Hi,
No. Your missing the point entirely regarding reducing the main
amplifiers output in the bass. Classic class-A is an utterly awful
solution for any sort of bass only drive, your way off base.
rgds, sreten.
What a relief. You've made it really clear.
… Not …
SRetten, I've come over the many years of reading this blog, to respect your opinions, as well as many others (e.g. DF96, azazello, Eli Duttman, kevinkr…). However, qualitative absolutist rebuttals hardly do other than self serve. Essentially you're almost (but not quite) saying, “you ignorant slût, you should know better”.
Which I don't doubt is true! But hey, I don't know better. Tell me why!!!
Respectfully Yours, GoatGuy
I think Sreten's concern was all of the power burned in the class A bass amp you "cooked up." 😉 (Sorry, couldn't resist)
As a guy running pretty large class A DHT based SE amps I don't have a leg to stand on, but it concerns me too. 😀
As a guy running pretty large class A DHT based SE amps I don't have a leg to stand on, but it concerns me too. 😀
Sounds great! Unfortunately, if we follow that line of reasoning then it is scurrilous to have top-shelf speakers that only produce 85 dB/W … because they're inglorious bâstards at wasting power. Which of course is hogwash: at the top-end, one aims to produce speakers of esquisite functionality, and within reason, the dB/W be dâmned. One can always produce a more powerful amplifier.
One of the amplifiers I always wanted to build is one where the input attenuator isn't 2-channel, but has a 3rd channel ganged to the shaft (supposing colinear). The third resistance element would control a PCM switched power supply that would limit power to the output finals. You know: if the nominal setting is '5' (about 1% of total power, 10% of voltage swing), then power supply is throttled down from ±50 V to ±10V. The total quiescent power draw of the amplifier would drop (10 ÷ 50)² = ¹/₂₅, or from 300+ quiescent watts to 12 watts (plus whatever the rest of it needs).
Of course, one could also do this more adaptively with a FWB fast rectifier (synthetic, zero voltage drop, "discriminator" type), and a low-Z charge, high-Z drain capacitor. A tracking servo. Microsecond response on the charge, 1 second RC on the drain. Techy types like this. I like the 3rd VC section. Pick your poison.
Having a digital PCM supply in turn is almost absurdly efficient. 90%+ or more, really significantly limiting wasteful heat production.
I think this may well be the best of both worlds. Essentially Class-A emitter-follower output which strongly decouples power supply variation from output signal level, and an efficient DC power source that doesn't waste the watts.
GoatGuy
Hi,
I don't think I'm stating any more than the obvious if
I say a class A unity gain buffer for the low bass only
is a very bad idea in all respects in pure inelegance.
PSRR is not improved by class A, and rail ripple
by definition being class A will be high. Power
dissipation will be pointlessly very high, and
the actual distortion irrelevant given the
application, Class aB would be sensible.
Its still not a great option given its limitations.
rgds, sreten.
I don't think I'm stating any more than the obvious if
I say a class A unity gain buffer for the low bass only
is a very bad idea in all respects in pure inelegance.
PSRR is not improved by class A, and rail ripple
by definition being class A will be high. Power
dissipation will be pointlessly very high, and
the actual distortion irrelevant given the
application, Class aB would be sensible.
Its still not a great option given its limitations.
rgds, sreten.
Last edited:
- Status
- Not open for further replies.