27 replies to this topic

[Roderick]

The ultra-low distortion, 20-Watt, Class A, stereo amplifier kit that has been described over the last few editions of Silicon Chip magazine is due to go on sale from Altronics about the middle of September,  for $599.  I'm seriously considering building one of these brutes.  THD is supposed to be so low that it's very difficult to measure -- something like 0.0006%.

Anyway, I'm off to the UK for a month, so I might order one when I return.  Should be a good challenge to build, and probably the equivalent of amps costing many times more.  Hope I don't get caught up in the security measure at the International Terminal in Sydney on Friday!  

I get to test out my PlaneQuiet NC7 headphones on the flights to the UK.

Rod

[wineds]

20W isn't a hell of a lot unless you have really efficient speakers.  And just because the distortion is low as measured doesn't mean it will sound good.  The Silicon chip stuff often isn't anything to write home about.  Their first priority is usually lowest cost not best sound quality.  I haven't seen the articles or design but there is sometimes room for improvement if you use better quality parts.  I still think you will probably want more power.  Maybe the design can be upscaled?

[mtv]

Here's the Altronics blurb.

http://www.altronics...a=item&id=K5125

Price on the website is $539

Way too high in my opinion for such a low-powered, 2-channel amp, especially one you have to construct yourself.

[Roderick]

wineds, on Sep 4 2007, 07:57 PM, said:

20W isn't a hell of a lot unless you have really efficient speakers.  And just because the distortion is low as measured doesn't mean it will sound good.  The Silicon chip stuff often isn't anything to write home about.  Their first priority is usually lowest cost not best sound quality.  I haven't seen the articles or design but there is sometimes room for improvement if you use better quality parts.  I still think you will probably want more power.  Maybe the design can be upscaled?

For a Class A amplifier, 20 Watts is pretty good -- they generally don't put out a lot of power.  If you have not read the articles how can you possibly commernt on the quality of the components and design?  They deliberately chose some of the best availible.

This is not some one-off weekend design.  It is the last in the series of a several Class A amps that were designed by one of the the best audio engineers in the country.

[Roderick]

If you want an 80 Watt/channel class A amplifier there is always the Jungson JA-88D monster at about $2250.  It weighs 30Kg and has a THD of about 0.05%.  You could proably cook toast on it .

According to the Silicon Chip articles, they originally considered designing a valve amplifier, but decided that it would be too much work to assemble, and too high a price. They also claim their current Class A offering gives far better sound than any valve amplifier, at a far lower cost.  Its audio performance does not degrade with time, as do valve amplifiers.  The secret to the ultra low distortion is a National Semiconductor LM4562 dual op amp used in the low-noise pre amp part of the amplifier (in preference to two Burr-Brown op amps).  The THD of this op amp is something like 0.00008%!

Rod

[wineds]

Quote

QUOTE(wineds @ Sep 4 2007, 07:57 PM) *
20W isn't a hell of a lot unless you have really efficient speakers. And just because the distortion is low as measured doesn't mean it will sound good. The Silicon chip stuff often isn't anything to write home about. Their first priority is usually lowest cost not best sound quality. I haven't seen the articles or design but there is sometimes room for improvement if you use better quality parts. I still think you will probably want more power. Maybe the design can be upscaled?


For a Class A amplifier, 20 Watts is pretty good -- they generally don't put out a lot of power. If you have not read the articles how can you possibly commernt on the quality of the components and design? They deliberately chose some of the best availible.

This is not some one-off weekend design. It is the last in the series of a several Class A amps that were designed by one of the the best audio engineers in the country.

Obviously this is your call.  But I recommend you take a look at diyaudio.com.  If you want class A I would look at building a Krell clone or a Pass Labs Adelph clone.

[Roderick]

wineds, on Sep 5 2007, 08:30 PM, said:

Obviously this is your call.  But I recommend you take a look at diyaudio.com.  If you want class A I would look at building a Krell clone or a Pass Labs Adelph clone.

Unless a kit comes with all components included I'm really not inclined to chase around after a miriad of parts.  I'm not doing this sort of thing for a living you know.   .

[Chicken Man]

These ultra low distortion amplifiers are highly dependent on the earthing arrangements of the modules, so wiring and terminations need to be of the highest order for such low distortion levels to be realized.

Actually ultra low distortion amplifiers have no sound 'quality' at all as any 'quality' is indicative of colourations induced through residual harmonic distortion of the circuit topology.

Tranparency is what becomes most obvious with such amplifiers, and as such the harmonic distortion from the source components and of the speakers then becomes highlighted.

As stated in an ealier post, high efficiency speakers of around 96db/1 metre/watt are needed to get any sonic headroom out of such a low powered amp so you don't drive it into clipping.

Personally, I think Ultra Low Distortion amplifiers are just a novelty in practice as speaker harmonic distortion is so high by comparison.

Each to their own I suppose.

C.M

[Roderick]

They have gone to some trouble with the grounding and layout of theis amp.  Certainly, the case is very roomy, and each panel of the case is individually grounded.

Actually, I agree that THD ceases to be a significant factor once you get below a particular point, and that speaker imperfections will swamp any advantage of such low figures.  Most commercial amps and receivers these days seem to be content with THD figures of about 0.05%.  I doubt that most people would detect any effects from this level of distortion.  With the steady advance of clas D amps the figures are not getting any better, either, but that does not seem to matter one whit.  I think most HT recievers will be class D before too long.  Many of them are right now.

You are also right about the fact that many people actually prefer amplifiers that introduce a bit of colouration.  Just look at how popular valve amplifiers are.  They are a bit like the limestone caves that use coloured lighting to wow the tourisrs!  Don't go for it myself   .

Rod

[miro]

Roderick, on Sep 4 2007, 07:42 PM, said:

The ultra-low distortion, 20-Watt, Class A, stereo amplifier kit that has been described over the last few editions of Silicon Chip magazine is due to go on sale from Altronics about the middle of September,  for $599.  I'm seriously considering building one of these brutes.  THD is supposed to be so low that it's very difficult to measure -- something like 0.0006%.

Anyway, I'm off to the UK for a month, so I might order one when I return.  Should be a good challenge to build, and probably the equivalent of amps costing many times more.  Hope I don't get caught up in the security measure at the International Terminal in Sydney on Friday!  

I get to test out my PlaneQuiet NC7 headphones on the flights to the UK.

Rod

This amp is wonderful. I built one myself just recently. Has a grate sound and I tested it on different speakers with different sensitivity and sounds grate.
I am highly suggesting building this amp.
You should also build your own speakers as I am doing that at the moment. You can find a good speaker kits on this veb site http://www.stonessoundstudio.com.au/.

[miro]

wineds, on Sep 4 2007, 08:57 PM, said:

20W isn't a hell of a lot unless you have really efficient speakers.  And just because the distortion is low as measured doesn't mean it will sound good.  The Silicon chip stuff often isn't anything to write home about.  Their first priority is usually lowest cost not best sound quality.  I haven't seen the articles or design but there is sometimes room for improvement if you use better quality parts.  I still think you will probably want more power.  Maybe the design can be upscaled?

Have you ever listened to a CLASS A stereo amplifier? If you did you would never comment on it. With class A amp 20W is plenty. If you would like to buy one in shop it will cost you any ware from $2000 to $5000..

[wineds]

miro, on Feb 15 2008, 06:27 PM, said:

Have you ever listened to a CLASS A stereo amplifier? If you did you would never comment on it. With class A amp 20W is plenty. If you would like to buy one in shop it will cost you any ware from $2000 to $5000..

What is the sensitivity of the speakers you are running?

[jliang70]

miro, on Feb 15 2008, 06:27 PM, said:

Have you ever listened to a CLASS A stereo amplifier? If you did you would never comment on it. With class A amp 20W is plenty. If you would like to buy one in shop it will cost you any ware from $2000 to $5000..

If you buy from China is much cheaper.  I had a quick look at some of the stuff from Verygoodstore on ebay.    100 watts power amp is available for about $1100.  If I am interested in a class A amp or a tube amp I will definitely seriously consider buying a chinese made tube amp.

[swm-makka]

Hi guys,

I've just built one of these for someone.  It blows driver transistors regularly.  It also oscillates at high frequency (above 160kHz).  I am an experienced fully qualified electronics engineer.  Buyer beware!  It could take you several days to just stabilise the thing.

The design is an almost straight cut-and-paste from a well-known designer so you might do better to design your own or copy other designs seen in other forums.  The design that this kit is modelled on is perhaps intentionally knobbled in the first place!  After all, if you were writing a book on how to design amplifiers, would you give the best design or a 'close-enough' approximation?  After a good amount of work and basically changing upwards of 15 components and mods I have managed to make it behave.  Whether or not it meets the stated performance figures is another matter.  You'd have to trust their word on that one.

The owners of the design did not readily accept criticism!  So there isn't a path forward in going back to the source.

Best of luck!

[Roderick]

Yes, I changed my mind about ordering one of these kits some time ago.  Basically, the thing is too big to fit easily into my entertainment setup, and the ultra low THD is really just a gimmick for anyone who is used to normal, middle-of-the-road HiFi gear.  Most speakers would not do it justice, so the ultra-low THD thing would be wasted anyway -- like casting pearls before swine!  I would not hear any benefit with my aging ears.

Much better to spend your money on good high-efficiency speakers, so that the power output of the amplifier does not become a significant factor.  Instead of the amplifier, I bought and built the Signature Series mains and centre speaker kits from Wagner Elecronics -- the ones with ribbon tweeters..   They have an efficiencyof 92.5dB at 1 metre.

When I upgrade my amplifier, which will be soon I hope, I wont need a hulking great beast with 250 watts per channel!  It should cost less and easily fit in the space meant for it.  Might even be Class D.

Rod

Edited by Roderick, 28 February 2008 - 11:56 AM.

[CConnor]

Interesting swm-makka. Do you have access to an oscilloscope.

Edited by CamCon, 28 February 2008 - 07:30 PM.

[wineds]

CamCon, on Feb 28 2008, 06:57 PM, said:

Interesting swm-makka. Do you have access to an oscilloscope.

I would assume so if he found it oscillates at above 160kHz.

[CConnor]

Well if he has an oscilloscope, he could measure the THD and let us know if the amp meets the claimed specs.

Edited by CamCon, 29 February 2008 - 09:41 AM.

[Roderick]

CamCon, on Feb 29 2008, 10:40 AM, said:

Well if he has an oscilloscope, he could measure the THD and let us know if the amp meets the claimed specs.

You need extra-special gear to measure THD this low -- most commercial oscilliscopes just can't do this job.

[Iforgot]

swm-makka, on Feb 27 2008, 09:09 PM, said:

Hi guys,

I've just built one of these for someone.  It blows driver transistors regularly.  It also oscillates at high frequency (above 160kHz).  I am an experienced fully qualified electronics engineer.  Buyer beware!  It could take you several days to just stabilise the thing.

Best of luck!

An "experienced fully qualified electronics engineer" and you could not do a 'simple paint by numbers' audio amplifier kit????

even thought it's a "design is an almost straight cut-and-paste from a well-known designer". ?????

I'm glade i'm just a dumb tech..  because it worked for me.
And also, even with extra personal mods..

Maybe you should read Douglas Self's book again, and this time try to understand it.

hang your head in shame Mr "experienced fully qualified electronics engineer"

[StolenNomenclature]

Chicken Man, on Sep 6 2007, 12:20 AM, said:

These ultra low distortion amplifiers are highly dependent on the earthing arrangements of the modules, so wiring and terminations need to be of the highest order for such low distortion levels to be realized.

Actually ultra low distortion amplifiers have no sound 'quality' at all as any 'quality' is indicative of colourations induced through residual harmonic distortion of the circuit topology.

Tranparency is what becomes most obvious with such amplifiers, and as such the harmonic distortion from the source components and of the speakers then becomes highlighted.

As stated in an ealier post, high efficiency speakers of around 96db/1 metre/watt are needed to get any sonic headroom out of such a low powered amp so you don't drive it into clipping.

Personally, I think Ultra Low Distortion amplifiers are just a novelty in practice as speaker harmonic distortion is so high by comparison.

Each to their own I suppose.

C.M

Don't make the mistake of thinking that the simplistic distortion figures obtained with single frquency sine waves driving the amps into a resistive load has bares any meaningful relationship to real world compex audio signals being amplified into actual loudspeakers (very complex reactive load). If these figures meant anything then any amplifiers with below 0.01 % distortion figures would all sound the same, but they don't. In fact different amplfier desgins with very low distortion figures typically sound more dissimilar than amplifiers with high distortion.
Judging amplfiers by harmonic distortion figures tells you as much about how they sound as the top speed of a car tells you about how it feels to drive. SIngle frequency sine wave signals used for typical harmonic distortion measurements are trillions of times less complex than a real audio signal. It is about the simplest test you could conduct, and is therefor of no use whatever in determining how the amp will really behave in real world situations.
If you want an amp to listen to single frequency sine waves, by all means take notice of these specs. If you want to listen to music instead, forget the specs.

[StolenNomenclature]

Chicken Man, on Sep 6 2007, 12:20 AM, said:

These ultra low distortion amplifiers are highly dependent on the earthing arrangements of the modules, so wiring and terminations need to be of the highest order for such low distortion levels to be realized.

Actually ultra low distortion amplifiers have no sound 'quality' at all as any 'quality' is indicative of colourations induced through residual harmonic distortion of the circuit topology.

Tranparency is what becomes most obvious with such amplifiers, and as such the harmonic distortion from the source components and of the speakers then becomes highlighted.

As stated in an ealier post, high efficiency speakers of around 96db/1 metre/watt are needed to get any sonic headroom out of such a low powered amp so you don't drive it into clipping.

Personally, I think Ultra Low Distortion amplifiers are just a novelty in practice as speaker harmonic distortion is so high by comparison.

Each to their own I suppose.

C.M

(Refer to my above comment - this is in addition).

The real distortion of an amplifier also includes the frequency response curve, which is not measured by the harmonic distortion measurements. If the amplifier amplifies frequencies of 10Khz +3db more than at 1Khz, this means the amp has doubled the power level of the 10Khz frequency, which in terms of the 10Khz signal is effectively 100% distortion. Any piece of music going thru such an amp will have its 10Khz components sounding twice as loud as they should. These kinds of distortions exist in all amplifiers - none has a completely flat frequency response curve. Most true h-fi amps would be better than +-3db so my example is an exxageration to prove a point, but even +-.5 db is significant distortion of the signal, and way way higher than 0.001 %.

Its a real shame that these simplistic measurements have entered into the public conciousness in such a way that people give assume that an amp which has 0.001% distortion really does amplify music suchy that the signal coming out of such an amp is 99.998% accurate. This is of course total nonense. Even were loudspeakers 100% perfect, such an amp having a 20-20Khz frequency response +-0.5DB would still be adding many percent of distortion, probably in the region of 20% even assuming the frequency response curve were very smooth. And this of course ignores amp - speaker interaction caused by reactive components in the speaker, and other forms of amp distortion such as intermodulation distortion, etc.

We all need to realise that these harmonic distortion measurements are doen because they are simple to do, not because they really are a good indication of amp performance in real situations. But we must understand that no amplifier ever acheves distortion figures below even 5% with real music sources into real speakers. I doubt if the figire is even below 20%, but then I cannot of course know for the reasons I have outlined above. No one does. There are no instruments that can measure the real performance of an amplifier.

[Chicken Man]

StolenNomenclature, on 05 September 2008 - 01:12 PM, said:

Don't make the mistake of thinking that the simplistic distortion figures obtained with single frquency sine waves driving the amps into a resistive load has bares any meaningful relationship to real world compex audio signals being amplified into actual loudspeakers (very complex reactive load). If these figures meant anything then any amplifiers with below 0.01 % distortion figures would all sound the same, but they don't. In fact different amplfier desgins with very low distortion figures typically sound more dissimilar than amplifiers with high distortion.
Judging amplfiers by harmonic distortion figures tells you as much about how they sound as the top speed of a car tells you about how it feels to drive. SIngle frequency sine wave signals used for typical harmonic distortion measurements are trillions of times less complex than a real audio signal. It is about the simplest test you could conduct, and is therefor of no use whatever in determining how the amp will really behave in real world situations.
If you want an amp to listen to single frequency sine waves, by all means take notice of these specs. If you want to listen to music instead, forget the specs.

StolenNomenclature, on 05 September 2008 - 01:40 PM, said:

(Refer to my above comment - this is in addition).

The real distortion of an amplifier also includes the frequency response curve, which is not measured by the harmonic distortion measurements. If the amplifier amplifies frequencies of 10Khz +3db more than at 1Khz, this means the amp has doubled the power level of the 10Khz frequency, which in terms of the 10Khz signal is effectively 100% distortion. Any piece of music going thru such an amp will have its 10Khz components sounding twice as loud as they should. These kinds of distortions exist in all amplifiers - none has a completely flat frequency response curve. Most true h-fi amps would be better than +-3db so my example is an exxageration to prove a point, but even +-.5 db is significant distortion of the signal, and way way higher than 0.001 %.

Its a real shame that these simplistic measurements have entered into the public conciousness in such a way that people give assume that an amp which has 0.001% distortion really does amplify music suchy that the signal coming out of such an amp is 99.998% accurate. This is of course total nonense. Even were loudspeakers 100% perfect, such an amp having a 20-20Khz frequency response +-0.5DB would still be adding many percent of distortion, probably in the region of 20% even assuming the frequency response curve were very smooth. And this of course ignores amp - speaker interaction caused by reactive components in the speaker, and other forms of amp distortion such as intermodulation distortion, etc.

We all need to realise that these harmonic distortion measurements are doen because they are simple to do, not because they really are a good indication of amp performance in real situations. But we must understand that no amplifier ever acheves distortion figures below even 5% with real music sources into real speakers. I doubt if the figire is even below 20%, but then I cannot of course know for the reasons I have outlined above. No one does. There are no instruments that can measure the real performance of an amplifier.

This is a very old thread of which the 2 previous posts had been directed toward me, for whatever reason, over comments I made in my initial post in this thread.  For the life of me I can not make sense of his ramblings.

I came upon it by accident while doing a Google search on very low distortion amplifier design. It probably doesn't deserve a response but I thought I'd highlight it anyway as the subject on low distortion amplifiers is of interest to many HiFi enthusiasts and DIY'ers.

So then, how low should amplifier distortion levels be if tonal transparency is to meet with the expectations of the audio enthusiast ?

C.M

Edited by Chicken Man, 02 November 2012 - 03:16 PM.

[POV]

StolenNomenclature, on 05 September 2008 - 01:40 PM, said:

There are no instruments that can measure the real performance of an amplifier.

Whilst I agree with this statement, I must tell you that it is not in line with popular opinion on this forum.  Indeed suggesting that not everything can be measured is considered heresy by some here....
...

[ajm]

I think that's a slight exaggeration Drew.

And by "slight" I mean "complete"