11-12-2014 07:46 PM
Y(s) = W^2/(S+W)^2 X(s) (Pardon my Greek)
I thought I had a fairly good demo of this filter. no overshoot- no stastical deviation from a "Simulink" model implementation (OK I have a "Simpson's Rule" error between the Simulink model and the discreete implementation (No, I cannot show the code)
The question comes down to more: "How do I educate the Engineer?" What I heard in the discussion was "What good is LabVIEW if I cannot drop a discrete transform block?"
OK, I can configure a 2nd order filter easilly enough but, how do I prove the implementation of an n-Order filtet in LabVEW meets his critea for a 2nd order function out of matlab? (Yup- that lvann*.dll sure does not document methods specific to filter order)
Setting up a discrete "Classic" fiter results in a non-critically dampend system- That really is bad in this case!
So, I'm not as "Trustworthy" as the Mathworks or National Instruments in developing this incredably limited filter but, I cannot find a generic filter to replace the Simulink block.
-All Ears!
11-17-2014 02:26 PM
Hi Jeff -
Would you mind clarifying your question? Are you looking for the .m version of the LabVIEW filter you created? Or are you looking for a LabVIEW block to replace your .m function? I'd like to help but I'm not quite sure what you are asking.
11-18-2014 07:37 PM - edited 11-18-2014 07:38 PM
@HollyBerry wrote:
Hi Jeff -
Would you mind clarifying your question? Are you looking for the .m version of the LabVIEW filter you created? Or are you looking for a LabVIEW block to replace your .m function? I'd like to help but I'm not quite sure what you are asking.
Sorry, This fell a few screens in my history list. Actually, The point has become moot once I requested the engineer change the solver settings to a 50 Hz update rate in simulink to better simulate the RT system on which it will deploy. (Face-Palm!) It should have been done in hardware! - Lesson Learned!