Cadence Power Guide
Cadence Power Guide
To measure the static power dissipation, we simply apply a static (DC) input signal so that no
switching occurs. For digital circuits, this amounts to a high (VDD) or low (ground) at the input,
which typically turns one side of the circuit off and eliminates any static short circuit current through
the transistors. For analog circuits, the input should be set to the appropriate DC operating point of the
circuit, typically somewhere between ground and VDD. This is one of the reasons that analog circuits
consume more power; in their static state many transistors are turned on and consume static power.
Example: For a CMOS inverter with pMOS 1.8u/0.6u and nMOS 1.5u/0.6u and a 5pF load
capacitance, we can use the Simple Method described below to measure the static power dissipation by
applying a high (3V) or low (0V) input signal.
if Vin = 3V Æ PVDD = 9.05E-12 W
if Vin = 0V Æ PVDD = 9.04E-12 W.
Notice that, as you should expect, the power consumption is nearly the same for both cases. The
difference is due the slightly different leakage in pMOS and nMOS transistors when they are biased
with different voltages. We could expect these results to vary more if one of the transistors were much
larger than the other.
To measure total power dissipation, we have to apply an input signal that varies with time, causing the
output node to charge/discharge. For digital circuits this simply requires applying a pulse input signal.
Example: For a CMOS inverter with pMOS 1.8u/0.6u and nMOS 1.5u/0.6u and a 5pF load
capacitance, we can use the Simple Method described below to measure the static power dissipation by
applying a pulse input that changes from VDD to ground.
if Vin is a pulse (pulse width = 5µs, period = 10µs). Æ PVDD = 8.64E-9 W.
For analog circuits, this approach is complicated by the fact that the inputs to analog circuits do not
typically switch from high to low like digital circuits. To use this method on analog circuits, one must
first determine the typical input signal swing and then apply the appropriate input signal, which in
many cases would be an AC signal (often with a DC bias). The Alternative Method described below
may be more useful for measuring the total power dissipation in analog circuits. However, it is
important to note that power dissipation in most analog circuits is dominated by the static power, so
applying the appropriate DC bias to all inputs and measuring static power may be sufficient in many
cases.
To measure dynamic power dissipation, we can subtract the static power from the total power to
estimate the contribution of dynamic sources.
Alternative Method
There may be a case where the above method does not provide the information you need. The steps
below illustrate how to use the results calculator tool to collect additional data.