A demonstration of why decoupling capacitors are required to reduce supply rail transients caused by so-called shoot-through current spikes.
A second sheet included to illustrate the additional effects of the same circuits driving a resistive load.
Shoot-through current spikes are caused by push-pull output stages that are controlled by signals that allows both devices to be in a low resistance state for some period of time. The output stage of the bipolar version of the 555 timer is perhaps the most well known example of this behaviour although much of the increase in supply current drain with signal transition frequency of CMOS logic devices is due to the same effect.
In this example, the gate drive signal, GRDV, is initially at 0V so the upper P channel MOSFET, M2, is fully on whilst the lower N channel MOSFET, M1, is fully off.
As the gate voltage ramps up from 0V, M1 starts to turn ON while M2 is still held almost fully ON so there is a low resistance path created through the two devices in series across the supply rail. With a 9V VDD supply as in this example, when GRDV reaches about mid supply voltage, both devices are turned near fully on so the current through the devices reaches a maximum. As GRDV increases further M1 is turned a little harder ON but M2 turns further OFF so the shoot-through current reduces again.
Although the peak shoot-through current does not change significantly, the width of the current spike is reduced by reducing the rise and fall time of GRDV. Reducing VDD and hence the associated voltage swing on GRDV reduces the peak current because it brings the turn ON and turn OFF thresholds of the two MOSFETs closer together.
To run a simulation do:
To view the simulation results of one or more particular circuits, select and set the probe command(s) then - in the right hand panel - set:
text type = spice
Set all other instances of the probe commands to:
text type = comment
More background information here: