What isn’t addressed is, what happens “electrically and sonically” with Autoformers when 8 ohm speakers, like 802D3s, dip down to 3 ohms or exceed 20 ohms???With either an autoformer or output transformer the operation is similar, so we don’t have to separate them.
With typical output transformers (common with tube amps) speakers with wild impedance swings (and phase) present a challenging load to the amp and fidelity can suffer.
Can the same be said of autoformers? Or, are these electrically different enough from transformers as to not present the same issues?
Transformers **transform** impedance. To do this, it goes both ways- if a lower impedance as you suggest above is on the 8 ohm tap, a lower impedance is thus also seen by the output devices.
This is exactly the same as if there were no output transformer at all- think about it- if an amp has no output transformer, it too will see a varying load.
How the designer set up the amplifier with the transformer plays a big role. The reason you do this BTW is to reduce distortion- all amps have higher distortion playing lower impedances, so if you can raise the overall impedance seen by the output devices distortion will be reduced. This is why Mac uses autoformers.
Now there is the issue of the varying load, and a lot depends on what the designer of that load expects. You might want to read the article at this link:
http://www.atma-sphere.com/Resources/Paradigms_in_Amplifier_Design.php
ESLs have an impedance curve that typically varies by about 10:1 over their range. Yet it works better for them if the amp can make constant power rather than constant voltage. A lot of box speakers don’t behave this way- for them, constant voltage over the range is more important, this in order to control the normal resonance and resulting impedance peak usually in the bass region near the speaker’s cutoff. To accomplish the latter, enough loop negative feedback applied will cause any amplifier, tube or solid state, output transformer or not, to behave as a voltage source and the resulting output will be constant voltage. This is why this comment:
Large impedance swings and severe phase angles do exactly what you’d expect; they produce dips and humps in response. McIntosh is about the only solid state manufacturer to use output coupling transformers. If you’re driving benign, modestly reactive loads, all is peachy. But you throw some significantly reactive speakers at it, not so much.
- is only partially true- if the amp is acting a voltage source, the above statement is false. Specifically in the case of MacIntosh, the statement is false; Mac led the way in the late 1950s towards developing the idea that speakers be ’voltage driven’ (IOW act like a voltage source) and I don’t think they’ve backed away from that, since most of the industry has followed their lead.