Sign in with
Sign up | Sign in
Your question

New "digital" amplifiers. What's the technical scoup?

Tags:
Last response: in Home Audio
Share
Anonymous
January 1, 2005 12:40:18 PM

Archived from groups: rec.audio.tech (More info?)

Hi group,

I'm curious as to the technical differences between these new breed of
amplifier circuits and standard. From what I've read, they seem to use
digital pulsing to reproduce AC signal instead of true harmonic
(sinusoidal) AC.

Wessel
Anonymous
January 2, 2005 12:10:07 AM

Archived from groups: rec.audio.tech (More info?)

> I'm curious as to the technical differences between these new breed of
> amplifier circuits and standard. From what I've read, they seem to use
> digital pulsing to reproduce AC signal instead of true harmonic
> (sinusoidal) AC.

Here's a link to a technical description of how these amplifiers work:
http://users.ece.gatech.edu/~mleach/ece4435/f01/ClassD2...

The so-called "digital" or class "D" amplifiers use pulse-width modulation
of a square wave that is then filtered to analog. The speaker gets an analog
signal, just like any other amplifier output. Pulse width modulation is
actually an analog process. A square pulse can have any width: It can
smoothly go from "always off" to "always on". The output filter "integrates
the area under the curve" [1] -- also an analog process.

The advantages of class "D" are very high efficiency (lower power
consumption and less heat) compared to traditional class "A", "AB" or "B"
amplifiers. The have been around in experimental form since the '70s, but
they seem to be gaining in popularity due to the large number of power
amplifiers required for multichannel surround sound. The primary
disadvantage, as far as I can tell, is class "D" switching amplifiers don't
attain the performance of the highest quality traditional designs, but they
might eventually.

This is a digression, but... The class "D" modulated output is actually very
similar to the unfiltered output of a single-bit sigma-delta D/A converter.
I don't know if anyone uses this D/A output signal directly as a raw signal
source for class "D" power amps. As you can see, the distinction between
analog and digital can become blurred, but that's not surprising when you
are converting between digital data and the end product, which must still be
analog sound.

[1] A complete coverage of integration is probably beyond the scope of this
NG, but if you're interested, see "integration" in any calculus text.
Related resources
Can't find your answer ? Ask !
January 2, 2005 3:46:54 PM

Archived from groups: rec.audio.tech (More info?)

"Wessel Dirksen" <wdirksen@gmail.com> wrote in message
news:1104601218.516536.192920@z14g2000cwz.googlegroups.com...
> Hi group,
>
> I'm curious as to the technical differences between these new breed of
> amplifier circuits and standard. From what I've read, they seem to use
> digital pulsing to reproduce AC signal instead of true harmonic
> (sinusoidal) AC.

Sir Clive Sinclair was doing this in the 1960's. See
http://www.psaudio.com/articles/sdat.asp

Paul
Anonymous
January 2, 2005 4:12:29 PM

Archived from groups: rec.audio.tech (More info?)

Paul wrote:

> Sir Clive Sinclair was doing this in the 1960's. See
> http://www.psaudio.com/articles/sdat.asp

Uncle Clive was 'doing' his investors and customers as usual
and never produced a working class D amp.

--
Eiron.
Anonymous
January 2, 2005 9:42:54 PM

Archived from groups: rec.audio.tech (More info?)

Wessel Dirksen wrote:
> Hi group,
>
> I'm curious as to the technical differences between these new breed
of
> amplifier circuits and standard. From what I've read, they seem to
use
> digital pulsing to reproduce AC signal instead of true harmonic
> (sinusoidal) AC.

These amps are more properly called switchmode amplifiers, because they
operate their output stages in two states - either on or off. High
speed switching is used to create waveforms that average out to be
audio signals.

The advantage of low power loss is offset by the practical difficulties
involved with keeping the switching pulses away from the speaker load
where they might destroy tweeters.

Switchmode amplifiers aren't the only practical approach for building
more efficient power amps. Another popular approach is called "Class
G". It varies the supply voltages applied to the output stage so that
it follows the output wave. This minimizes power loss in the output
stage at the cost of a more expensive power supply. One advantage is
that the problems of switchmode related to reduction of switching pulse
amplitude are eliminated.
Anonymous
January 2, 2005 11:12:56 PM

Archived from groups: rec.audio.tech (More info?)

"Karl Uppiano" <karl.uppiano@verizon.net> wrote in message
news:p kEBd.20938$2X6.11419@trnddc07...
>> I'm curious as to the technical differences between these new breed of
>> amplifier circuits and standard. From what I've read, they seem to use
>> digital pulsing to reproduce AC signal instead of true harmonic
>> (sinusoidal) AC.
>
> Here's a link to a technical description of how these amplifiers work:
> http://users.ece.gatech.edu/~mleach/ece4435/f01/ClassD2...
>
> The so-called "digital" or class "D" amplifiers use pulse-width modulation
> of a square wave that is then filtered to analog. The speaker gets an
> analog

Nothing "new" about Class D. Been around for decades.

There is now also Class T, which is a far cry from class D in principal and
sound quality.

http://www.tripath.com/downloads/an1.pdf

geoff
Anonymous
January 2, 2005 11:12:57 PM

Archived from groups: rec.audio.tech (More info?)

"Geoff Wood" <geoff@nospam-paf.co.nz> wrote in message
news:41d79ef8$1@clear.net.nz...
>
> "Karl Uppiano" <karl.uppiano@verizon.net> wrote in message
> news:p kEBd.20938$2X6.11419@trnddc07...
>>> I'm curious as to the technical differences between these new breed of
>>> amplifier circuits and standard. From what I've read, they seem to use
>>> digital pulsing to reproduce AC signal instead of true harmonic
>>> (sinusoidal) AC.
>>
>> Here's a link to a technical description of how these amplifiers work:
>> http://users.ece.gatech.edu/~mleach/ece4435/f01/ClassD2...
>>
>> The so-called "digital" or class "D" amplifiers use pulse-width
>> modulation of a square wave that is then filtered to analog. The speaker
>> gets an analog
>
> Nothing "new" about Class D. Been around for decades.

I first became aware of class "D" amplifiers in the mid '70s, when I saw it
written up in Popular Electronics, I think it was. I toyed with designing
and building one for experimentation numerous times in the intervening 30
years, but alas, it never happened.

> There is now also Class T, which is a far cry from class D in principal
> and sound quality.
>
> http://www.tripath.com/downloads/an1.pdf

At the end of the Marshall Leach article I cited earlier, he mentions a
push-pull version of the class D amp that requires no output filter. But the
output transistors require idling bias, like class AB. What goes around,
comes around!
Anonymous
January 3, 2005 3:58:27 AM

Archived from groups: rec.audio.tech (More info?)

"Karl Uppiano" <karl.uppiano@verizon.net> wrote in message
>
> I first became aware of class "D" amplifiers in the mid '70s, when I saw
> it written up in Popular Electronics, I think it was. I toyed with
> designing and building one for experimentation numerous times in the
> intervening 30 years, but alas, it never happened.

I made one with a 555 ( IIRC ). It actually worked !

geoff
Anonymous
January 3, 2005 11:08:12 AM

Archived from groups: rec.audio.tech (More info?)

Thanks everyone. The Georgia tech doc is exactly what I was looking for.
Anonymous
January 7, 2005 8:37:57 PM

Archived from groups: rec.audio.tech (More info?)

On 1 Jan 2005 09:40:18 -0800, "Wessel Dirksen" <wdirksen@gmail.com>
wrote:

>Hi group,
>
>I'm curious as to the technical differences between these new breed of
>amplifier circuits and standard. From what I've read, they seem to use
>digital pulsing to reproduce AC signal instead of true harmonic
>(sinusoidal) AC.

The *output* is a true sinusoid (given that the input is!), just like
with a DAC. Indeed, the truly 'digital' ones may be thought of as
giant power DACs. IIRC, Wadia even used that as a model name.
--

Stewart Pinkerton | Music is Art - Audio is Engineering
Anonymous
January 7, 2005 8:37:58 PM

Archived from groups: rec.audio.tech (More info?)

Stewart Pinkerton <patent3@dircon.co.uk> writes:

> On 1 Jan 2005 09:40:18 -0800, "Wessel Dirksen" <wdirksen@gmail.com>
> wrote:
>
> >Hi group,
> >
> >I'm curious as to the technical differences between these new breed of
> >amplifier circuits and standard. From what I've read, they seem to use
> >digital pulsing to reproduce AC signal instead of true harmonic
> >(sinusoidal) AC.
>
> The *output* is a true sinusoid (given that the input is!), just like
> with a DAC. Indeed, the truly 'digital' ones may be thought of as
> giant power DACs. IIRC, Wadia even used that as a model name.

Hi Stewart,

I'm assuming you're referring to a 1-bit delta sigma DAC?

Since there are only two possible output states (prior to filtering)
for either a delta sigma or a class D amp, they are in some real sense
both digital. However, there are significant differences in how their
outputs are generated.

A class D amp essentially uses pulse width modulation, and PWM can
be generated using only a forward loop (i.e., no feedback).

A one-bit delta sigma DAC uses pulse *density* modulation and its
output is generated with feedback.
--
Randy Yates
Sony Ericsson Mobile Communications
Research Triangle Park, NC, USA
randy.yates@sonyericsson.com, 919-472-1124
Anonymous
January 8, 2005 10:45:00 AM

Archived from groups: rec.audio.tech (More info?)

On 07 Jan 2005 13:08:37 -0500, Randy Yates
<randy.yates@sonyericsson.com> wrote:

>Stewart Pinkerton <patent3@dircon.co.uk> writes:
>
>> On 1 Jan 2005 09:40:18 -0800, "Wessel Dirksen" <wdirksen@gmail.com>
>> wrote:
>>
>> >Hi group,
>> >
>> >I'm curious as to the technical differences between these new breed of
>> >amplifier circuits and standard. From what I've read, they seem to use
>> >digital pulsing to reproduce AC signal instead of true harmonic
>> >(sinusoidal) AC.
>>
>> The *output* is a true sinusoid (given that the input is!), just like
>> with a DAC. Indeed, the truly 'digital' ones may be thought of as
>> giant power DACs. IIRC, Wadia even used that as a model name.

>I'm assuming you're referring to a 1-bit delta sigma DAC?

Yes, of course.

>Since there are only two possible output states (prior to filtering)
>for either a delta sigma or a class D amp, they are in some real sense
>both digital. However, there are significant differences in how their
>outputs are generated.
>
>A class D amp essentially uses pulse width modulation, and PWM can
>be generated using only a forward loop (i.e., no feedback).
>
>A one-bit delta sigma DAC uses pulse *density* modulation and its
>output is generated with feedback.

Depends. Both PDM (Philips Bitstream) and PWM (Sony) techniques have
been used in audio DACs. As you rightly say, conventional class D amps
are PWM, although of course any techique *could* be used. A proper
'digital' amplifier can certainly be thought of as simply a 1-bit DAC
with more powerful switching devices.

--

Stewart Pinkerton | Music is Art - Audio is Engineering
!