[time-nuts] Curve fitting

Mark Sims holrum at hotmail.com
Wed Dec 23 22:46:38 UTC 2009


Here are a couple of images of what can be done with a cardboard box,  a small fan,  and a DC solid state relay (or transistor)...   

The first is of the DAC voltage over 8 hours with the temperature stabilized (note with an RMS error of under .001C over 8 hours...  much of which is due to those false temperature sensor spikes).   

The temp controller PID parameters were not tuned for the box/fan configuration.  With optimized values and no furnace to contend with,  I have seen rms errors in the low tens of microdegrees over 8 hours...

It also shows the FFT of the temperature data over that time...  some temp control wobbles around 1200-1400 seconds due to furnace cycling).

The second image shows the calculated oscillator drift at -2.239E-9 parts/day.  The osc tempco number is rather bogus since it needs to be calculated when the temp is moving a lot more. 		 	   		  
_________________________________________________________________
Your E-mail and More On-the-Go. Get Windows Live Hotmail Free.
http://clk.atdmt.com/GBL/go/171222985/direct/01/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: drift.gif
Type: image/gif
Size: 33824 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20091223/792847a8/attachment-0002.gif>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: drift2.gif
Type: image/gif
Size: 23161 bytes
Desc: not available
URL: <http://www.febo.com/pipermail/time-nuts/attachments/20091223/792847a8/attachment-0003.gif>


More information about the time-nuts mailing list