[Elecraft] KAF2 - accurizing the RTC

[Randy Tompson]

Adding my experience to the previous notes on accurizing the Real Time 
Clock  (see October 15th and 17th, 2001 Elecraft Digest archive).

I tacked an 11pF cap (a 10pF ceramic, measured 11pF on the meter) in 
parallel to C19.  A previous writer added 14pF, correcting a 1.3 sec/day 
(fast) error to 0.5 sec/day (fast).

My clock had been running about 0.6 sec/day fast (8 seconds over a 14-day 
period).  Adding 11 pF brought it down to under 0.07 sec/day fast.  It has 
been two weeks now, and it's under one second fast.

Your accurizing value may vary, depending on the exact frequency of your 
crystal.  The specified tolerance is +/- 1.7 seconds/day, according to 
Wayne.  And your C19 may have a higher or lower capacitance than mine... 
But perhaps a value of 20 pF per second/day of correction would be a good 
place to start.   (My accuracy improved from 0.57 sec/day to 0.07 sec/day, 
a movement of 0.50 sec/day, for 11 pF = 22pF per sec/day.  The other fellow 
in the EleDig got 1 sec/day correction per 18pF).

By the way, I use the "skip a second" indication as my mark point.   The 
RTC display skips a second every 7-9 seconds.   That means that the point 
of a given second's popping up on the display is not on the beginning of 
the second, but somewhere in the middle.  To keep the measurement 
consistent vs. the reference (e.g., WWV broadcast), it's best to use a 
consistent marker point.  This "skip a second" point serves that purpose.

72's,

WA4CIT

h