|
|
|
Output protection of amplifiers (and *from* amplifiers if viewed from a loudspeaker standpoint) is of paramount importance wherever power and current management impose to an output stage to work near about upper limits permitted by his power devices and by manufacturing constraints (as may be, for example, limited dissipation surface).
Protection strategies may be "active" (based on dedicated circuit which "sense" the output stage working point and wake-on when some predefined limits are excedeed) or "passive" (based mainly on thermal fuses of carefully chosen value and on "supply starvation" which rely on tailored - and apparently "undersized" - capacitive reservoirs in main equipment PSU: the latter, with it's high ripple, limit istantaneous output power to a mean value that, if exceeded for medium to long time, drive the fuses to blow). This latter type of solution may be employed, provided that output stage use enough rugged devices, only in amplifiers with not more than of 20-25 Watt. Above this limit an active protection scheme is recommended and, for very high power equipments, mandatory.
In the field of "active" protection circuits, these divide in two mainstreams: the first, more old and "classical", which directly limit current and power delivering, usually acting on the output devices driving; the second more modern and sophisticated (but also a bit less sure than previous) which leave untouched "internal" machinery of the amplifier and act cutting off the load (loudpeakers) or, less used, shorting to ground the signal input when predefined limits are reached.
CLASSICAL DIRECT PROTECTION CIRCUITS
Classic current direct limiter circuits, wherever are employed (also in regulated PSU), appear deceiptively simple and straightforward to design but they have at least a couple of subtleties which require careful thought in order to obtain proper operation of the circuits itself.
First of all is worth remembering that they act usually on a NFB amplifier which, due to its internal dynamic, contrast very strongly the action of the limiter which, if not helped by auxiliary means, may prove itself to be practically ineffective in honouring the task for that they are designed and implemented.
Other than this, a second tips has to be highlighted: these kind of limiter really protect output devices shorting their "driveness" but, at same time, they *connect the VAS output directly to the load* and this, is not taken in account, may turn the VAS itself in a "power amplifier" (without the helping of the nominal output devices!) and quickly drive it to a plain thermal blow, which, in turn, bring the entire amplifier to hang to one of supply rails with consequent appearing of *very strong* DC component on output (effectively near the entire rail voltage) and immediate danger for loudspeakers, especially if the amplifier is a dual supply/direct coupled to load type.
In absence of adequate countermeasures for preventing this occurrence, a plain short circuit on outputs is usually fatal for the amplifier, despite of the fact that output devices are "protected" (at best *they only* are really protected, not the totality of the amplifier!).
Once provided a mean of protecting, toghether with output devices, the VAS stage (may be just a couple of well placed resistor or more sophisticated approach), still remain the choice of kind of protection to be established. Max current sensing is enough? Need also an output voltage-current cross sensing circuit? If so, V-I sensing should be single or dual (or more) slope? Protecting circuit should be aware of second breakdown limit of output devices? And should be aware of working temperature of output stages and of the temperature dependence of its maximum limits? All these questions shows how much complex a circuit protector may be or, in hewvy duty/high power amplifier, *must be*.
(to be continued)