There are basically 3 types of factors which affect the working of a flip flop:
1. Setup Time: This is defined as minimum amount of time required for which an input should be stable just before the clock transition occurs. Suppose we have a positive edged JK flip-flop and setup time is t= 1ns seconds. If clock pulse with period 5 ns is going from 0 to 1 i.e. first positive edge is coming at time t=1 ns then both inputs J & K should be stable for 1 ns from time t=0ns to t=1ns & t=5ns to t=6 ns & t=10 ns to t= 11 ns i.e. J & K should not change during this period as shown:
2. Hold Time: This is defined as minimum amount of time required for which an input should be stable just after the clock transition occurs. Suppose we have a positive edged JK flip-flop and Hold time is t= 1ns seconds. If clock pulse with period 5 ns is going from 0 to 1 i.e. first positive edge is coming at time t=1 ns then both inputs J & K should be stable for 1 ns from time t=1ns to t=2ns & t=5ns to t=6 ns & t=11 ns to t= 12 ns i.e. J & K should not change during this period as shown:
And both conditions can be represented as follow and inputs should be stable for at least 2 ns:
Well there are some special timing requirements which must be fulfilled by the input signal to get a valid output at the output terminal. If any of the above requirements is not followed and inputs change their value within any of setup time window or hold time window, then output of the flip-flop can not be predicted and flip-flop is said to enter in METASTABLE STATE and output can be either zero or one. This whole process is called METASTABILITY.
3. PROPAGATION TIME:
It is defined as the time after the clock transition, required for a flip-flop to generate output. This is also called CLOCK TO Q delay T clock-to-q.
e.g. Let’s draw the wave forms of inputs and outputs for D FLIP-FLOP which would illustrate the above discussed:
