Labview gurus,
Here's the problem:
I am currently collecting photoacoustic waveforms from my PXI-5152 using an ultrasonic transducer and I am normalizing the output by measuring the laser energy photoacoustically using another ultrasonic transducer. Mathematically, I am taking Voltage (experiment) /Voltage (laser energy) and then I am plotting this vs. angle of incidence so that I can get photoacoustic signals from a material at varying angles of incidence.
Essentially, this gives me an inverse curve for Fresnel reflection as shown here: http://en.wikipedia.org/wiki/File:Fresnel_reflection.svg (So I get transmission instead of reflection, etc.)
What I've noticed is that at higher angles of incidence (i.e. smaller photoacoustic signal since there is more reflected light) I seem to have a less noisy curve than I do at lower angles of incidence. I realized that this may have to do with the fact that the number of digitization levels (8 bit digitizer at 100 mV range so each level is 0.390625 mV) changes dramatically since at lower angles my signals are ~100 mV (~256 levels) and at higher angles my signals are ~10 mV (~25 levels). This suggests to me that the lower noise I observe at higher angles has more to do with digitization error than anything.
Question: How can I use Labview to make the acquisition of voltage levels equal for different voltages? For example, if I wanted both 100 mV and 10 mV to be displayed by 25 levels as opposed to how it is now?
At first, it might seem counterintuitive to want less resolution for some of the measured voltages, but the issue is that I am using a curve fitting algorithm over all the points and I would like to keep the error approximately consistent in % over the range since this will affect the curve fitting, etc.
So, how do I do this? Btw, not that anyone would care, but I just graduated with a PhD in Bioengineering, so huzzah, one more engineer has entered the world.