Setting up the Q-switch

For the purposes of this discussion, we shall assume that the Q-switch is of the longitudinal KD*P variety as this is the most common form of Pockels cell used for Q-switching. We would recommend that the user familiarises themselves with the characteristics of the cell at a visible wavelength before attempting to set up for i.r. sources such as Nd:YAG. This will make it easier to understand what is happening when you adjust the cell in the final laser set-up.

The rotation of the Pockels cell in its mount should be adjusted such that the input laser polarisation axis is aligned with the axis of the connectors. This is the standard alignment for all Leysop longitudinal Pockels cells. The analysing polariser should be set orthogonal to the input polariser such that in the absence of the Pockels cell a clean extinction of the beam occurs. Place the Pockels cell in its position between the polarisers and arrange for the input beam to spread to fill the input aperture of the cell. This is easily achieved using a ground glass diffuser or even a piece of sticky tape a moderate distance from the input window. Now observe the scattered light which is transmitted onto a screen some distance after the analysing polariser.

The image on the screen should look something like that on the left. This is known as the isogyre pattern and is caused by interference of the divergent light at angles corresponding to directions at which an additional wavelength of path difference has been added by the birefringence of the crystal. They are effectively fringes of equal birefringence. When the pattern is centred with a good extinction of the HeNe beam (although some breakthrough of the beam is usually visible), the cell is axially aligned. If the pattern is off centre, the cell must be adjusted in altitude and azimuth until the pattern is centred. Please note that the pattern is not affected by displacements of the beam centre, only the angle of propagation through the cell.

It is helpful in setting up infra red lasers if this procedure can be followed using a low power HeNe laser which has been aligned to the axis of the laser optics. Often a retro-reflection off the laser rod can be used to set up the HeNe. Obviously, in all these discussions it is assumed that the user will observe the usual laser safety precautions to protect themselves and others present.

This process will have aligned the Pockels cell to the optical axis of the laser and orientated the cell to align the polarisation axis to the crystal axis of the Q-switch crystal, the diffuser may now be removed. The next task is to adjust the voltage applied by the Q-switch driver to produce the correct degree of phase retardation to give a 90° polarisation rotation (assuming half-wave and not quarter wave switching is to be employed, but that's to be the subject of another application note). This must usually be done dynamically as the voltage which produces the necessary rotation when applied under dc conditions, is lower than the dynamic voltage requirement by about 20%. In all cases, the peak switching voltage should be incremented gradually whilst monitoring the performance of the laser produced. If the device is set up for "on Q-switching" the laser should be held off by the polarisation selectivity of the cavity components and the increasing of the switching voltage will increase the laser output until the peak output is obtained. For "off Q-switching", the laser is held off by the rotation of the polarisation by the Pockels cell. In this case, the full half wave voltage may not always be required and it is recommended that the d.c. hold off voltage is increased until hold off is achieved at full optical pump power (with a little margin in hand). Any further increases will not reap any benefits and will just shorten the life of the Q-switch crystal (which will break down under long term application of high d.c. voltage).