I was thinking about spectral purity standards and the Si5351 chip. I realized that I didn't even know what the FCC standards for "close in" noise are. The standards for spurious emissions ARE well known, but these are for harmonics and parasitic emissions relatively far from the desired signal. What about unwanted signals CLOSE to the desired signal?
My old 2002 ARRL handbook indicates that the FCC has not established firm standards for this "close in" noise. (They call it "out of band" noise, but are clearly referring to noise that is close to the desired signal but spreading out beyond the desired bandwidth. Phase noise would be in their category.)
In the course of my Googling, I found the above spectral purity mask. I don't know where it comes from, but I think it is the kind of graph that would be very useful to us as we evaluate the merits and shortcomings of various frequency synthesizers. Would our DDS or PLL rigs fit in this mask?I think an Si5351 rig WOULD. According to KE5FX's measurements, at a mere 100 Hz from the center frequency, the Si5351 phase noise is already -90 db.
Does anyone have a similar mask showing current standards?
I still don't understand why so many folks believe that the Si570 is a useful part for homebrew rigs, but the Si5351 is not. Look at the numbers:
Si570 Clifton Labs measuring at 30 MHz carrier. At 10kHz from carrier: -109.6 dbc/Hz Silicon Labs web site (carrier freq not specified) At 10 kHz from carrier: -116 dbc/Hz
Si5351 KE5FX measuring at 19.99 MHz. At 10kHz from carrier: -127 dbc/Hz Silicon Labs measuring at 156.2 MHz. At 10 kHz from carrier -112 dbc/Hz.
Can anyone out there explain the technical basis for the belief that the Si570 is a useful part while the Si5351 is not?
It is important to keep things in perspective. ALL of these noise numbers represent VERY small noise levels. Let's keep is simple and assume a 100 watt carrier signal and a phase noise of -100 dbc/Hz. That means the phase noise per hertz would be .00000001 watts. That's watts/hertz. How much "noise power" would that represent in a typical SSB passband? Multiply by 2500 Hz and you get 25 microwatts. That's really low noise levels. Not enough to worry about. And as we've noted, we've happily used rigs with LC VFOs and crystal oscillators for all these years without every once measuring their phase noise.
In response to popular demand, "SolderSmoke -- Global Adventures in Wireless Electronics" is now available as an e-book for Amazon's Kindle.
Here's the site:
For the print version:
For shipping from a printer in the U.S. (probably better for N. American buyers) Click here: SolderSmoke USA Version
For shipping from a printer in the UK, Spain, or the USA (probably better for UK and other European buyers)
Click here: SolderSmoke EU Version
The two versions are identical, except for a minor difference in the paper used. That's why the prices are a bit different.
Bill's OTHER Book (Warning: Not About Radio)
Click on the image to learn more
W4HBK's QRSS Grabber: The Amazing Pensacola Snapper (Live!)