from:
www ieee-uffc org/freqcontrol/quartz/vig/vigcateg htm (add your own periods)
A crystal unit's resonance frequency varies with temperature. Typical frequency
vs. temperature (f vs. T) characteristics for crystals used in frequency
standards are shown in Figure 11. The three categories of crystal oscillators,
based on the method of dealing with the crystal unit's f vs. T characteristic
are XO, TCXO, and OCXO, (see Figure 12). A simple XO does not contain means for
reducing the crystal's f vs. T variation. A typical XO's f vs. T stability may
be 125 ppm for a temperature range of -550C to +850C.
My $.02:
unless our timers are substantially more temperature dependent than "a typical
XO," then the 3rd digit should be reliable. For a 50 second run, 25ppm
=0.00125 sec. So, if one person runs when the crystal of the timer is at -55,
and the other runs at +85, the timer is going to shift 1 thousandth. Since
more likely temperature variations over an event are 10 to 35 (50 to 95 deg F),
the variation is pretty likely to be .05ppm. (1/5 the temp range, 1/5 the
variation)
(yes, I know that a hyperbola isn't a straight line, but for the area around 0,
we can approximate it as such.)
Since most runs of a class are done at the same time of day, the temperature
variation will be minimized. Small local temp differences (a cloud passing
over the sun, etc.) should be handled by the insulation effect caused by the
timer casing.
|