the shift is
> 10 Hz, not 100 Hz. It appears to me that is an audio spectrogram. I have
> no idea if the frequency shift is in the generator or receiver.
>

Well done!. I thought I was told it was 100 Hz, but looking at the diagram,
it is closer to about 10~15 Hz. The setup his end is a software defined
radio, so the measurement of frequency is being done at audio.
Unfortunately, I've been unable to get a response by phone, so I can't seem
to get much in the way of technical details at this minute.

I have an HP 5342A -- 10Hz to 18GHz Microwave Frequency Counter with 1 Hz
resolution, but I belive this is moving (it is is), is quicker than the
time it would take to measure to 10 GHz to 1 Hz. I can lock that to GPS
easy enough, so the signal generator would be fed from the same reference
as the frequency counter, but I don't think that would help.

It would appear that measuring such small changes at microwave frequencies
is not easy. Mixing with very good quality signal generator or known
performance would probably be ok, but even old 20 GHz units are expensive.
To get a good quality new signal generator would be a small fortune.

I don't think even sending this to Keysight for calibration would achieve
anything useful. I have not looked at the specification in detail, but
phase noise is specification is -76 dBc at 10 kHz offset at 20 GHz. The
calibration of this is high too ($1600 in the USA).

>
> Bill Byrom N5BB
>

Dave



More information about the Time-nuts_lists.febo.com mailing list