Douglas Self has writen about a 'load invariant power amp' in the sense that it is a power amp whose output voltage does not change when the load changes. It is a voltage output, low (ideally zero) output impedance amplifier.
Jan
Jan
This is clearly not a transimpedance amplifier:
With the drains of Q3 and Q3 tied together, such a hard (unity) feedback is needed to avoid the output stuck on a supply rail.
I went through these nested posts (dark side 2015 & good old tech 2013, both closed) to review this odd constant power / load invariant claims.
As per agreed with elvee, the NGTT does not what it claims. I've the impression that it was an invention by accident: the current mirror was connected wrong (the bases of T6 & T7 on the cathodes instead of the anodes), resulting in a 'rounded' transfer function, resembling a square law, and claiming the constant power (square-root) function, and thus 'tube-sound' too!
Attachments
... and...
Why would someone prefer constant power? What's its use?
Constant voltage or constant current I do understand. But if the impedance of a loudspeaker varies by frequency (it is a very complex electrical-mechanical-acoustic system), it demands more or less energy or dumps energy back in the amplifier. And this dumping (output voltage rises, current flowing back into the amplifer) is what directs to a perfect voltage source, which is capable doing so.
In audio, constant power isn't useful or even desired, I agree.
I just wanted to clear up what people were meaning by their phrases, as they seemed at odds with some of the discussions.
Jan
I just wanted to clear up what people were meaning by their phrases, as they seemed at odds with some of the discussions.
Jan