I do not have any sort of high resolution AC voltmeter. I have the Tek DA4084 that provides 3 digit precision, and an old Leader LMV-181A analog meter that is sort of good to 3-digits. So, sitting around trying to figure out how I could in some manner quantify the PCGU1000’s ability to resolve output amplitude, it came to me that the Bessel nulls of an FM wave are quite sensitive to the RMS value of the modulation frequency–and that I had the necessary instrumentation to observe same.
So I set up my Leader 3215 AM/FM RF generator for a 101.1 MHx carrier (my usual pirate FM frequency) and fed it a 13.587 kHz signal from the PCGU1000. 13.587 kHz is the 2nd Bessel null frequency for a 75 kHz deviation of an FM carrier–I.e. when an FM carrier is modulated at 13.587 kHz, at a specific level, the carrier will be nulled out and only the sidebands will remain in the frequency spectrum.
The Leader RF generator expects the external modulation signal to be 1.0 Vrms, so that was the first value I applied. Here is the spectrum of the resulting wave (PCGU100 at 13.587 kHz and 1.0 Vrms).
Hot damn, you can see that the carrier at 101.1 MHz is gone, 40 dB down and completely nulled out on the first try (the PCGU1000 frequency was tweaked by the 1.000015 factor I had previously established). The first upper sideband is at 101.114 MHz where it’s supposed to be.
To see what effect adjusting the RMS output value might have I tried setting the modulation wave 0.995 Vrms.
You can see the carrier starting to pop-up, 31.1 dB down from the sideband. At 0.990 Vrms (below) it’s a bit higher at only 25.2 dB down from the sideband.
Going the other way I tested 1.005 vrms and 1.010 Vrms
Both producing delightfully similar results (-32.3 and -25.2 dB) on the “uphill” side of the null…
So, the bottom line is that the PCGU1000 can resolve amplitude settings down to 0.005 quite nicely. In case anyone remains confused, I really like this thing! I’d like to get Velleman’s matching 'scope and see what the pair can do, but this is not likely to happen on my mostly-fixed, semi-retired income, with things the way there are…
If anyone cares, the spectra were gathered via an Instek GSP810 spectrum analyser, using software I created for that instrument. These are .png files created directly from the application.