italie |
|
MAME owes italie many thank yous, hah
|
|
|
|
|
|
Reged: 09/20/03
|
|
Posts: 15246
|
|
Loc: BoomTown
|
|
|
|
Send PM
|
|
|
And now, the $0.10 part they are putting in your clock. An uncompensated, 10ppm quantified piece of crap. You have little to no control over aging, phase shift, jitter, hysterisis, etc. They are guaranteeing your spec based upon averages in most cases. The error you are compensating for today might drift out by next week. You might get +.2 second accuracy today, and -1.2 tomorrow.
You might get lucky, and room temp might fall on a linear part of your response curve and leave you with a relatively predictable crystal. You might not get luck just as easily. There are no promises about anything, other than +- 10ppm.
Show, and not to scale, a plot of all three we've briefly touched on. In real-world plotting, the green "uncompensated" curve would be ~10x larger than the other two.
[ATTACHED IMAGE - CLICK FOR FULL SIZE]
|
|
Index 
Flat 
Xtal compensation theory 101: Part 3
Reply
Xtal compensation theory 101: Part 1
