On Sun, January 29, 2012 12:40 pm, Nathan House wrote: >> If you REALLY don't want to use a PLL, use an oscillator. Tons of >> options out there. I regularly use 100 and 125MHz oscillators. > > I don't have anything against using the PLL :-) Are there advantages > to using oscillator, aside from a higher speed? The 2-16 MHz range has historically been a sweet spot with respect to low cost and accuracy. The higher you go, the physically smaller and thinner the crystal had to be, so it was more sensitive to microphonics and damage from physical motion. Go lower, and the crystal gets proportionately bigger (I believe that your typical 32KHz watch crystal is a somewhat different technology than what you see in the MHz range, so it isn't quite apples-apples). > I found this excellent page on crystals, > http://www.sxlist.com/techref/clocks.htm, and have a few more > questions! The page I just linked to suggested splitting the > calculated capacitor into two separate capacitors (on each side of the > crystal) and connecting one to GND and one to VCC. Is that good > advice? Honestly, I've seen many, many crystal circuits, and I've never seen the GND/VDD split used. At the frequencies crystals run, GND and VDD are equivalent (if they're not, you've got bigger problems than crystal stability to face). It makes routing much harder, as well. If you think about what the crystals are doing, there should be a 180 degree phase shift between ends of the crystal- whatever current is going in to one crystal should be coming out of the other- I'd think that keeping both capacitors tied to the same plane would be better as it keeps the overall loop of current much smaller. > Also, say I want to use the PLL to bump the system clock up to 48 MHz. > If I used an 8 MHz crystal instead of a 4 MHz crystal (assuming both > are valid options to use with the PLL), would that cut down the > jitter/noise by a factor of two? Ideally, yes, but crystals aren't particularly known for their cycle-cycle stability- their stability is more apparent in a longer term sense. Cycle-cycle, RC oscillators can be quite stable. If you want to learn lots about oscillators, look up papers by John R. Vig (here's one for example: = ). In a real circuit, without a $100,000 analyzer, you'll probably never notice a difference. Matt Bennett Just outside of Austin, TX 30.51,-97.91 The views I express are my own, not that of my employer, a large multinational corporation that you are familiar with. --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .