spectrophotometer

[toni23]

I Hope somebody can help us.

We have to make a program for a spectrophotometer with labview. We have to measure the pick place, pick high and pick surface....
We take a power supply to generate or pikes.
Now we make a beginning, but we don't know how to generate a pickplace and pick surface

http://www.techzine.nl/f/g/45909php9IrMAM.JPG

maybe can somebody help us....

[DFGray]

Could you please define pick place, pick high, and pick surface.  I would be surprised if LabVIEW could not do what you want.  I have interfaced to spectrophotometers in the past, so should be able to give some help.

From your code, it appears you are trying to find the highest peak and integrate it to find its area.  It is unclear whether you are trying to find all the peaks, or just the highest one.

Please give us a few more details and we can help you out.

Message Edited by DFGray on 04-04-2007 08:40 AM

[shoneill]

Regarding "pick place" "pick high" and "pick surface", I believe he means peak center, peak amplitude and peak area.

Beyond that, I don't know.

Seems like a gauss fit via Lev-Mar would be the best option......

Shane.

[DFGray]

Without some representative data, it is hard to say what the best approach would be.  As a first try, I would use Waveform Peak Detection to find the peaks and valleys of the waveform.  I would use the valleys with a robust curve fit to generate the background.  The type of curve (line, polynomial, etc.) would depend on the background.  I would then integrate each peak using Integral x(t).  The only real problem here is determining the interval over which to integrate.  That could be hard-coded (if the peaks are a known width due to width being determined by the spectrophotometer) or determined using some sort of algorithm (Gaussian Peak Fit using bisquare or least absolute residual would work if the peaks are symmetric, some sort of level determination if not).  I would not use the gauss fit coefficients to do the integral unless I knew the peaks were gaussian in shape, since distortions in the peak shape could lead to unacceptable errors.  In addition, overlapping peaks can cause major errors as well.

Take home message - without some representative data, we can't really supply more than generalities which may not work.