Class B amplifier with SINAD of 120, how is that possible?

[terryforsythe]

Never forget that THD+N here (Topping) is achieved with a very low and impractical gain.

They have a high gain setting, which is 19.9dB for XLR and 25.9dB for RCA. In high gain, Sinad is 117.6dB for the XLR input and 113.3dB for RCA.

In contrast, the Purify 1ET9040BA class D amplifier has a single gain setting of 14.4dB. An I/O module is needed if higher gain is desired, but that probably adds noise.

 

[Ken Tajalli]

Just pouring on feedback will not fix inherent crossover distortion from Class B. The key issue is that during that crossover, all of the output devices are switched OFF and the feedback can't have any effect because the output is off (simplified explanation). Also, just increasing feedback will generally cause stability issues.

I don't know what Topping did and would love to know more about it, but there must be a very clever workaround. Speculation hat on, some feedforward? A sliding bias?

We will all kick ourselves, if Topping comes back and explain this "Class B" operation, as being a sort of Class AB operation! Or simply it was a typo.

 

[Timcognito]

Not even close to any kind of circuit designer. Last electronics class was '77 at UC Berkeley for non-majors. Here's how they say they did Class B a Nakamichi in the mid '70's.

 

[wwenze]

We will all kick ourselves, if Topping comes back and explain this "Class B" operation, as being a sort of Class AB operation! Or simply it was a typo.

Never disrespect Occam's Razor.

all of the output devices are switched OFF and the feedback can't have any effect because the output is off

Yea but the feedback would turn the output device back on. The moment the output node is lower than expected, the NFB would increase the base or gate voltage.

 

[DonH56]

I am pretty sure most of these "class B" amplifiers, including Topping's and Nakamichi's, actually have a bit of idle current so are actually class AB but with very low bias current (so class A for only a little while). That and feedback does the rest.

 

[wwenze]

It's hard to define class B in reality. Mathematically it conducts exactly 180º, and reality is not exact. Also real transistors don't change from completely off to completely linear on the dot, but would instead behave half-assedly around that 180º point, so what now.

 

[Ken Tajalli]

Never disrespect Occam's Razor.

Disrespect?
I am quoting him!

 

[Timcognito]

I am pretty sure most of these "class B" amplifiers, including Topping's and Nakamichi's, actually have a bit of idle current so are actually class AB but with very low bias current (so class A for only a little while). That and feedback does the rest.

Nakamichi admits to idle current in their lit above. So is pure class B from zero output a myth?

 

[pma]

I am pretty sure most of these "class B" amplifiers, including Topping's and Nakamichi's, actually have a bit of idle current so are actually class AB but with very low bias current (so class A for only a little while). That and feedback does the rest.

Don, this amplifier (about 1971), is class B or not?

(BTW, R313 was a design nonsense. Omitting it distortion goes 1 - 2 orders down).

 

[DonH56]

Don, this amplifier (about 1971), is class B or not?

View attachment 392841

(BTW, R313 was a design nonsense. Omitting it distortion goes 1 - 2 orders down).

My first thought was "why do I care?", I'm retired! Second thought was that with the bases of T304 and T305 shorted and output capacitively coupled, though is a DC path via the feedback loop, it's hard to see how it would be anything but class B. Last thought is I am too lazy and/or do not care enough to wade through the circuit and figure out the biasing.

OK, after looking at your last note, I am scratching my head about why R313 is in the circuit, regardless of its biasing... No significant loop gain at that point, seems like it just messes up the feedback loop.

If it was an interview question I'd make some assumptions and scribble it out. You could establish a little idle current through the feedback path, I suppose. One bad thing about a career doing high-speed stuff is that the wonderful more complex circuits you find in low-speed audio and op-amps are not possible.

 

[pjug]

In the review thread Amir said this amp idles at 9W. What else would be drawing this much power if the output stage is not biased?

 

[pma]

My first thought was "why do I care?", I'm retired! Second thought was that with the bases of T304 and T305 shorted and output capacitively coupled, though is a DC path via the feedback loop, it's hard to see how it would be anything but class B. Last thought is I am too lazy and/or do not care enough to wade through the circuit and figure out the biasing.

It is a class B, no idle current of the output stage, except for leakage current. I built about 20 pcs of these amplifiers when I was a student and made some pizza lunch profit of them.

 

[pma]

In the review thread Amir said this amp idles at 9W. What else would be drawing this much power if the output stage is not biased?

That is a good question, and another good question is if 9W is with or without the power supply. It might well be the power supply idle power. If it was without the power supply, then the amp is not a class B.

 

[amirm]

In the review thread Amir said this amp idles at 9W. What else would be drawing this much power if the output stage is not biased?

There is a microprocessor running and I *think* a relay that turns on the output. These are extras beyond the amp.

 

[Timcognito]

Maybe post here. Much to be said and learned.

Class B amplifier with SINAD of 120, how is that possible?

I thought that class B based amplifiers advantage were a lot of power in combination with the disadvantage of high distortion. To solve the negative aspect class A design was integrated, creating the classic combination: Class AB amplifier. BUT why bother with this AB combination if it can be...

www.audiosciencereview.com

 

[amirm]

10 KHz is not a valid audio signal as there is no source that can create it. With CD sampling, it would be the sum of just two sine waves instead of infinite for example.

 

[sergeauckland]

Peter Blomley's Class B amplifier from 1971 was a Class B amplifier but with some standing current, but not a Class AB. It worked by using current-driven diodes to switch the audio + and -ve half-cycles to the upper and lower output transistors, thus meeting the Class B requirement that each output device was on for 180 ° only, with no overlap. By definition, Class A has both outputs on at all times, Class AB has some overlap, more than 180°, but less than 360°. Class B has exactly 180°

S.

 

[pjug]

Honestly, where is the downside to this amplifier? You know, other than the power supply.

If it is in fact Class B it has the same inefficiency as Class AB when playing loud. So Class D is better in that regard.

 

[pma]

10 KHz is not a valid audio signal

10kHz square is a very good test signal. It may show hidden stability issues that are not clearly seen from measurements with strictly limited bandwidth.

 

[pma]

If it is in fact Class B it has the same inefficiency as Class AB when playing loud.

Definitely.