There is an error in all the example programs using differential channels to measure the four radiation output (CNR4 instruction manual revision 11/10): the VoltDiff function is called so that it will use only the first calibration constant cnr4_mult(1) and not the 3 others
VoltDiff (cnr4(1),4,mV25C,1,True ,0,_60Hz,cnr4_mult(1),0)
The multiplier argument should be mult()(1) otherwise three out of four CNR4 radiation measurements will be wrong (unless the multipliers are all equal..). This error is difficult to detect in the output data especially if you check only the net radiation or albedo.
We found the same error in the default program generated by the Shortcut software.
* Last updated by: SimonG on 7/25/2013 @ 2:34 AM *
Have you tried out the "mult()(1)" syntax and confirmed that it works properly ? I have an email thread that indicates the desired syntax might be "mult(1)()" instead.
You are absolutely right!
the correct syntax is
mult() or mult(1)()
in both cases the multiplier is incremented with each repetition
Thank you
Your question is resolved (I'm glad it worked for you), but I will provide some additional information here.
The longer syntax is for use when you need to start in the middle of the array.
If mult() is an array of size 10, then the datalogger should be able to utilize all 10 by specifying "mult()" as a starting point and reps of 10. (this is the basic use case)
However, if I wanted to change just four items in the middle [for example, elements 5 through 8 -- mult(5), mult(6), mult(7) and mult(8) ], I should be able to specify "mult(5)()" and set the reps to 4.
I think that example shows a little more clearly what that syntax is intended for.
In summary, you have a way to
A) set all 10 to the same thing using "mult(1)" or "mult(3)"
B) cycle through all 10 using "mult()" and reps of 10.
or
C) just change a few using "mult(5)()" and reps of 4.
Just in case someone reads this and thinks the original error report of there being an error in the manual, was wrong, it is not. The debate about the correct syntax that should be in the manual slightly distracts from this. The manual and Shortcut do appear to be incorrect in the way the calibrations are applied.
I have raised various requests for changes to be made to the manuals etc.
We have checked and this error is not in Campbell Scientific supplied programs for Eddy Covariance systems.
The change in SCWin's CNR4 sensor file to correct the problem with the calibration values is in a new release that should be posted to the web site by the end of the week. When it's posted, you will be able to find this new version of SCWin on our website at www.campbellsci.com/downloads .
Note that if you are signed up with our Customer Center, you can choose to get email alerts when new downloads are posted.
Dana W.