To understand how our microdisplay works, consider the display as a mirror with an electrically switchable quarter-wave plate (formed by the liquid crystal layer). 

If a light source, a mirror, and a polarizing beamsplitter (PBS) are arranged as shown below and a quarter-wave plate (l/4) is inserted between the mirror and beam-splitter, and orientated at an angle of 45 relative to the S-polarized light, then the following sequence of events occurs:

  1. The illuminator produces S- and P-polarized light, which hits the PBS.
  2. The P-polarized light passes through the PBS and leaves the system.
  3. The S-polarized light is reflected by the PBS onto the mirror, this time via the quarter-wave plate, so it is circularly polarized when it reaches the mirror.
  4. The reflected light, is circularly polarized in the opposite direction, and having passed through the quarter-wave plate a second time, is once again linearly polarized, but this time rotated by 90, i.e. P-polarized.
  5. The PBS passes the P-polarized light through to the observer.

Now replace the mirror and quarter-wave plate with our microdisplay, the same sequence of events occurs when the panel is switched to its "ON" state. That is, it behaves just like the mirror and quarter-wave plate. However, when the panel is switched to its "OFF" state, then the S-light incident on the microdisplay does not have its' polarization state changed. It therefore reflects back to the PBS and is reflected back to the light source. The observer then does not see any light coming from the microdisplay, as it appears black.