Looking at the output of my Portsdown on a spectrum analyser at 3kHz resolution BW, I'm seeing spurs when using 125 and 333 kS/s. They seem to be related to symbol rate and not (for example) to ripple on the power supply (though I'm happy to be corrected). They are not visible at 1000 kS/s and above.
Would appreciate any views on whether this is expected / acceptable?
Many thanks and 73
Ray
This is the fairly clean SR 1000 spectrum for comparison. All the screenshots are 5MHz / division.
Here is the SR 333 spectrum. The nearest spurs to the carrier are at + and - 5.533MHz and are about 47dB down on the carrier.
And here is the SR 125 spectrum. The nearest spurs are at + and - 2.066MHz and 4.133 MHz and are about 55dB down on the carrier. I noticed that 333/125 is pretty much the same ratio as 5.533/2.066.
Spurs on output spectrum
Re: Spurs on output spectrum
Hi Ray
Yes, these are expected, and we consider that they are acceptable. They are filter clock breakthrough on the signal path. We put in extra filtering (the V1A modifications) after the first batch of boards to reduce their level to where they are now. If you compare them to the main carrier when unmodulated, we found them to be well below-50dB.
As the filter clock is not used at 1MS and above, it only affects the lower SRs.
Dave
Yes, these are expected, and we consider that they are acceptable. They are filter clock breakthrough on the signal path. We put in extra filtering (the V1A modifications) after the first batch of boards to reduce their level to where they are now. If you compare them to the main carrier when unmodulated, we found them to be well below-50dB.
As the filter clock is not used at 1MS and above, it only affects the lower SRs.
Dave