[time-nuts] A Time-Nut's Worst Nightmare
magnus at rubidium.dyndns.org
Sat May 11 07:32:53 EDT 2013
On 11/05/13 06:48, Jim Lux wrote:
> On 5/10/13 6:52 PM, Ed Palmer wrote:
>> Part of me thinks it's cute, part of me wants to kill it. :-)
> but.. what is the actual distribution? Is it white phase or white
Instead of added wideband noise-modulation, think about more systematic
Notice how it only affects on every second. Also note how after lagging
it makes huge jumps. Also notice how it does not jump backwards.
What it tells is is that it makes a random decision wither to jump
forward or not, and if to jump forward one sec or the error sec. Keeping
the amount of time it lags, it can then every now and then jump into
place. It would make sense to have a maximum lag so that it is forced to
jump into place. There is a practical amount as to how high speed you
run the handle without there is a risc of it skipping past the intended
second without you noticing, so let say that it does not do more errors
than 5 seconds or so to keep it manageable.
The end result will be a clock which in long term is showing the right
time, but have short term variations. Since it is a lag scheme, it will
also on average be behind.
Not to hard to achieve with a PIC and not to hard to implement, but
highly annoying none the less.
Looking forward to seeing NIST having one of these in their lobby. :)
More information about the time-nuts