Holoeye SLM Calibration

[yyyaaaaaaa]

Hi！
I am currently using a Pluto NIR-149 SLM. I have loaded the factory-provided LUT file:

NIR-149_1065nm_9-5_linear-2.00pi_0.95V-3.30V.

As per the filename, I have configured the voltage limits to 0.95V - 3.30V in the driver settings. In my software (using slmsuite), I have set the operating wavelength parameter wav_um = 1.035 to account for my 1035 nm laser source.

However, when I display a blazed grating to test the diffraction efficiency, I am observing a significantly high zero-order intensity (approximately 35%).
I have also attempted to adjust the polarization angle of the polarizer, and found that 35% is the minimum zero-order intensity I could achieve.

Does this indicate an incorrect calibration or configuration? Could you please provide any suggestions to resolve this?

[ichristen]

Hi @yyyaaaaaaa ,

It sounds like you're doing everything correctly apart from a small issue that the design wavelength should be wav_design_um = 1.065 according to the LUT. But this is a small effect and would not explain your large 0th order intensity.

What angle are you trying to blaze to? If you have a short blaze period (<10 pixels), then your blaze might be blurring in a way that has a significant 0th order component.

Otherwise, I suspect the issue is on the hardware side.

• In the future, we'll actually finish up the code which can generate a LUT pixel calibration inside slmsuite (TODO: finish analysis for pixel_calibrate @ichristen @cpanuski ).

Best,
Ian

[yyyaaaaaaa]

Hi @ichristen ,

All these configurations are targeted for 1.065 nm. I have reloaded the default settings and applied the same LUT, except that the voltage limits remain at their defaults (0, 3.6 V) instead of those indicated in the filename (0.95, 3.3 V). This resulted in a lower 0-level light intensity (around 3%) and a higher 1-level light intensity (approximately 75%), which is consistent with the 71-76% range specified in the official manual.

The parameter for the blazed grating in my slmsuite is set to vector=([0.01, 0.01]).

My concern is whether this setup still satisfies the linear gray-scale to phase relationship. If I intend to apply a π phase shift in slmsuite under these conditions, could the corresponding gray-scale value deviate from the center value of the scaled gray-scale range?

Best，
yyy

[ichristen]

Hi @yyyaaaaaaa

All these configurations are targeted for 1.065 nm. I have reloaded the default settings and applied the same LUT, except that the voltage limits remain at their defaults (0, 3.6 V) instead of those indicated in the filename (0.95, 3.3 V). This resulted in a lower 0-level light intensity (around 3%) and a higher 1-level light intensity (approximately 75%), which is consistent with the 71-76% range specified in the official manual.

Okay, it sounds like maybe Holoeye mislabeled the LUT file.

The parameter for the blazed grating in my slmsuite is set to vector=([0.01, 0.01]).

That corresponds to larger than a 10 pixel period (period_pix ~ rad_1pix / blaze_rad ~ (1.065 um / 8 um) * (1 pix) / 0.01 mrad ~ 12.5 pix). So you're good there.

My concern is whether this setup still satisfies the linear gray-scale to phase relationship. If I intend to apply a π phase shift in slmsuite under these conditions, could the corresponding gray-scale value deviate from the center value of the scaled gray-scale range?

You can run a testing step of wavefront calibration. If the sine wave plot is skewed (if the x axis is squished on one side or something), this might indicate a bad LUT.

This also depends on how good of a calibration you want. Is 10% deviation from linear acceptable? Or perhaps even 1% deviation from linear is unacceptable? (1% is probably hard to measure via the wavefront calibration test, will require proper pixel_calibrate).

You can make the wavefront calibration test more sensitive by increasing phase_steps.

Best,
Ian

[ichristen]

^ note that the wavefront calibration test will only give information about the LUT if your calibration region is as close as possible to the 0th order. Try to minimize the number of blaze fringes on the superpixel to 1-2.