required accuracy for crystal oscillator
Propability
Posts: 142
·OK so I did in my pll on my protoboard and I want to see if driving it with a 80 Mhz crystal oscillator will bring the video back to life. I see there are varying degrees of accuracy from 100 ppm to 25 ppm with the 25ppm being the costliest of course. If the crystal (5Mhz) might have a 10ppm (might be a little higher -just going by memory)· and with the pll set at 16x what is the accuracy of the clock? Is it 16times 10ppm and if so is a 100ppm crystal oscillator good enough for video.
Pete
Pete
Comments
The accuracy of an external oscillator is both long-term (how close is the frequency to that stamped on the case?) and short-term (how much does the frequency vary with temperature, power supply voltage, aging?) In both cases, video depends on a fairly stable clock over short periods of time (fractions of a second to seconds). Absolute frequency accuracy needs to be on the order of percent and tenths of a percent.
A 100ppm crystal oscillator should be fine for video.
Pete
www.mouser.com/Search/Refine.aspx?FS=TRUE&Ntt=*hc49us*+*6MHz*&Ne=688437+688748&N=5115911+4294622349+4294622948+1323038&Ntx=mode%2bmatchall&Ns=P_SField&OriginalKeyword=hc49us+6MHz&Ntk=Mouser_Wildcards
I am not sure if it is in the right capacitance range though at 18pf.
I'm out of luck trying to use a crystal (still have the 5 Mhz that came with the protoboard) and the PLL (at 16x) to drive any video since I somehow broke the PLL in the Prop. It does run with the crystal but not with any of the PLL settings. I was just curious as to the needed accuracy the clock had to be since the 100 ppm is about $1.60 versus about $3.80 for a 25ppm crystal oscillator which runs at 80Mhz which will bypass the PLL , so hopefully· I can get the video back without having to remove the Prop chip from the board, which has been populated with a few connectors I don't want to remove.
Pete