Well... I read the orange amps posting about impedance matching. Sounds like a lot of BS to me.
Listen:
The output transformer, or any transformer for that matter, is like a big spring. You put energy into the spring by compressing it, and you let energy out of the spring by releasing it. Some of the energy is lost as heat. It is the same exactly with the transformer: electrical energy is put into the transformer and an electric field is produced, electric energy is removed and the field collaspses giving back the energy applied minus the energy lost as heat. This field collapse BTW, is how the coil in you car generates the voltage it does.
Now lets move to the front suspension of your car. The spring is preloaded by the weight of the car, in an output transformer we can think of this as the DC current that flows thru the primary winding at idle. A field is generated but the energy is never given back because DC dosen't change; hence: a preload. As you drive, the spring is compressed and uncompressed over unevenesses in the road. In an amplifier an AC signal is applied to the primary with the changing field generating a signal in the output winding. Aside: The shock absorber in the car is akin to the damping factor in your amp, they do the same job. (the British call shock absorbers "dampers" and you want the damping factor to be just right to make the suspension work).
Now lets think about load: 4 ohms, 8 ohms, 16 ohms. 1000 pounds, 1500 pounds, 2000 pounds. The effect is to make the springs operate over a different area of their compression range. If the load is well matched to the spring, everything will work as designed, the ride will be smooth. But if you put a 4 cylinder into a car designed for a Olds 455 or visa-versa the ride quality will suffer. The same applies to speaker impedances, if it is designed for 4 ohms and you connect 16 ohms to it, the sound quality will suffer. Within certain design limits tho, you won't break anything. With impedances, if you put a 4 ohm load on a 16 ohm tap, tube life may be shortened by the extra current required to drive the primary to the voltages demanded, but I doubt it. (if you halve the resistance, the current must double to produce the same voltage. V=IR) And if you put 16 ohms on a 4 ohm tap you are actually unloading the transformer, making its job easier. ( and the tubes might last longer, but I doubt it) The sound quality, however, will suffer.
Now we come to the damage part. The reason the coil in you car produces the 10s of thousands of volts it does is because the field is in freefall collapse when the points open. This is the same effect as disconnecting the load on a tube amp with a signal present. The secondary field goes into freefall collapse, the field lines cut across the primary and since there are so many more primary windings than secondary, thousands of volts are produced on the primary, zorching your tubes. Just like taking a big crowbar to your front suspension and levering out the spring; it will take a big chunk out of your head as it passes into the next county.
Finally: Look at the graph entitled Operation Characteristics here:
http://frank.pocnet.net/sheets/093/5/50C5.pdf
Plate current goes down as load goes up as you would expect. Power output peaks and drops off slightly with increasing load. And distortion drops to a minimun when load is well matched to output device. The range of resistance used here is 500 ohms to 5000 ohms; a tenfold difference. In our discussion of speaker impedance we are going from 4 ohms to 16 ohms, a fourfold increase, well withing the design limits.
Note also there is no disastrous region of the graph.
To conclude: Impedance mismatching is an inconvienience for your amp but accidental load disconnection is a disaster.
Next week: The Oscillator or Taking the Shocks Off Your Car.
2 weeks: Frequency Response or Driving Faster Over Washboard Can Make it Smoother.
Don