In the NPD sensor a particle is only detected if it generates a start and corresponding stop pulse. A start pulse is generated if at least one secondary electron is released and if this electron initiates an electron cascade in the start MCP. For the stop pulse it is further necessary that the particle is scattered towards the stop surface and not in the surrounding structure, that at least one secondary electron is released at the stop surface upon particle impact and that this electrons initiate an electron cascade in the stop MCP.
To compare different start surfaces the relative detection efficiency
was defined as 
where depicts the reflection efficiency and the term in parenthesis the approximate probability for the generation of a start pulse if the secondary electron yield of the start surface is equal to . The factor of 0.6 accounts for the active area of the MCP. For higher energies the probability to get a start pulse approaches one since . The probability to generate a stop pulse is not included in the Equation above since it does not depend on the start surface.
March 2001 - Martin Wieser, Physikalisches Institut, University of Berne, Switzerland