An improved 
electron multiplier bias network that limits the response of the multiplier when the multiplier is faced with very large input signals, but then permits the multiplier to recover quickly following the large input 
signal. In one aspect, this invention provides an 
electron multiplier, having a 
cathode that emits electrons in response to receiving a particle, wherein the particle is one of a 
charged particle, a 
neutral particle, or a 
photon; an ordered chain of dynodes wherein each 
dynode receives electrons from a preceding 
dynode and emits a larger number of electrons to be received by the next 
dynode in the chain, wherein the first dynode of the ordered chain of dynodes receives electrons emitted by the 
cathode; an 
anode that collects the electrons emitted by the last dynode of the ordered chain of dynodes; a biasing 
system that biases each dynode of the ordered chain of dynodes to a specific potential; a set of charge reservoirs, wherein each charge reservoir of the set of charge reservoirs is connected with one of the dynodes of the ordered chain of dynodes; and an isolating element placed between one of the dynodes and its corresponding charge reservoir, where the isolating element is configured to control the response of the 
electron multiplier when the multiplier receives a large input 
signal, so as to permit the multiplier to enter into and exit from saturation in a controlled and rapid manner.