[time-nuts] Digital temperature compensation

[Alan Melia]

Mmmm yes you can see the equation evaluation starting to rise in your 
"Warmer" plot, as Mark says, which will make a nonsense of the formula if 
your summer temps get above 28.

Why not a table and then interpolate between the table data points?. You 
might have more points where the changes are greater.
The colder plot looks cubic maybe for a crystal made for 20 deg C ?? But 
depending on the oscillator electronics you may have component tempcos 
affecting the frequency as well?
I suspect the turnover at 21deg C should be a smooth curve not as your 
formula predicts. Which suggests that you have too too high an order of 
polynomial I think, but you may not get a good fit with a cubic if other 
effects are present.

Alan
G3NYK


----- Original Message ----- 
From: "Dan Drown" <dan-timenuts at drown.org>
To: <time-nuts at febo.com>
Sent: Saturday, November 01, 2014 10:07 PM
Subject: Re: [time-nuts] Digital temperature compensation


>I gave zunzun a try and the one with the lowest root mean squared error 
>was:
> f(x) = a( x**0.5) + b( x ) + c( sin(x) ) + d( cos(x) )
>
> It got 0.202 RMSE, so I guess I'll stick with my original function as  it 
> seems to be closer to what I expect will happen at colder/hotter  temps.
>
> You have a good point about temperatures outside my data samples.   Once 
> it gets hot again in the summertime, I'm sure I'll have to  re-evaluate 
> this.
>
> Quoting Mark Sims <holrum at hotmail.com>:
>> You could try submitting your data to zunzun.com   It will fit it to 
>> around 40,000 different curves and find the best ones.
>>
>> Beware that with all curve fitting formulas,  once your live data  starts 
>> to wander out of the range of your original curve fit data,   things can 
>> go rather badly...
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to 
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.