no parasitic elements interfere significantly.

The waveform can be significantly less easy to in-

terpret if the load has not the perfect symmetry of

that case, as showed below. For instance, it is

enough to connect the resistive load to ground, or

the switching phases described.

If the load is connected to ground, the waveform

stays stuck to ground as long as the output stage

is in high impedance; viceversa when the load is

the supply voltage as long as possible.

If an output load made of an inductor and a resis-

tor in series is used, the inductive kick at the be-

ginning of every output transition generates the

equivalent effect of an ”anticipated” switching

when the inductor can discharge; while the

switching looks ”delayed” if the output transition

tends to initiate a charging phase (see figure 7).

With a load almost free from parasitic elements,

the waveforms resemble the ones of the purely

resistive cases.

With a real, more composite load, the effect of the

inductive kick in comparison to the resistive load,

would be more apparent.

With a capacitor and a resistor in parallel as a

load, another type of waveform can be seen (re-

ported in figure 8).

As long as the output stage stays in the transient

high impedance state, the output voltage will fol-

low the classic exponential law of an RC relaxa-

tion.

As soon as the other transistor is switched on and

takes charge, the waveform is quickly

forcibly

brought to its steady state value.

From the above it is possible to see how the

switching times, inherently very fast, of the output

stages, may be difficult to identify in a waveform if

the output load is not accurately taken into con-

sideration.

Figure 9 show typical switching waveform for in-

puts and outputs.

Ω connected to ground.

L6374

7/13