Digital Audio Amplification & Future Trends

[gene]

<font color='#000000'>It seems more and more apparant that digital amplifiers have arrived on the high end audio scene as evident by some of the leading hardware manufacturers embracing these new technologies. &nbsp;

The advantage of digital amplification is increase in efficiency, leading to less power consumption and required chassis size.

I have opened this thread to discuss the various flavors of Digital Amplification (IE. PWM Class D, Tripath Class T, etc) as well as their associated implementation options.

Starting out, here is a white paper on Class T amplifiers from Tri-Path. &nbsp;Discuss.

Class T Tri-Path Amplifiers</font>

 

[Guest]

<font color='#000000'>I know that carver uses tripath technology in the ZR amps.

In the interst of cutting costs will does this mean that tripath amplifiers will cost consumers less dollar-wise since a smaller power supply will be required to yield an acceptable amount of output power?</font>

 

[Yamahaluver]

<font color='#0000FF'>The only class D amp I have heard is in my sub as well as my Alpine PWM amp on my SUV.

The only concern here is how they sound, not all of us are SPL power obsessed megalomaniacs so to some fidelity and quality is the main issue. The PWM works fine in the auto enviroment where power is the main requirment whereas in a home enviroment we do have the space to accomodate big amps as long as they give us musical sound.

&nbsp;

</font>

 

[Rob Babcock]

<font color='#000000'>I've been reading as much as I can about digital amps for awhile now. &nbsp;Maybe some of you have seen my thread detailing the ongoing tests that John Curl (yes, that John Curl) and Brian Cheney of VMPS are conducting. &nbsp;Latest bench tests show some shockingly bad measured performance. &nbsp;Brian has been corresponding w/the designer; apparently some of the methodology used for measuring analog amps yeilds screwy results with digital ones. &nbsp;They're working out some changes to see how Carver came up with their specs.

However, measurements aside, BC's listening tests are very encouraging. &nbsp;His initial reports are that the ZR sounds &quot;very good&quot;, and is still burning in. &nbsp;He'll report the results of some exhaustive listening tests in a week or so. &nbsp;Many folks of my aquaintance say the Carvers are the best amps they've ever heard, so we'll see.

This month I'll likely be buying a ZR-1000 or ZR-1600 to see for myself if they are as amazing as I've been told.

I'm stoked about the technology in general. &nbsp;Just think of the day just on the horizon where a digital bitstream will pass unmolested from the disc (including SACD &amp; DVD-A), thru the preamp (where any tone controls, delay, etc will be done transparently in the digital domain), on to a digital amp and to the speakers without any analog conversion whatsoever. &nbsp;Our ears will be the DAC!

Dream further of the day that small, cheap, cool running amps make affordable &amp; compact 200 W x 7 HT receivers commonplace, with the same pure digital signal path.

Some impressive, almost outlandish claims are being made for the modest Sony SACD/reciever units, and they're just the tip of the iceberg. &nbsp;I think we standing on the edge and looking down on a revolution in amp technology.

Or maybe I'm just getting carried away! &nbsp;I'll try to rein in my enthusiasm til I actually get to try one!</font>

 

[gene]

<font color='#000000'>Rob;

I also share your vision, especially since it eliminates all of the snake oil, oops, I mean cable vendors from the equation

</font>

 

[Rob Babcock]

<font color='#000000'>Ha ha! &nbsp;I agree there. &nbsp;I do want to be careful to qualify that I haven't heard them yet, and I don't wanna come off as a drooling, digital fanboy (although I am!

).</font>

 

[gene]

<font color='#000000'>Rob;

I haven't seen the thread regarding the digital amp test results. &nbsp;Please post a link to it here if you have it. &nbsp;Thanks.</font>

 

[PaulF]

<font color='#000000'>Rob, Gene,

I share your enthusiasm for the new generation of digital amplification products. To date unfortunately we have seen most implementations in low cost HTIBs, or car audio and subwoofers as Yamahalover points out. For HTIBs and car audio it simply makes a cost effective and small solution. For subwoofers it provides a compact and efficient method of providing the power necessary to drive these units. For a while Class D stayed in the subwoofer realm for audiophile grade equipment due to the distortion created by the switching lag of the output stage at higher frequencies.

As far as I can tell all first generation digital designs were open loop. We are just starting to see products such as B&amp;O's Icepower and Mueta (as yet not commercialized as I can tell) that take feedback from the output as part of the control loop for the switching stage. It's also ironic that the best performing implentations are the ones that use an analog input signal. It seems to be difficult working a feedback loop around the jitter of a digital signal. Class T tends to modulate the switching frequency with the input signal, but as far as I can tell is still open loop.

Rob, yes the standard way of measuring THD can be problematic in class D designs. There may be other ways now, but the traditional method was to inject a very pure tone into the amp and notch that frequency out with a filter on the output. Anything left on the output must be harmonics created by the non linearity of the amp. The harmonics are &nbsp;referenced to the input level of the tone to determine the perentage distortion. Problem is you can just filter out a single tone on a class D amp, both the music signal and a bunch of unwanted harmonics are created through the switching process. The output low pass filter will extract the audio signal but high frequncy harmonics will still remain, They won't bother us as they're well above the audible range, but to a wide bandwidth test measuring instrument, they're there.

I personally can't wait to see combo class D amps fed by switchmode supplies. Sony did it in their new receiver lineup but at the cost of THD. Jeff Rowland's 300 series performs reasonably well but they forgot to take the toroid out when they put in the switchmode modules. It must have escaped them that users keen on an efficient amp would prefer it not to use 85W when turned off.</font>

 

[Audioholics]

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>

PaulF : <font color='#000000'>It must have escaped them that users keen on an efficient amp would prefer it not to use 85W when turned off.</font>

<font color='#000080'>Ouch.</font>

 

[Yamahaluver]

<font color='#0000FF'>http://www.audiocircle.com/circles/viewtopic.php?t=5123&amp;postdays=0&amp;postorder=asc&amp;start=60

Please check out Brian Putzey's post who is the cheif engineer for class D and Phillips DSL.

He gives some great insights into the implementation of class D and its future.

Yamaha has come out with a real expensive digital amp with MOSFET output stage, would be interesting to hear its performance.</font>

 

[PaulF]

<font color='#000000'>Gene,

I just discovered the Spectron Troubadour uses both a class D amp and switchmode power supply. Perhaps you can get a sample in house and put that new test equipment to work. I would also be interested in seeing how Spectron's fancy remote sense cables work out.

Paul</font>

 

[Dan Banquer]

<font color='#000000'>I have a question for Paul F. You mentioned in a previous post that the Rowland amp that uses the switchmode amp draws a constant 85 watts at idle. I don't understand how a switchmode amp could do that because most of them draw an incredibly small amount with no signal applied.
The other question I have is the mention of the toroidal transformer possibly being the culprit. I again don't understand this because transformers in this application have small amounts of &quot;leakage current&quot; but no where near 85 watts.
Any help you could give me to understand this would be greatly appreciated.
                                  d.b.</font>

 

[PaulF]

<font color='#000000'>Hi Dan,

Yes according to Jeff Rowland's own specs the 302 draws 85W in the off position, please see the specifications at the attached link.

Jeff Rowland 302

I also mentioned that while this design does make use of switchmode power supplies, it did not remove the transformer, for what reason I don't know why. Here is a link to their white paper on the 300 series design.

300 series white paper

They also have a breakout diagram which shows the various components here.

302 Breakout diagram

Finally there is some info on the switchmode modules at this site.

Switchmode power supplies

This is the vendor mentioned in their white paper. If you note from JRDG's specs, the amp uses 1200W max and the amp appears to have two modules. The modules put out between 500-600W, even the 600W is borderline for the power spec rated. My guess is that some parts of the amp are running off an unregulated supply. They have a 1000W module but it requires at least 180V input, and I doubt the toroid is being used to step up the line voltage.

It's not clear to me how they put the toroid to use however I never attributed the 85W consumption in the off mode to the toroid.</font>

 

[Dan Banquer]

<font color='#000000'>Hi Paul;
  Well I am not exactly sure where the 85 watts at standby comes from but I will hazard a guess that there is some form of linear transistor output stage on the amplifier( after the filter of the switching  power amp). This will give a low uniform output impedance that is pretty non reactive.
The toroid is there to double the nominal AC voltage of 120 Volts because the power supplies he is using are more efficent at 220 volts. It will also double as an Isolation Transformer.
   Thanks for the links.
                   d.b.</font>

 

[PaulF]

<font color='#000000'>I hope the increase in switchmode efficiency makes up for the transformer losses. Typically the transformer after the switching stage in the switchmode provides sufficient isolation, after all this is not powering an ECG machine.

If you follow the link to the review, they have a better picture of the insides. The ICEpower class D modules are clearly visible. I still think this amp is way too bulky for something that employs both switchode power and class D amplification.</font>

 

[Rob Babcock]

<font color='#000000'>Solar HiFi has tossed their hat into the ring as well with a new line of digital amps. &nbsp;Owner/designer Kevin Wilden has a new digital stereo amp, a digital integrated amp, and new digital monoblocs.

I also see that Peter of Acoustic Reality has updated his line of digital amps (utilizing B&amp;O's ICEpower modules) and dropped the prices.

Some AC'ers have had the chance to A/B the new Solar amp against the Carver ZR; it was judged to be the better amp, although it costs 3 times what the carver does. &nbsp;Also, those in attendance said the ZR wasn't fully broken in.</font>