|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
75.168.142.52
In Reply to: RE: Class ab and class a posted by itolduso1000timesb4 on March 29, 2022 at 07:31:08
The reason to run class A is to place the output devices in their most linear region, thus reducing distortion.
Its not really to eliminate crossover distortion, should the design be Push-pull. The reason is that as long as the amplifier is properly biased, there simply will be no crossover distortion; if there were, the amp would not be class AB but simply class B.
Transistors in particular (unless your amp uses Static Induction Transistors, also known as VFETs) are not all that linear, so unless you jump through some serious hoops to produce a zero feedback amp (such as Ayre) the amp will have feedback to improve linearity.
As soon as feedback is in the picture you have an entirely different can of worms! If feedback is used in insufficient amounts (less than 30dB-35dB) it will generate audible distortion of its own thru its application. If you look at the schematic of a tube or solid state amp using feedback you can see one reason why- the feedback point is applied to a device (such as one transistor of a differential amplifier) which has linearity issues of its own, meaning that the feedback signal isn't being seen by the amp in an undistorted way. This causes the generation of higher ordered harmonics thru bifurcation in the said device (which might also be a tube).
You may notice at this point that we are well off into the weeds- class A having little to do with it! It is helpful to have the amp make as little distortion as possible prior to the inclusion of the feedback though.
Put another way, this issue of feedback has to be solved if you really want to get down to the nub of it. To do that, the amp has to have two things:
1) a nice high Gain Bandwidth Product, such that the amp can support 35dB or more of feedback. If the amp does not have the GBP, the feedback will decrease as frequency rises, causing distortion to increase, thus contributing to brightness and harshness (since the ear assigns this tonality to the higher ordered harmonics)
2) The amp must have a lack of high frequency poles that cause phase shift; if it does not the feedback will be limited by something called the 'phase margin' of the amp where at some high frequency the phase (caused by those pesky poles I just mentioned) will become so shifted that the feedback is positive rather than negative- and that results in oscillation. This can and does limit how much feedback you can run!!
There are very few traditional solid state designs (meaning: not class D) that are capable of this and there are no tube amps whatsoever capable of this.
By this time you can see that the class of operation is still not a major player... unless the amp uses no feedback at all!
Class D offers a way around this, by allowing you to run so much feedback that the amp oscillates as soon as you turn it on, and then that oscillation is used as the switching frequency. In this way you can run 35dB of feedback and satisfy the requirements I laid out earlier.
Put another way, don't sweat it.
Follow Ups:
Nice explanation, neatly put.I always enjoyed zero-feedback SET amps (which I hope to try again soonish) and am now hugely enjoying my Class D Marantz Reference PM-10 a great deal.
Big J
"... only a very few individuals understand as yet that personal salvation is a contradiction in terms."
Edits: 04/16/22 04/16/22
nt
Dmitri Shostakovich
"The reason to run class A is to place the output devices in their most linear region, thus reducing distortion."
I don't disagree with the rest of what you said but putting all of that aside, a Class A/B amplifier never "places the output devices in their most linear region, thus reducing distortion" so none of the output power produced by a Class A/B amplifier can properly be called Class A.
And that was my only point.
My point was meant to be very narrow. Not a statement meant to predict how an amplifier might sound, etc.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
If my Class A amp didn't run hot as hell I'd want my money back !!
If my Class D amp ran hot as hell I'd want my money back !!
If my Op-Amps didn't have incredible GBP I'd want my money back !!
opamps didn't have all that much GBP, and so they had a 'sound' if you asked too much gain of them. So certain guitar effect pedals have these earlier opamps, and if you want them to sound right, you have to seek out those older semiconductors.
Totally FWIW dept...
That's interesting about guitar pedals. My first IC op-amp experiments as a hobbyist was with the Fairchild uA741 which was probably the most popular of all time.I still have a few books on the topic and some are probably classics in the field but I was just a hobbyist [ham radio and related] so who knows.
IC op-amp Cookbook by Walter G. Jung, 1974.
Active Filter Cookbook by Don Lancaster, 1975.
Op-Amp Handbook 2nd Edition by Fredrick W. Hughes, 1986.
Transducer Interfacing Handbook by Analog Devices, 1981 [about transducers, signal conditioning and IC op-amp instrumentation amplifiers].
Forest Mims III had a series of "Engineer's Notebooks" and other short topic books on various practical projects. I think Don Lancaster had a CMOS Cookbook. I lost all of those during a move.
Edits: 03/31/22
This approach seems like the best of both worlds. The voltage is Class A, while the current is sourced via Class D. The measurements achieved with this approach are outstanding. It looks like a variation of the current dumping approach.
The DAC being integrated into the ADH amp also helps. No interconnects required. I'm surprised no one else has tried to emulate this topology.
"What this country needs is a good 5 watt amplifier!" (Paul Klipsch)
I took a look at the website where they push this idea.
There's a lot of misinformation. I'm guessing its written by the marketing department, by someone not having a grasp of the actual engineering.
The most obvious example is stating that a class D amp is digital and conflating the two. Class D amps are analog and employ switching. They are not digital; if there is a digital amplifier its not class D. Just so we're clear on that.
Its an interesting idea but one question coming to mind right away is since the class D is providing the current, it therefore has to be providing the voltage as well, so the class A portion would be doing nothing! So I have to assume that the explanation is simply incorrect.
I think it's more involved than that. I get the fact that Class D is analog, but when the engineers demonstrated the amp, they disconnected the Class D section, and just use the Class A portion. That only outputs a few watts. The claim is that the Class D provides the current, while the Class A supplies the voltage. It does seem like the Class A does switch the current source.No other amp I'm aware of uses this approach. I can report that it sounds outstanding, and seems to have lots of power.
I'm not sure how it actually works, but it sure sounds better than most any setup I've come across. I agree the write up is likely written by marketing, but I think the approach is unique, to be sure.
"What this country needs is a good 5 watt amplifier!" (Paul Klipsch)
Edits: 03/31/22
There are two amps: a high power but lower accuracy amp and a low power but very hi accuracy amp. The low power amp is driven with a correction signal and it's output is added to the output of the high power amp such that the combined output is as precise as that of the low power amp alone. Or something like that. I've heard it described as a form of feed-forward.
Edits: 03/31/22
Peter Walker must have been born on another planet. I have Quad 57's and a current dumping amp for the quads as well as a few different amps.I remember something like the low power amp is class A driving 2 class C amps, one + the other - , and the waveform put together to then drive the speaker. That's a lousy explanation and I forget how it's done exactly but it sounds quite good. If I also remember correctly, he also invented tertiary windings on output transformers that was later used by McIntosh but received no credit for that.
I once mailed Quad across the pond for help restoring my Uncle's 1957 57's and he personally mailed me guidance and original picture manual of the Quads and how to update them. Back then at 21 Y.O. I didn't recognize the significance of that, what can I say.
edit - I was wrong, it's 2 class B sections not class C
Edits: 04/03/22
Best sound I ever got out of my partially refurbed 57s was with a passive attenuator driving a pair of completely refurbed Quad II tube amps. Never seriously tried them with anything solid state.
NT
"What this country needs is a good 5 watt amplifier!" (Paul Klipsch)
over the last 40 years of playing music thru them with different amps, they can be extremely revealing. Back then I tried a Phase Linear 400 through them and that was bad. The usual import receiver with loads of feedback was bad. I don't know anybody local with Quad tube amps, my Uncle had a Dynaco Stereo 70 and I have a few different including Mark III's.
Quads have euphonic qualities, at least the 57's. They also throw a wide spacious soundstage beyond the left and right and have a good sense of depth. You can't tell where the mid/tweeter panel is, even with my plain speaker cloth and metal screen long gone. Revealing of sources, yes, so much so that deficiencies in recordings are plainly evident. My quest is over except for music where the money should go.
and not an actual class of operation.
Here is the power formula (which is an extraction of Ohm's Law):
1 Watt= 1 Volt times 1 Amp
From this you can see that if you have a voltage, you also have an amperage. The two cannot be separated, else you cause a new branch of physics!
So if you are driving an 8 Ohm speaker with 10 watts (solving for current first):
10 = 8 times the current squared.
Thus the current is about 1.12 Amps. That means the voltage is about 8.93 Volts. Quite literally you can't have one without the other. I don't know what those 'engineers' did but they didn't switch back and forth between just current and just voltage.
Again, I'm not contesting how the amp sounds nor that it works. I am contesting the explanations I've seen, which fly in the face of basic electrical laws.
Understand your point. They are not very clear on how it actually works.
I think Steve O is onto something with the current dumping/feed forward concept. I had read that one of the Devialet engineers at a demonstration initially played a track with only the Class A amp, then connected the Class D stage.
However it actually works, can't argue with the results. Would have thought by now some other manufacturers would have come up with a similar approach. Maybe it's because it's trademarked?
"What this country needs is a good 5 watt amplifier!" (Paul Klipsch)
.
Have Fun and Enjoy the Music
"Still Working the Problem"
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: