Jens, your analysis is faulty from several standpoints.
First you fail to list several parameters such as accuracy and jitter. These are very important parameters in determining which type of circuitry to use.
The two lower frequencies are for audio output only, they are divided by such a high divisor that jitter does not play any significant role. Frequency stability is fine at 100ppm (cheapest crystal grade I can get). The 28.322 Mhz clock should be a little more stable, I have used 50ppm canned oscillators in all predecessors of the product. Jitter does not play a big role, the overall frequency stability is the cruicial value, but that's met by the 50ppm grade. The circuits only use the rising edge of the clock, so the 45/55% duty cycle that is produced by these spread-spectrum oscillators would not hurt at all (if I understood that technique correctly).
Second, you can use a PLL. It is not necessary to be harmonically related. You only need one common denominator (try 100KHz) . This will be the best solution if you need and phase coherency. However that may be a difficult design for 3 frequencies.
The three clocks drive independant circuits. Syncing is only necessary at interfaces that need Fifo memories anyway.
If you indeed have a solution costing $1.30 that works I would be inclined to stick with it. I can't buy three non-standard crystals or quality PLL components for that price.
These crystals are not really non-standard. 14.31818Mhz is 4 times NTSC video carrier (used on all PC mainboards today), the 17.7344Mhz is 4 times PAL video carrier, and 28.322Mhz is an old VGA frequency, I think I can even get overstock crystals of that frequency. The problem might be EMI - I thought with clock management chips being in fashion at many chip makers, I could replace my three discrete oscillators with something more sophisticated with spread spectrum. I managed to pass CE with the canned oscillators before, and they don't have spread spectrum at all, so it's nothing that I seriously worry about. It just would have made me feel better to design something that's more up-to-date - being a little vain, you know...
ciao, -- Jens Sch�nfeld
-- Author: Jens =?iso-8859-1?Q?Sch=F6nfeld?= INET: [EMAIL PROTECTED]
Fat City Hosting, San Diego, California -- http://www.fatcity.com --------------------------------------------------------------------- To REMOVE yourself from this mailing list, send an E-Mail message to: [EMAIL PROTECTED] (note EXACT spelling of 'ListGuru') and in the message BODY, include a line containing: UNSUB CHIPDIR-L (or the name of mailing list you want to be removed from). You may also send the HELP command for other information (like subscribing).
