what does mean by deadband in switches? and how to select the value of deadband and figure it in plc and fsc?
This is the range upon which an input may change without any action in controller output. Selection of deadband depends on the process under control, it may be fixed or adjustable.
There is a slight delay when switching from one position to the next on a standard break-before-make switch which is due to the contact(s) changing position. This is usually in msec or usec. I would not be overly concerned with the issue of trying to compensate for this as even with a good PLC your program scan time is going to be greater than the time it takes the contacts to move from one position to the other. Though the plc may have the ability to scan at a rate of usec, add a few thousand lines of code and your easily going to have a scan time of 30 - 100 msec.
You should be concerend with issues regarding filtering of your DI's and many good PLCs have the ability to set the input filter time to reduce noise from ac sources - we use a filter time of 1 msec or less.
Deadband is also known as hysteresis and is the amount of a measured variable (pressure, temperature, etc.) between the point where a switch closes and then re-opens. It can be implemented is software if you have an analog input (i.e., open at 1.02 PSI, close at 0.98 PSI) or designed into a switching mechanism. Often the deadband of a switch is fixed and cannot be adjusted.
A typical example is in a wall thermostat in your house. It closes when the temperature is under 69 deg. F and opens at some higher temperature (say 70.3 deg. F). The deadband would then be 1.3 deg. F. Most of the time the deadband is a relatively small part of the range and can be ignored. However, in more demanding applications,
the difference is noticable and can effect your operations.
I wrote an article about this in a bi-directional electric motor that controlled a gas valve on a burner system. The hysteresis kept the valve open until the position signal changed by almost 2%. This allowed the temperature to vary by about + - 7 deg. F. After compensating for this deadband I was able to lower the variance to about + - 1.5 deg. F. The article appeared in Motion Control magazine, September, 1998 (ISA controls the archives of this magazine which is no longer published, you may be able to find a reprint on their site if you are interested).