List of PWM Amplifiers. - Page 4 - AVS Forum
Forum Jump: 
Reply
 
Thread Tools
post #91 of 357 Old 10-31-2005, 04:09 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by millwood View Post

that wouldn't be necessary: you can simply multiple a pwm or pcm or bitstream signals in the digital domain and then feed that signal to the amplifier. For example, multiplying the signal by a factor greater than 1 increases its volume and multiplying the signal by a factor less than 1 decreases its volume.

let me use a simple example. see you are dealing with a 4-bit pcm signal, which takes the form of 0110 (equates to a value of 6) at this particular moment. see you want to double its volume (to 12), you just shift it 1 bit to the left, 1100 (equates to 12). feeding that
signal to a class d amp (either directly or through a dac) will double the output voltage.

or you can shift it to the right by 1 bit to reduce the volume by 50%, 0011 (which equates to 3).

the same principle applies to your typical 16-bit PCM used by your CD, or DTS / DD signal on your DVD.


In other words, you want to do volume control or processing (of which volume control is a form) in the digital domain, and in the pre-amp section, not in the amplifier.

no need to alter rail voltage dynamically for a class d amp - too complicated and without any meaningful gain.

A very good point and it is logically sound.
However, this procedure will require tremendous processing. Imagine recalculating every bit (no pun intended) of information even with the slight change in volume (gain).

Each one of these digi-amps have a unique way of handling volume control. Equibit's approach maybe the best that I have seen so far. It uses PSVC (Power Supply Volume Control) which varies the (fixed) voltage of the PS via a controller built in the chip and is completely separated from the signal path therefore providing the cleanest unhindered volume control system.


...
GoldenEarDrums is offline  
Sponsored Links
Advertisement
 
post #92 of 357 Old 10-31-2005, 04:19 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by Yves Smolders View Post

You don't alter the input signals amplitude to control the volume - you control the rail voltage that is fed to the power transistors. Since the transistors are always on or off, the same feeded signal can be made "harder" or "softer" at the output if you change the voltage on the rails feeding the transistors. In conventional designs, this rail voltage is fixed (in class A, AB, and D) - Class H designs alter the voltage of the rails also, but for a different means: They track the incoming signal to increase efficiency.

Again, these digi-amps varies in designs and voltage control implementation. The PSAudio digi-amp for example uses its Gain-Cell module technology to adjust its volume at the preamp stage.


...
GoldenEarDrums is offline  
post #93 of 357 Old 10-31-2005, 04:39 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by millwood View Post

a class D amp doesn't feed its digitzed output (PWM waveforms) directly into a receiver. It does so through a low-pass filter, which converts the pwm waveform into analog voltage (effectively a DAC).

So the combination of a pwm output stage and the low-pass filter forms the digital amp and outputs an analog signal.

You can feed the output directly to a speaker and basically making it the LPF and the speaker itself will interpret this signal and output it in analog form (soundwave). Of course with the conventional design, noises and transient frequencies will be part of that signal as well. The LPF's purpose is to take out these noises and unwanted signals out and to somehow "smooth" the signal before it reaches the speaker.

The new Yamaha LSI chips are the only one designed to bypass the LPF completely AFAIK.


...
GoldenEarDrums is offline  
post #94 of 357 Old 10-31-2005, 05:13 AM
Advanced Member
 
ManicMiner's Avatar
 
Join Date: May 2004
Posts: 592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Here are two more for your list, both use ICE amps

Midgard Audio

http://www.midgardaudio.no/www/

Bertram

http://www.audiokabel.dk/eng-main.htm

Bertram are the cheapest ICE based amps you can find, but he needs to get some help with his homepage cause it's a mess

Thomas. Rated R for non-stop dumbness, language and some nudity
ManicMiner is offline  
post #95 of 357 Old 10-31-2005, 05:29 AM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
thanks

whats it with europeans and digital amps..

and strangely, the Brits are oblivious (well, except Sonneteer)

Boo!
SiriuslyCold is offline  
post #96 of 357 Old 10-31-2005, 05:55 AM
 
millwood's Avatar
 
Join Date: Dec 2002
Posts: 1,523
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by GoldenEarDrums View Post

However, this procedure will require tremendous processing. Imagine recalculating every bit (no pun intended) of information even with the slight change in volume (gain).

I am not quite sure how much "processing" power it needs. those digital volume controls chips are a dollar a dozen (the upc line from National for example) and each controls 2 or 3 channels too.

Multipliers are some of the most basic elements of digital circuitry. and if you want, you can even use a memory chip for that - I actually have a patent on that.

Quote:
Originally Posted by GoldenEarDrums View Post

It uses PSVC (Power Supply Volume Control) which varies the (fixed) voltage of the PS via a controller built in the chip

it essentially controls the resistor divider on the voltage feedback path in the smps.

Quote:
Originally Posted by GoldenEarDrums View Post

and is completely separated from the signal path therefore providing the cleanest unhindered volume control system....

multiplying in the digital domain would be the cleanest path.
millwood is offline  
post #97 of 357 Old 10-31-2005, 07:43 AM
Advanced Member
 
ManicMiner's Avatar
 
Join Date: May 2004
Posts: 592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Here's one more

http://www.lcaudio.com/index.php?curr=2

Thomas. Rated R for non-stop dumbness, language and some nudity
ManicMiner is offline  
post #98 of 357 Old 10-31-2005, 08:07 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
SC,

Also add Flying Mole. Those guys are ex-Yamaha "Digital" engineers that formed their own company. Great products but looks so generic.

http://www.flyingmoleelectronics.com/


...
GoldenEarDrums is offline  
post #99 of 357 Old 10-31-2005, 08:24 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by millwood View Post

...
multiplying in the digital domain would be the cleanest path.

I agree with you, it is ideal. In fact similar logic in DSP for EQ, effects and etc but the processing on these pales in comparison. To reprocess the whole dynamic sound in realtime either bit-by-bit or even byte-by-byte is a tremendous processing tasks.

Quote:


... those digital volume controls chips are a dollar a dozen (the upc line from National for example) and each controls 2 or 3 channels too.

I'll go check but IFAIK these digital attenuators are for "mechanical" control. Unless I missed the one for audio. Can you give out the link to National? TI Equibits PSVC controller circuit is similar to what you describe but it still just a controlled step-up/step-down circuit to control the PS.

Quote:


Multipliers are some of the most basic elements of digital circuitry. and if you want, you can even use a memory chip for that - I actually have a patent on that.

Well now... I'd like to hear more on watcha got.


...
GoldenEarDrums is offline  
post #100 of 357 Old 10-31-2005, 08:24 AM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
LCAudio and the Moles were there from the beginning

Boo!
SiriuslyCold is offline  
post #101 of 357 Old 10-31-2005, 08:26 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Sorry SC!


...
GoldenEarDrums is offline  
post #102 of 357 Old 10-31-2005, 08:33 AM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
haha no worries

it's amazing how European companies have taken PWM to heart - B&O & Hypex and the people who use their amps dominate the list,

Boo!
SiriuslyCold is offline  
post #103 of 357 Old 10-31-2005, 08:53 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by millwood View Post

Back to the point that timing of a pwm amp being analog. You could conceivably use the pcm signal to control the duration of the pwm on / off state and have the amp being entirely digital.

For example, if you see a signal of 0101 (again, in a 4 bit pcm format), you turn on the output device for 5 cycles (or a multiple of 5 cycles). in this way, all you need is a sqaure wave generator and a down counter that automatically resets itself.

something like this can be even implemented ina fpga device. and can be very fast.

This is how a pure digi-amp work (like Equibit) if I am getting you right. Except it is coupled with a triangle wave instead (to produce a square wave).



...
GoldenEarDrums is offline  
post #104 of 357 Old 10-31-2005, 09:04 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by millwood View Post

those pwm amplifiers are essentially your switching mode power supply. Except that in the case of a power supply, the "input' signal is a reference voltage / current. You can take such a power supply and feed its ref node with an audio signal and it should sign.

As a matter of fact, there are people using sg352x devices and mosfet gate drivers to make high performance pwm amps - IRF's IRAudAmp1 is such an example.

IRAudAmp1 is a self-oscillating design similar to UcD of which, looking at from a different angle is also similar to a Tracking PS amplifier or "Tracking Down Converter" as Bob Carver would name it.


...
GoldenEarDrums is offline  
post #105 of 357 Old 10-31-2005, 10:58 AM
 
millwood's Avatar
 
Join Date: Dec 2002
Posts: 1,523
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
the National chips are LM197x series digital pots.

The IRAudAmp1 (and 2) are those 'analog PMW" amps, to follow the prior discussion, that take an analog input, digitze it and send it into a LPF.

The Carver scheme is a little different in that it alters the rail voltage "discretely", meaning that it has a high and low rail voltage and if the amp senses a high signal, it will switch to the high voltage rail. the amplifier itself is still linear.
millwood is offline  
post #106 of 357 Old 10-31-2005, 12:21 PM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Thanks for the info millwood. I did check out the National LM197x series digital pots and they are in fact built for audio but the only digital part of it is the gate control for the volume. It also takes in an analog audio input and an attenuated analog audio output. So it is not bit-dynamic.

The TI Equibit chips has similar built in circuit for controlling the PSVC.

I do like your idea of this bit-dynamic VC and would like to know more about your patented product.

Regarding Carver's work, that is a pretty good description with the exception that UNLIKE CLASS-H it only has one voltage rail and not two or more. Sorry, just had to VENT that out as I had a lenghty discussion in another thread. Just in case somebody is lurking.


...
GoldenEarDrums is offline  
post #107 of 357 Old 10-31-2005, 07:46 PM
Member
 
stewartstewart's Avatar
 
Join Date: Jun 2005
Posts: 30
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
I know this is about two pages late.....but instead of wondering what the proper terminology is for D-class amps.....why not just call them D-class....that is completely accurate...PWM, Digital, Digi, are all either inaccurate or ubiquitous on some level but not nearly descriptive enough to cover the entire class.....break it down from D-class at the individual amp....D-class, emphatically, isn't digital....but by some it can be viewed as such....


so simplify....

D-class


that's just a complete novice's take on the subject (or lack there of)
stewartstewart is offline  
post #108 of 357 Old 11-01-2005, 01:52 AM
Advanced Member
 
ManicMiner's Avatar
 
Join Date: May 2004
Posts: 592
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Here's yet another one

Embla. ICE based

http://www.emblaaudio.no/

Thomas. Rated R for non-stop dumbness, language and some nudity
ManicMiner is offline  
post #109 of 357 Old 11-01-2005, 02:05 AM
AVS Special Member
 
blazar's Avatar
 
Join Date: Oct 2004
Location: USA
Posts: 2,328
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 79 Post(s)
Liked: 270
with these modules it appears that just about anyone with a soldering gun and the ability to fabricate a chassis can put these together and come up with a product.

It would appear that the main differences will be in chassis, connectors, transformer....

Well I guess for the most part this has been the case with other solid state designs as well.... only god knows what you actually get from boutique manufacturers anyway.

Blazar!
blazar is online now  
post #110 of 357 Old 11-01-2005, 06:17 AM
 
millwood's Avatar
 
Join Date: Dec 2002
Posts: 1,523
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Quote:
Originally Posted by GoldenEarDrums View Post

I do like your idea of this bit-dynamic VC and would like to know more about your patented product.

it is actually very simple. think of a pot as a multiplier, that takes two inputs and outputs one. then, think about a multiplification table: it does multiplification by maping two inputs on a table and outputing the result. Well, a memory chip does that: it takes an input (on the address line), and outputs a result (on the output line).

for example, take a memory chip with 32 bit address line and 16 bit data line. You split the address line into hi 16 bits and low 16 bits. the high 16 bits contain your pcm data, and your low 16 bits contain your volume control levels. and the memory chip is pre-programmed with all the possible combinations. viola, you have a hardware implemented multiplier.

By programming the chip differently, you can pretty much implemented any multiplification scheme you want.
millwood is offline  
post #111 of 357 Old 11-07-2005, 05:42 PM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
TEAC embraces Class-D with their Esoteric line AZ-1 pre-main amplifier


Boo!
SiriuslyCold is offline  
post #112 of 357 Old 11-08-2005, 12:07 AM
 
GoldenEarDrums's Avatar
 
Join Date: Aug 2005
Location: Sin City - Las Vegas
Posts: 751
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Another cool find SC.

I wonder how many will be displayed at CES 2006


...
GoldenEarDrums is offline  
post #113 of 357 Old 11-08-2005, 03:29 AM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
at $4.5K, the amp is pretty expensive (pretty, and expensive), though; and meagre specs. I suppose we'll get more info when they actually start selling.

50W (1kHz,8 Ω load), 80W (1kHz,4 Ω load)
Impedance: 4 Ω - 16 Ω
Frequency characteristic: 10Hz - 80kHz

The attraction must be the iLink with the Universal player paired with it.

Boo!
SiriuslyCold is offline  
post #114 of 357 Old 01-30-2006, 03:49 AM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Cary Audio A 306 power amplifier


Boo!
SiriuslyCold is offline  
post #115 of 357 Old 02-02-2006, 09:45 PM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Say hello to Rotel RB1092


Boo!
SiriuslyCold is offline  
post #116 of 357 Old 02-02-2006, 11:11 PM
Member
 
KB9KXH's Avatar
 
Join Date: Feb 2004
Location: st louis area
Posts: 162
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
pwm stands for pulse width modulation, modulation is not what an amplifier does, in fact, modulation in an amplifier is distortion. an amp can have a pwm power supply but any audio amplifier that reproduces its input with pulse width modulation would have a very distorted signal. usually pwm is used to impose intelligence on a carrier, or to adjust to changes in a load impedance in power supplies.
KB9KXH is offline  
post #117 of 357 Old 02-02-2006, 11:57 PM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
KB9

dang, I thought we got this out of the way on page 1 of this thread

Boo!
SiriuslyCold is offline  
post #118 of 357 Old 02-03-2006, 12:13 PM
AVS Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 20,777
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 399 Post(s)
Liked: 183
Also, we're listening to distorted PWM sources

Noah
noah katz is offline  
post #119 of 357 Old 02-03-2006, 12:33 PM - Thread Starter
AVS Special Member
 
SiriuslyCold's Avatar
 
Join Date: Jul 2003
Location: KL, Malaysia
Posts: 2,745
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
must be the 'good' distortion everyone keeps talking about similar to tube amps

KB9KXH, take a peek: http://sound.westhost.com/articles/pwm.htm

Boo!
SiriuslyCold is offline  
post #120 of 357 Old 02-09-2006, 06:41 PM
AVS Special Member
 
Schadenfreude's Avatar
 
Join Date: Dec 2005
Location: Middle Earth
Posts: 2,469
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 11
Who has done extensive listening of the new Rotel 7 x 100 class D unit?

The stereo (2 x 500) and 500w mono units are coming out now.

Quote:


must be the 'good' distortion everyone keeps talking about similar to tube amps

THE TUBE AMP SOUND IS due to the fact that some of these units are rooled off at 15k.
If the distortion can be pushed to frequencies above audible range then one can roll off higher , if one wishes to design the amps that way.
The reason they use to use class d mostly in subs was because, if you don't adress it, the "funky wave", comes in at around 3,000 cycles, and it is expensive to fix.

Quote:
Maybe someday in the future we will be able to quantify perceived Sound Quality .
(But not today....)

Earl Geddes Ph.D.
Schadenfreude is offline  
Reply Receivers, Amps, and Processors



Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off