rebiasing from 4 to 8 ohms?

Sean...If what you say is true, the transformer isn't doing its job. (Must be lousy hardware). I agree with you that transformers are generally undesirable, and their elimination is the greatest benefit of solid state amps.

Expensive too. Did you know that Dynaco, aka Dyna Kit, started out as a OEM manufacturer of transformers. They came to realize that between the power transformer and the audio output transformer they accounted for about 80 percent of the cost of hardware in a power amp (this was when tubes were common and cheap). They added a chassis, a few resistors and capacitors, and called it a kit.

Slight differences might be expected because, for example, windings with more turns will have more resistance. Are such second-order effects enough to matter? Can you direct me to the reading matter that will set me straight?

El: You better do some reading on the subject other than theory found in textbooks. Not only do the tubes see different loads at the various taps, the speakers see a different source impedance. When all is said and done, they are all inter-acting and modulating each other.

For a look at a REALLY poorly designed piece of junk that demonstrates this situation to an extreme, take a look at the March Stereophile review of the Antique Sound Labs tube amp. As JA stated in the measurements section, most EE's would call an amp that performed like this "broken", yet the ASL Hurricane's ( another one of their amps ) are on The Absolute Sound's "luv it to death" list. Go figure... Sean
>

Sean...The frequency-sensitive impedance variation seen by the tubes will be the same for 4 ohm as for 8 ohm speakers (assuming the speakers have similar characteristics). The transformer makes different impedance speakers look the same. It does not make them look like resistors.

El: The isolating and impedance coupling benefits of output transformers only work in theory, not reality. Since there are no speakers / drivers made that maintain a consistent impedance and level of reactance across the audible range of operation, the load that the transformer sees and therefore the tubes load into varies with frequency. That's why getting rid of the huge inductor located smack dab between the tubes and the drivers cleans up the sound so much. That is, if the amp is built and designed robust enough to deal with the speakers being used. This is besides the fact that the transformers themselves introduce non-linear distortions into the system. Sean
>

The purpose of the transformer is to make all speaker loads look the same to the tubes. The tubes don't know what the final load is, so bias should not be speaker load sensitive.

I basically agree with Newbee but the stability of the design may have something to do with this. Some "bottleheads" may find this far-fetched, but it will never hurt to check such things just to be sure. Once you've done so, you know what to expect in the future. Sean
>

No. Biasing the amps has nothing to do with which taps you choose to power your speakers.