Hi,
I do not know that brand at all, I will just extrapolate from what I already encountered, probably not everything is relevant in your case.
I made up a simple line grating with a pitch of 60 um (6 pixels) and diffract the zero-order to the 1st order when the grating is On. I then changed the level of phase modulation at all lines (increase the value in the csv file) and observe increasing power at the 1st spot.
That is not a bad idea, however, I would do it a bit differently. By putting a grating, you end up with a spatial square signal, that has many harmonic components in the Fourier domain, hence multiple order of diffractions.
As you have phase SLM, I would do a phase ramp. Over N pixels, I would linearly increase the pixel value between 0 and V_max and I will then increase V_max gradually. I will repeat this ramp to have it periodic of course.
If V_max corresponds to 2pi, you have a tilted plane wave with ideally (if everything was perfect and pixels infinitely small) one only spot (+1 order) in the Fourier plane.
Well, you will always have intensity in the 0-order due to the limited diffraction efficiency and filling fraction, but the other orders should be very small.
That being said, your experiment should work too, It is just not as easy (visually) to find the maximum of intensity in the +1 compared to a minimum elsewhere.
So, what can go wrong?
First, the calibration is usually given for one wavelength, meaning that increasing linearly the pixel value increase proportionally the phase for this wavelength, but for another one, the relation is not linear anymore. That is probably not the cause, but it is worth keeping in mind.
Another thing is that SLMs do not have a plane response, meaning that if you send a constant value, you do not have a plane wave. Companies usually provide a correction mask, that also depends on the wavelength.
Moreover, the value for pi depends also on the wavelength. The phase shift is proportional to 2pi/lambda*delta_n, with delta_n the birefringence induced by a given voltage. So for the same voltage (pixel value), you have much less phase modulation at 1550 nm compared to visible. So if the pi value is 500 at 775nm, it will be 1000 at 1550nm.
So most likely, your experiment worked and you find the pi value around 1000.
I may be wrong, but that would be my first guess.
Best,
Sebastien