"Jim Lesurf" wrote in message

In article , Nick Gorham
wrote:
Fleetie wrote:



Mind you, if there's feedback and so on happening round it all,
goodness only knows.

Its not a feedback issue, just inductors at play.

Agreed. If there is a current flowing through an inductor, and the source of
that current is removed, then the voltage across the inductor will tend to
increase until the given level of current flow is maintained across whatever
impedance effectively parallels the inductor, at least briefly. If the
parallel impedance is very high, then the voltage kick will tend to be very
high.

This is the well-known inductive kick-back, which is commonly used as a
basis for automotive ignition systems, for example. Many people do not
realize that while the DC voltage applied to the primary of an automotive
ignition coil may only be 12 volts, when the points or their modern SS
equivalent opens up, there may an inductive kick of several hundred volts.
(Note: other technology may be used.)

Not quite. The o/p transformer is a pair of *coupled* inductors, and
there may be feedback from either/both sides.

Agreed. The amount of mutual coupling between the primary and secondary can
vary.

This may then control
the voltage gain and prevent problems. Depends upon the details of
the design. Removing the speaker load may drop the current levels,
but this does not *guarantee* that the plate swing will rise to a
dangerous level. There is a risk of, this, though. So you must assume
it possible with an 'unknown' design.

If there is an effective negative feedback loop, then its unlikely that
dangerous voltages will be developed in the output transformer, no matter
what the speaker load is, ranging from zero to infinity.

Trouble can begin when the amp lacks loop feedback, which is strangely
enough increasingly popular.

Or, the feedback loop can be present but ineffective, for example if the amp
is operating in a highly nonlinear way.