HowInternal power depends on output capacitance and slew provide din the .lib file

[prateek.agrawal]

Hi,
Internal power is the power consumed by a cell when the input switches. The internal capacitance's of the logic might be charged or discharged causing energy to be dissipated in the form of heat. Also the internal power includes the energy dissipated when a short occurs between the pmos logic and the nmos logic. This is also in the form of heat. Now when we are picking up the information of leakage power from the .lib file why is it that the internal power depends on the output capacitance even though in some cases only the internal caps may be switching..also how does slew effects these powers internally???in other words why is the internal power calculated based on output capacitance and slew??please help asap

 

[tgootee]

Is it not reasonable to expect that the device's power dissipation (and other internal behaviors) will depend on the load it must drive?

 

[prateek.agrawal]

Is it not reasonable to expect that the device's power dissipation (and other internal behaviors) will depend on the load it must drive?

NO its not.If the output load is not switching the output load wont have any effect on the power dissipated due to internal switching of capacitance's..as far as i have learned .libs do not make the internal power of a cell depend upon output load if the output is not switching..it is a function of output slew only..i wanted to be sure thats all..

 

[tgootee]

Sorry for being "high-density". I know almost nothing of your application and the "switching" terminology is a little ambiguous. But if there is output slew (of current?) into a capacitive load, then wouldn't the required total slew time depend on the output capacitance (assuming it needs to reach some required voltage)? Or, alternatively, the current would have to increase to finish slewing in some fixed shorter time. Either way, wouldn't the total internal dissipation over the slew time depend on the value of the load capacitance? i.e. same current slew rate for longer time or higher current slew rate for shorter time but either the time or the slew rate would depend on C. (Slewing of voltage instead of current would involve a similar argument.) Or is the input changing but not the output?

 

[prateek.agrawal]

Yes i can now understand why the slew is having an effect on the power dissipation..when the input switches the rate of change of switching from 10 01 will have an effect of how much current flows through..but if the output capacitance doesnot change say for a nand gate the internal power consumed by the nand gate wont depend upon the output capacitances..