We have been using DMDs from Vialux for a few years now, and I already published few posts about different effects that need to be taken into account when working with such devices (in particular aberrations and diffraction effects). One more trivial, but potentially troublesome, effect is due to vibrations, that come from the FPGA board and transmitted through the rigid flat cables. In this quick post, I show the damaging effect of vibrations and how to easily get rid of them, at least partially.
In our experiment, we measure the complex field at the output of multimode fibers using off-axis holography. We noticed some instabilities of the fringes as shown in Fig. 1.
Figure 1. Fringes recording without isolation.
After investigation, it turned out that the sources of the instabilities were vibrations coming from the FPGA board, that has a fan. Since the board is linked to the DMD chip via a rigid flat cable, those vibrations were directly transmitted to the DMD, resulting in angle fluctuations.
To get rid of these vibrations, a quick and dirty solution turned out to be quite effective; we simply clamped the cable using some foam to absorb the vibrations and isolate the DMD from the board. Fig 2. and Fig 3. shows how it looks like for a DMD working in the horizontal position and at 45 degrees.
Figure 2. Isolated setup for a DMD in the horizontal configuration.
The second setup has two cables, it is a bit more difficult to achieve good contact with both cables, especially at a 45 degrees angle. However, having a piece of foam on one side, as long as the cable firmly presses against the foam, is sufficient to isolate the DMD.
Figure 3. Isolated setup for a DMD in the 45 degrees configuration.
We measure again the fringes and observe the movie shown in Fig. 4. We see a drastic stability improvement.
Figure 4. Fringes recording after isolation.
To quantify the difference between the system with and without the isolation setup, we use off-axis holography to measure the evolution of the phase as a function of the time at one position. The graph in Fig 5. proves that we get rid of fast fluctuations, likely due to the fan, but also more slow variations, probably due to some tension of the rigid cable that rotates slightly the DMD in time.
Figure 5. Evolution of the phase of one point of the complex measured field as a function of time (in seconds) without (red) and with (blue) isolation.