Accompanying the range of knowledge, there is a range of designs, from good to terrible. Keep this in mind, too.
No load: As the voltage swings around (presumably, under control of the feedback loop, if any), little current is used, which means little grid voltage is required. Loop gain is high, and the parasitic resonances of the OPT will be accentuated. (A properly designed amp will accommodate this change in loading without affecting frequency response, gain or stability.)
As physics go, the transformer's inductance is the same as it has always been; magnetizing current will dominate the total current draw, but this is far less than nominal load current, at least for (volts / freq) below saturation (which is usually designed to occur around 20-60Hz at full power, lower at less power).
The conditions under which voltage spikes can be produced are in clipping, where the output tubes are overdriven. This is possible in any load condition, but with no load to dampen the spikes, the voltages can become particularly high. The physics behind this phenomenon are, even though the amp may be attempting to control the output voltage (by feedback), this feedback is effectively delayed by the transformer's leakage inductance, which acts as a series inductor between primary and secondary. When a tube turns off while carrying current (magnetizing current, in this case), the primary voltage leads the secondary.
Under short circuit conditions, loop gain is minimized (indeed, zero), so all the gain goes into driving the output tube grids, pushing them into cutoff sooner and pushing current way up. Meanwhile, since voltage isn't saturating, current *stays* high, so the average current draw and power dissipation are much greater than under proper load conditions.
Tim
--
Deep Friar: a very philosophical monk.
Website: http://webpages.charter.net/dawill/tmoranwms
"Abraham" wrote in message
news:ebb7fb9b-0308-4cda-88d0-7c2a558f8e6d@googlegroups.com...
There seems to be a lot of wrong reasoning when it comes to the load on a
tube amp.
Anyone with real knowledge, both theoretical and experimental, mind setting
me straight?
My understanding of it is this:
No load:
Without any load, e.g., infinite impedance on the secondary, the output
transformer acts like a simple inductor(as if there were no secondary
winding).
The inductor, will charge up and eventually act like somewhat of a short.
Hence larger than expected currents will flow through both the primary
winding and tube when the tube conducts. Class A and class B amps will have
a slightly different response but both may or may not be ruined.
Infinite load:
With an infinite load, e.g., the secondary is shorted, the output
transformer acts much differently. The shorted secondary allows a back emf
to counter the driving voltage of the primary somewhat canceling out some of
the current. The effect is to reduce the maximum current in the no load
case.