Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
Most of the times, it is because of the timing issues, say, you are drving a net
with a big distributed RC, you may need some cascaded buffers. Some times,
it is because of logic reasons, say, you are implementing an AND logic, you may
need a NAND and an inverter.
usually we put a buffer between to circuits so that one will not act as a load of the other.so if you change something in one circuit it won't affect the "working point" (dc) of the other.
second reason is to make the circuit work faster.when you change let's say from '0' to '1' there are parasitic capcitance that need to be charged, so if you use a buffer your load time,and discharge time is shorter because the buffer has very small parasitic capacitance.
third reason is to improve the throughput rate by creating a "pipeline" -for example if you have 3 gates one after the other, and for each of them it takes 10ms for the data to pass you will have to wait 30ms to get the result (if you want the data to be stable you will not change the input until you get a stable output),but if you put 3 buffers and clock them, you can get a result every 10ms, so your throughput is now a lot bigger and the frequency you can work in is higher (1/10ms instead of 1/30ms).
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
