[AKSA]

Folks,

Recently I decided to examine the so-called 'digital', Class D amplifiers.  There are now chips from TaCT (Danish company recently acquired by TI), UcD (Philips), ICEpower (OEM only, from Bang and Olafsen), Tripath (Tripathi Corporation), and, amongst the Japanese, a Sharp offering.  Most of you have heard of Bruce Candy's Australian Class D amplifier, the Halcro, which is emerging as one of the best amplifiers in the world.  There is a lot of buzz at Audioasylum and DIYAudio with comment ranging from the rhapsodic to the ho-hum;  so it deserved a good look.  I was captivated by this very efficient concept in my youth from a UK entrepreneur called Clive Sinclair, and was interested to see how it had progressed in forty years.

The amps operate at around 300-400 KHz, depending on implementation.  By varying the on and off period of a mosfet output stage operating in push pull - by altering the mark/space ratio -and by using a (second order) LCR filter at the output, an analog waveform can be constituted from an integrated train of very fast on/off pulses, allowing a near perfect waveform with low distortion and very good transient response.

I bought a well known ready-built amp module, connected it to a Nirvana Plus 55W power supply, and began the listening tests.

Strengths:
 
1.  Very good transient/impulse control.  Slam and impact is amazing, and probably a tad better than the 100W AKSA.
2.  Good detail in midrange, with some coloration in the top end.  The sound is arresting, but not overly natural.
3.  Astonishing bass.  Tight, powerful, even a little 'wet', I suspect there's a little H2/H3 in there.  Very appealing for subwoofers!
4.  Good spatially.  I hear a clear soundstage.  In Class AB design (as you probably realize!) soundstage is incredibly difficult to achieve.  This amp has good imaging with width and depth but it's not quite up there with the AKSA.
5.  High efficiency but a surprising amount of heat generated in the output filter, which runs too hot to touch for more than a few seconds.  At idle, I'd say the PWM amps are not much more efficient than low bias Class AB.
6.  Good vocals;  but a lack of warmth, not much engagement.
7.  Outstanding load independence.
 
Weaknesses:
 
8.  There is a rawness to the sound, almost abrasive, which is not quite fatigueing, but certainly not smooth, and lacking refinement.  It's really only apparent on soft chamber music and acoustic vocal, and shows more as an edginess to the voice rather than obvious sibilance.  Immediate impressions are very good, but quiet, low level listening reveals refinement levels not quite to very good Class AB levels.
9.  On some electronic music ('Song for Olabi', by a Danish group called 'Bliss'), there is some sort of digital noise, low frequency.  I don't know, but it's quite unnerving.  It only occurs on some electronic music, NOTHING else.  I will check this more, of course.
10.  Sometimes there is an impression of a lack of 'body', as though a singer is not pushing hard through the diaphragm when holding a note.  This creates a lack of conviction in the performance;  one is not involved.


I had been concerned that PWM amps might now be surpassing good analog designs.  However, after extensive listening I've come to the conclusion that they cannot match good Class AB or A analog designs at present until they increase their resolution;  this probably means they must increase their operating frequency to beyond 1MHz.  I'm informed that only at 2.7MHz operating frequency can they give true 20 bit resolution;  this is probably more resolution than is required, but it indicates a clear evolutionary path.  The other area of deficiency appears to be emotional engagement;  this cannot be easily measured or engineered into these designs.  So for perhaps three or four years to come, we can expect the better analog amps to hold their position!

Ben and I plan to write a white paper on this topic for the website in due course.

Cheers,

Hugh

[andyr]

Hi, Hugh,

Verrry inneresting post ... I had no idea the famed Halcro amps were class D!

Given the things obsessing me at the moment, I wondered whether you were actually doing a "like" comparison.  Are you shoor:

1.  the digital amp module you were listening to, is non-phase inverting?  (I know you've told me the AKSAs do not invert phase.)

2.  the transformers were connected up to the mains in the same orientation as the one in the AKSA 55N+?  (NB: in a stereo amp, the two transformers need to be orientated the same way - preferably the "correct" way!  I found out last weekend that in my (only!) stereo AKSA 55, I had put one channel one way and the other channel the other way!  Doh!!)

I would surmise your #10 could possibly arise from one of the above, if the answer was "no"?

And I presume you installed the digital amp module in an earthed metal case (re. your #8 )?

Regards,

Andy

[tinears02]

Hugh.

Thanks for sharing your impressions.
A couple of things that concern me immediately when I
look at those 'eval boards' or modules supplied by
manufacturers:

1. Emi/emc contro.? One manuf in your list
you mentioned has a high freq square wave coming out
of the output pin (after the output filter) with no input signal.
This is on their ref
demo board. With the desire to reduce parts count for
lower bill of materials cost, this is probably not surprising.
Although the sq wave is out of the audio band, there is
no guarantee that intermodulation phenomenon will not
occur. May go some way to explaining the brash sound
you hear on quiet chamber music.
My guess is that there is a lot more the manuf can do to
mitigate emi. I am sure you already know what emi can do to cd
players and dacs at those relatively low levels (DAKSA). What more with
the high switching levels in these amps.

2. Almost all of these demo boards use a ferrite core inductor
as the output filter. Almost no manuf of ferrite core inductors
show inductance vs current thru them, save one that I know.
They saturate badly with dc current and high peaks.
Perhaps this explains the heat that you mentioned. What this
means is that as the audio signal level increases the output
filter network is a signal-level dependent filter .

My feeling is that class-D or more generically, switching power
amps are great for dynamics. But when it comes to tone, such
as being able to differentiate which Stradivarius model is being
played or the vocal inflections in female voices,
they are not quite there - yet.
The times, they are a changin.

cheers
yh

[andyr]

... as being able to differentiate which Stradivarius model is being played or the vocal inflections in female voices ...

Maaate,

What you really need is an amp which can tell you whether the Strad was made in summer or winter ... or whether the female singer got lucky the night before!  

I reckon an AKSA 25N+ with Ginger's additional mods should come close!  

Regards,

Andy

[jules]

The tight bass sounds like great news for HT applications so that will no doubt make it very popular.

When developed to the level Hugh suggests it can be, will "more information" and "better information" in the mid-range necessarily make a better sound? Is it ever going to be possible to reproduce all the nuances of a human voice and all the complexity of an orchestra through those paper cone things? It looks as though amps have to be tailored to their shortcomings rather than to a possible ideal.

Although we might rate an amp in terms of 8 or 9 out of ten I wouldn't mind betting that what we get compared to an original performance is about 10%. A piece of toughened paper capable of faithfully reproducing the sound of a few feet of tapered brass tube, an organ pipe and a violin simultaneously ... ? Anyone who's lives in a house with someone who plays a musical instruement will know the huge difference between the "live" sound and the reproduced sound. So I'm sceptical, but as with everything in this field there's no way of measuring the really fine details and no way of knowing with certainty.

The end of SET and AKSAs as we know them? I'll believe it when I hear it.

trogloditically

jules

[muralman1]

IMHO the time has arrived.

I have yet to hear an AB or A amp that can get as close to the real female voice as the H2O ICE powered amp. People have left behind the Parasound HC1, Pass Labs X, Plinius, Classe, and Krell for an H2O. The same goes for tube, like Quicksilver, and other class D amps, like Bel Canto.

Knowing some things about the development of the H2O, I can say with some certainty that one can't throw an amp together around the ICE module, and get a decent sound. It takes sound engineering skills thoughtfully applied.

As was mentioned by jules, the paper cone things may not be able to simultaneously handle all four pieces of a string quartet. My 100 feet of feather weight naked aluminum ribbons centered in a powerful magnetic field sure can.

BTW, I am not commercially affiliated with the H2O company. just a devoted groupie.

[andyr]

... My 100 feet of feather weight naked aluminum ribbons centered in a powerful magnetic field sure can ...

Apogees, muralman  

Regards,

Andy

[AKSA]

Thanks Guys,

Some very perceptive, informed comment here.  I'm always blown away at how expert a lot of people are in these forums, and thank you all for your valued input.

I see Class D/PWM v. Class AB/A amps as akin to EFI/VVT v. carbs/pushrods with auto engines.  However, I would remind all those intoxicated with new technologies that the very best of them are usually the ones just declared obsolete....     Objectively, I see good mileage in the old audio technologies yet, and to illustrate my point by analogy I would remind petrol heads everywhere that there are many motorcycle engines still being made with carburettors.  This is done for a reason, not just cost.......

Nevertheless, it is clear to me that the high efficiency and improving speed of Class D will eventually deliver a better product - specifically in terms of measured performance.  But we all know measured performance is a flawed indicator - the best SETs in the world, adulated for their 'emotional communication' (whatever the heck that is!), run at 2-3% THD, so impeccable measured specification may not be entirely believable.  No, there is art as well as engineering in the design future of Class D, and this will ensure that the market will be large, diverse, and seemingly contradictory.

I was struck by two stunning features of the Class D example I tested, notably impulse response and deep bass.  Since these features are hallmarks of good home theatre, I can't see a better application......

Maybe a new Aspen product in this area?  Could this be?    

Thanks for all the sage comments on this topic.  Magical!!  

Cheers,

Hugh

[mac]

Since these features are hallmarks of good home theatre, I can't see a better application.

I wouldn't go as far as proclaiming class-D [only] suitable for HT applications.  I've personally owned lots of amps over the years including tube, hybrid tube and various true SS amps including a one-of-a-kind class-A Stax DA-80.  The class-D amps that I currently have in my non-HT system eclipse most of the amps I've ever heard.  As someone who’s selling conventional competing [kit] amps, it's understandable that your findings fall somewhat short of what my experience has been with the technology.

[andyr]

Hi, mac,

Can U hexplain what the great-looking amp in the piccie is?  The no. of power transformers and filter caps suggest it's a 3-channel amp but, if the PCBs on the left with lots of wires atached are the amp modules ... there only seems to be 2 of them?  

And BTW, have U auditioned an AKSA N+ in your non-HT home system?

Regards,

Andy

[EchiDna]

..... As someone who’s selling conventional competing [kit] amps, it's understandable that your findings fall somewhat short of wh ...

Sorry dude, but as a digital amp owner you are just as likely to be guilty of exactly the same thing suggest Hugh is doing here... we all have an inherent greater like for the object we own when compared to something else, whether that is a car, an amplifier, a house....

To be fair, have you personally heard an AKSA?

not to mention, considered the costing of the AKSA when compared to the *cough* DIY *cough* digital kits on the market?

I've heard some decent digital amps, and for straight out sound I find them too analytical and not involving, less emotional than I hear in the best SS amps... for me, tube power amps go too far the other way in that they lack resolving power and (typically) don't have the grunt I enjoy...

In summary, for my preferences the AKSA is still right up there.... in 5 years time? who knows, my hearing will be worse and my preferences might have changed.

[andyr]

... we all have an inherent greater like for the object we own when compared to something else, whether that is a car, an amplifier, a house....

...

Haha, Echidna,

I agree that we inherently like the car, amp, house .... etc. that we own over others but, funny 'bout that, the analogy seems to break down when we consider ... the wife/girlfriend that someone else "owns"!!  

Regards,

Andy

[bubba966]

I was struck by two stunning features of the Class D example I tested, notably impulse response and deep bass. Since these features are hallmarks of good home theatre, I can't see a better application......

Maybe a new Aspen product in this area? Could this be?

I could use a 7ch amp sometime in the hopefully-not-to-distant future. And last time I tried to calculate what it'd cost to build a 7ch AKSA 100wN it was more than a bit spendy. I'd guess that might not be the same using some type of Class D modules...

[rha61]

Hi

The class D modules Mac and Hugh are speaking about have an input buffer build around ne5532 or opa2134 opamps and noname coupling caps
Already with these buffers , they have a surprising sound
You have just to work a little on this buffer stage for better psu and op amp to get the best amp i 've ever connected on my system
( i have not heard an AKSA amp , but i've owned a lot of high end amps )

alain

[Carlman]

If Aspen decided to implement a digital HT amp, that would make the most sense to me.  However, would that mean the development of a pre/pro of some sort also?  If you want an all-AKSA system it would.

Maybe an HT integrated with digital inputs?  You could use some of the DAKSA findings as well.  It'd be a good way to sell a lot of speakers as well.

How all of this could be transformed into a reasonable DIY kit is beyond me but if anyone can do it, Hugh can.

[mac]

To be fair, have you personally heard an AKSA?

Yes.

[Felipe]

If Aspen decided to implement a digital HT amp, that would make the most sense to me.  However, would that mean the development of a pre/pro of some sort also?  If you want an all-AKSA system it would.

Maybe an HT integrated with digital inputs?  You could use some of the DAKSA findings as well.  It'd be a good way to sell a lot of speakers as well.

Sorry to disagree here, but i found this statement a litle (if not a lot) of reason.  
A HT preamp is the most complicated piece of equipment one could find: it uses a digital input (or more than one) , has lots of DAC's for stereo & the various types of multi-channel and it has a 5.1 or 7.1 channel preamp !!

Aspen amps are DIY and to produce such a product you would have to put together 3x Gk-1 for 6 channel HT , plus a DAKSA for stereo inputs, plus a DAC for multi-channel wich is NOT in Hugh's plans (sorry if i am mistaken here) and join this in one box......too dare maybe $$$  

A 6-channel power amp Class-D is one thing..... a 6-channel GK-1 is another, though perfectly possible and i HAVE thought about it - problem is dont have the dough yet $  .  To add a processor...well....that is a complete different story, and to my point of view, very off-topic.
I see a lot of decent processors out there, DAC's are almost all the same, diference is the analog stage.

[Carlman]

Hugh can do anything he wants and my suggestion was just that... a pie in the sky lofty suggestion... that he can do with as he pleases... I agree it'd be a big project to create an integrated digital HT amp... Maybe it doesn't have to be DIY...

edited to remove misunderstanding...

[Felipe]

Sorry if i sounded ofending. Not my intention.
Maybe used the wrong smiley  

I agree Hugh should do what he pleases, i was just expressing my opinion about it...

[AKSA]

Mac, you are welcome here!  You certainly know how to provoke discussion!  Should we invite Occam to comment??    

Bubba, thanks for your post - yes, could be done, but with all the 'why me' robot assembled seven (or five) channel amps presently in the market, you'd need a good differentiator to make it attractive.  And remembering that they are mostly made in China, it would need to be offered at no more than about $US1200 for a kit - which runs the margins very, very tight, particularly given that my costs aren't much less than the US and there's a big brakish swamp between us......

Carl, ditto your post - good thinking, but there are a few manufacturing issues we need to know about.  First up, a processor needs a manufacturing licence from George Lucas/Ray Dolby (I believe) before you can legally sell 'em.  Not sure of unit cost here, but I understand quite high.  Then, being a digital beast, you'd need to make it along motherboard lines, so four layer SMD is de rigeur.  OUCH!!  How can this be done for DIY;  we are having enough difficulty with the DAKSA, and that's only two layer!  So, it would seem Felipe is right;  this is a minefield.  

The reality is that most of us in the developed (and developing!) world have grown used to cheap Chinese imports, whose quality and range is improving all the time.  DIY kits from the developed world are not price competitive with fully built, store bought products - witness the price of the European kits!  As decades pass, fewer people have the necessary manual skills for DIY anyway, and the consumer ethic fostered by the marketing world propels all of us into habituated 'box opening' with our hifi components.  So the DIY marketing drawcard has to be service and very high sonic quality.  While service is relatively easy with email, it requires advanced technical (and typing!) skills and many daily hours at the keyboard.  I love this game, which is just as well, because I'd make more income as a suburban gardener!

Rha61 and Echidna, thank you both for your valued input.  

Cheers,

Hugh