[wxqc] Non-linear Humidity Sensor

[jwilson]

Hi,
 
I have a question on how to handle a non-linear humidity sensor.  I have
occasionally been getting error reports indicating my dew point is in error.
In order to see if I could determine a better calibration for my system
(Peat U2000) I purchased a sling psychrometer and have been running tests
periodically for the past two months.  What I have discovered is that the
humidity sensor is non-linear.  Particularly at low humidity readings,
although it is also off at high humidity levels also.  This information
seems to correspond to the days I get the big red X on my QC data.
 
I determined the non-linearity by plotting the data from the psychrometer
and from my weather station in an X-Y scatter plot and then applying a trend
line to the data (using Excel).  The "best fit" provides me an equation of y
= -0.0041x2 + 1.6786x - 25.872, where x is the psychrometer data and y is
the weather station data.
 
http://pond1.gladstonefamily.net:8080/cgi-bin/wxqchart.pl?site=AP677
 
Does anyone have any ideas as to how to correct this problem?  Is this a
common problem with humidity sensors?
 
Thanks
 
Monty Wilson, AP677
jwilson at ncfcomm.com
 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://pond1.gladstonefamily.net:8080/pipermail/wxqc/attachments/20060415/aab631bb/attachment.html