[Amps] Tiny SA ultra, part 2

[Jim Brown]

On 5/5/2023 9:28 AM, jim.thom jim.thom at telus.net wrote:
> Several folks told me that the Tiny SA was useless for IMD testing, due
> to it's 3 khz resolution.   Apparently somebody listened, and the newer /
> bigger version, with it's 200 hz resolution, should fare better.

That's quite coarse. The Elecraft P3 can get down to a few Hz; the SVGA 
card, which does a higher res FFT of the same data, gets down to 
fractional Hz. This is the sort of resolution needed to measure 
transmitted bandwidth.

Some measurements I did about ten years ago are shown in these pdf 
files.  The first, for a talk by Elecraft engineer Bob Wolbert, K6XX, 
was done before the SVGA card was developed.

http://k9yc.com/K6XXAmpTalk.pdf

The second file, which uses Power Point as a convenient means of showing 
graphical data, was not intended for presentation to a club or at an 
event. The measurement method is mostly described with text slides.

http://k9yc.com/P3_Spectrum_Measurements.pdf

The P3, with or without the SVGA board, measures 100 dB of dynamic 
range, but can display only 80 dB on the screen at a time. It turns out 
that 80 dB is more than enough for even the best rigs I've measured, but 
if a cleaner rig showed up (or perhaps with Pure Signal), the bottom 20 
dB could be displayed simply by changing the zero reference for the 
display.

The various SDRs shown in the talk and text pdfs on chasing RFI by 
studying their spectra, mostly have very good frequency resolution. 
NR0V, who authored both Pure Signal and computer software for ANAN 
radios, told me that the little 10W Anan rig is good for fractional Hz

http://k9yc.com/KillingReceiveNoise.pdf  Text
http://k9yc.com/KillingRXNoiseVisalia.pdf  Slide show

I used the P3 and SVGA board because they could do the job and I had 
them. Both are discontinued by Elecraft. If I were doing those 
measurements today, I'd probably use the ANAN.

I own both HP8590D and Rigol spectrum analyzers. Their frequency 
resolution specs are comparable, MUCH to wide (at least two orders of 
magnitude) for transmitted bandwidth measurements of transmission modes 
used by hams.

73, Jim K9YC