Re: Today's News
- Hi Harold and All,
Great question. If the Master Oscillator were "rock solid", the
minimum bandwidth final xtal filter would be about 3 Hz. This is due
to the "uncertainty" of the DDS 1. In reality, the Master Oscillator
is going to "dither" +/- about 50 Hz once it is warmed to
stabilization, and very slowly, at that. This means that the
frequency of LO 1 (PLO1) is going to dither approximately 781 Hz at
it's low end (MSA commanded to 1000 MHz) and 1562 Hz at the high end
(MSA commanded to 1000 MHz). LO2 is fixed at 1024 MHz and it will
dither about 800 Hz (at exactly the same direction as LO1). Therefore
the final I.F. frequency is going to dither about 19 Hz at the low end
and 762 Hz at the high end.
Now, this Master Oscillator dither is quite slow and is a function
of supply voltage and more importantly, the Master Oscillator
temperature. If it is well thermally isolated, this dither rate can
be as slow a few hertz per hour (after a 30 minute warm up). And best
of all, the software will allow you to "calibrate" to the final
"resting" frequency of the Master oscillator. The frequency
measurement of the MSA can be as accurate as +/- 3 Hz at the low end
and 6 Hz at the high end.
You wanna use 500 Hz. Go for it. I built and tested a 180 Hz
bandwidth final xtal filter in the MSA for a trial test. I didn't
really like it because it had more loss than I cared for, but it
proves that the MSA oscillators have low enough phase noise (low
dither) to allow for extremely narrow filters.
So, how do you see this Master Oscillator "dither"? Run the MSA
program from cold. Input an extremely stable RF signal into the MSA.
Command the MSA for the input signal to be the center frequency.
Command the MSA sweep width for about 10 times the bandwidth of the
final xtal filter. Watch the resulting waveform to "walk" from either
the left or the right as the Master Oscillator warms up. When thermal
stability is reached, the waveform should be in the exact center of
the graph. (If it is not, then go into the code and change the Master
oscillator frequency and start over.) Now, watch that waveform over a
period of several hours and see how much it "dithers" around the
center of the Graph. Less movement, the better your Master Oscillator is.
The great thing is, even if your final xtal filter is too narrow for
your MSA "dithering" it will still give you a result in the Graph
Window. It may not be in the center of the Graph, but it will still
By the way, if any of you have "non-phase locked" Spectrum
Analyzers, try this test. You will pull your hair out. Years ago, I
tried to get my department to upgrade our old HP141 SA's. Our cheap
#$$%% Director said no. I took him back to my lab and showed him this
test and he approved my $500,000 budget for new SA's.
--- In email@example.com, "Harold E. Johnson"
> Scotty, (for planning purposes) what would your guesstimate be as tothe
> minimum filter bandwidth that can be supported considering theoscillators?
> I have 35 KHz, 8 KHz and 4 KHz filters in hand, wondering whether tobuild a
> narrower filter at perhaps 500 Hz.
Ha ha! It is much to early to begin defending yourself. I haven't
looked into it yet. Besides, whining does not become you!
--- In firstname.lastname@example.org, Bob Fish <rwfish@...> wrote:
> Uh- Oh
> > __"" The PDM output is not "quite right" so there is an area I
> > on next.,_._,___
- It's Funny,
I had a job once where this engineer would bring me his designs and it
was my job to breadboard up a prototype and see if I could get it to
work. If I couldn't get it to work, I would go find the engineer, he
would come sit at my bench and we would go through it together. If the
error turned out to be mine, he would raise an eyebrow and leave the
room without a word. He was a good engineer and a nice guy, but he used
to love to publicly needle me when I screwed up. If he was trying to
motivate me, it certainly worked. I always concentrated a little harder
when I worked on one of his designs.
> Ha ha! It is much to early to begin defending yourself. I haven't
> looked into it yet. Besides, whining does not become you!
> --- In email@example.com
> <mailto:spectrumanalyzer%40yahoogroups.com>, Bob Fish <rwfish@...> wrote:
> > Uh- Oh
> > >
> > > __"" The PDM output is not "quite right" so there is an area I
> will focus
> > > on next.,_._,___