What makes a high end audio amp "better" then a low end unit?

[Electro Fan]

Next we get to your subject:  the amplifier.  Fwiw, I'm with you.  I think the current consensus in this forum is a bit dogmatic when we say that everything that needs to be solved with amplifiers has been solved.

Let's put it another way: Do you think amplifiers are going to get much better in the future?

That's a good question.

There are reasons to think amplifiers will get better, possibly much better in some respects, the further we look out into the future.  Certainly, one would think there aren't a lot reasons that amplifiers should get worse.  In general, we not only have advancing technologies (including digital and ever increasing software defined and software controlled capabilities) which could conceivably improve amplifier performance and possibly lower prices for a given amount of performance.  Further, also due to digital technology, we have the Internet which probably makes 80% or more of previously known information about anything of value available to anyone with a web browser.  So one would think that the various lessons of amplifier design knowledge accumulated throughout history aren’t going to be lost and if anything will become more widely available to anyone working on advancing amplifier performance.

There are also some reasons to think that amplifiers won't get much better.  In fact, if someone is of the school that says already today amplifier performance exceeds the capabilities of human hearing, then there would be little reason to expect improvements in amplifiers, much less reason to expect amplifiers to get “much better”.

This question regarding future developments however is somewhat different than the question of “can anyone hear any differences between amplifiers?”  I tend to think the answer is yes but I realize this is probably a minority position on this forum and possibly elsewhere.  Having said that, I will be first to say that most differences in amplifier performance are subtle and in general tend to have a lesser impact on overall system sound than speakers.  I’ve listened to enough integrated amps, power amps (solid state and tube), and receivers over decades to be pretty confident that not all amplifiers sound the same.  Certainly, there can be differences depending on output power and depending on the how the amplifier works with the particular speakers being driven.  But even with comparably powered amps and the same set of speakers two amplifiers can sound perceptibly different.  Just my opinion, of course.

One of the reasons I’m not so bullish on amplifiers improving much is that my experience indicates that hifi in many respects (overall, not just amplifiers) peaked around the 1970s.  So when I say that I think I can hear differences between amplifiers it’s in part because some of my favorite amplifiers were made in the 1970s and I’m not convinced that many or even most of the amplifiers made since then do sound better, or even as good – although many are much more expensive today (even after accounting for inflation).

My theory is that sometimes there is confluence of events that focuses society, business, and sometimes the military on a particular area of innovation and we get eras with relatively high performance, and in some cases peak performance that endures for a period of time.  I’m pretty sure the Romans with their aqueducts and columns would be impressed with how far we have advanced cement technology, although while they would marvel at our interstate clover leafs I’m not sure they would think our clover leaf roads would be prettier than their columns.  The point is that their cement technology (and other technologies) represented a peak performance that endured for many years (several hundred through the Dark Ages, and beyond in some cases).

Likewise during the Renaissance history showed some innovation and peak performance in fields such as the arts and political thinking as well as some good improvements in physics.

While hifi probably will get at most a page or a few paragraphs in history books compared to the chapters and books earned by the Romans and the Renaissance, I think the 1970s were a peak performance era and perhaps the “golden era” of hifi.  Of course the 70s were built on developments in the 40s, 50s, and 60s.  The 40s brought the end of WWII and freed up time and funds for new endeavors including early consumer electronics, and especially hifi (in addition to TV, automobiles, and home construction).  The 50s brought the development of stereo vs. mono, and thereby brought a big improvement in "high fidelity" made possible by the spatial imaging attributes as described in our “imaging” discussions, not to mention stereo gave the industry a reason and a way to sell twice as many speakers.

The 50s and especially the 60s brought huge advances in transistor design and commercialization while continuing to make many markets for tube electronics.  By the time we reached the 70s we had large numbers of companies of all sizes from the U.S. and Japan competing for hifi market share.  Additionally, by the 1970s the 1960s rock and roll music had been written and distributed widely on records giving an abundance of music to drive hifi consumption and usage.  In parallel, the development of the electric guitar and guitar amplifiers drove incremental demand for similar audio technologies including a demand for ever bigger amps and speakers. 

By the 1970s a lot of intersecting forces created a crescendo for hifi.  In my opinion the competition for audio amplifiers hit a peak in the 1970s as tubes were still popular and transistor designs had not yet become highly evolved.  The concept of negative feedback was just starting to make it’s way into the consciousness of hifi enthusiasts and high powered amps (from Phase Linear and others) were just emerging.  At the same time tubes were perhaps at their best, least expensive, and most abundant.  One of my favorite power tubes, the GE6550 was manufactured by General Electric in a version called the JAN 6550.   JAN stood for Joint Army Navy.  The demand for tubes was still strong in the military and history has shown that military investments often lead to or enhance commercial products and markets.

Of all these reasons for the 70s becoming a peak in hfii I would not under estimate the impact of rock and roll.  The 60s gave us the Beatles, the Rolling Stones, Led Zeppelin, Jimi Hendrix, and a very large number of other bands that remain classics today.  The demand to feed “hardware” with “software” was significant.

So, what happened to hifi starting in the 80s?  Computers and digital music among other things.  Computers peeled off increasing amounts of people’s time; more time spent with computers and less time spent with hifi.  Digital nearly killed vinyl but it took awhile to realize that just because music could be digitized it might not necessarily sound the same as or even as good as analog.  Much but perhaps not all of the gap has been closed in this area.  But mostly, overall, the addressable market for hifi manufacturers became smaller and smaller as new markets took mindshare from former hifi enthusiasts.  And we shouldn’t forget to mention that many of us who grew up during this timeframe moved on to careers and families.  If some things drove hifi adoption upward, the WAF wasn’t one of them.

All of that is to say – somewhat to your point and the point of others here – amplifier technology has not been on a huge slope of upward improvement since the 1970s, and if you look at price performance it’s possibly declined.  Having said all of that, I’m still not personally convinced that amplifiers contribute nothing to overall hifi system performance.  I believe that while terms such as definition, air, transparency, imaging, and more while exceedingly difficult to quantify and perhaps impossible to measure are discernible attributes and it’s my experience (empirical evidence?) and my sense of history that indicate some of the best amplifiers of the 1970s if A-B’d side by side with many of today’s amplifiers might in fact produce surprisingly good and in some cases superior results, even with DBTs.  I can’t prove the differences would be discernible and they probably wouldn’t be discernible for many people.  So - I can’t prove these things, but heck some people were pretty sure the Sun wasn’t the center of the Universe in 1500 and even though they couldn’t prove it, their thinking turned out to have some merit. 

Just my impressions.  YMMV, EF 

[David Hess]

So if I want to make a high end amp with very low distortion, how important are specs of the components to design specs?

With some exceptions, the component tolerances and variations are not important.

A good example of an exception is aluminum electrolytic coupling capacitors if used; a bad one can increase distortion by orders of magnitude over a good one and good ones are expensive which makes direct coupled amplifier designs popular.

Could I take a low end amp with a decent power supply; change out the parts that are very specific to the value they are supposed to be (take all parts that are say 10% tolerance and hand select them to 1% or less) and see an improvement?

Any improvement would be marginal.  Good designs are tolerant of component variation.

I guess my question is what makes the high end amps different then the crappy ones? Big filter caps? Big transformers? Well matched transistors (I'm guessing most amps are push/pull designs)?

Larger capacitors and transformers are expensive.  Good output transistors (ring emitter, perforated emitter, MOSFETs) are more expensive than poor ones.  Using more output transistors in parallel to reduce Ft and hfe droop is more expensive.  Using matched transistors is more expensive.

Design is critically important.  There is lots of extra complexity like cascodes and active loads which can be used to improve distortion performance and other things.  Cheap amplifiers leave out protection circuits which under normal operating conditions do essentially nothing but you will sure notice if they are needed and not present.  Or cheap amplifier include protection circuits which activate under normal operating conditions distorting the output when they should not because the design was not robust enough to start with.  Layout is critical to prevent ground loops and even magnetic coupling between circuits.

Let's put it another way: Do you think amplifiers are going to get much better in the future?

Better for whom?

They are not going to get better as far as performance compared to the best amplifiers made except in efficiency.  And they do not need to because as other have pointed out, amplifier performance long ago reached the point of diminishing returns as a contributor to system performance.

I would settle for an amplifier which does not contribute significantly to distortion of the system as a whole, operates correctly under adverse conditions, is reliable, and is user friendly.  That leaves out a majority of consumer level designs most of which cannot be evaluated anyway.

[BrianHG]

9. What are the best speakers you have heard that make the best compromise between bass response and stereo imagery?

Headphones.

My Clements RT7 tower speakers outperforms GradoLabs SR1 headphones with stereo imagery even down to their astonishing flat 25Hz response.  In fact, with properly down-converted to stereo DTS bluray movies, I've had sub-sonic effects start out in front of you, move right through you, then felt behind you as imaged by the location of the effects in the movie.  Try doing that with a mono sub-woffer.

Are you able to explain how we avoid constructive/destructive interference effects that occur when you have two (or more) speakers, vs when you have a single source (like an real musical instrument)?

An A6 (1760 HZ) is about 20cm long, so with a sound on the center all it takes is a a speaker 10cm closer than the other (about half a head) to have destructive interference on the fundamental, leaving you only with reverb from the room...

     Yes, this is how your ears judge position of the original mono source.  Having your speakers in an awkward position does alter the imaged location of the instruments depending on their source stereo recording as well, however, I have my speakers facing forward at one end of my room and I don't seem to have any issues of where I sit, though, if I am in a different room, I don't hear any stereo imaging.
     Also, in many loudspeaker designs, the woofer crossover, being an LC filter usually, doesn't throw thing out of wack at different signal levels with it's physical interplay of it's case being physical cavity tuned object.  My Clements RT7 got around this by using a simple RC filter instead, though, only being less than 50% efficient at transferring the amp's voltage to the woofer being a great loss, the 0 error delay which otherwise would have been there with a series inductor in place of the resistor was a trade-off to get a full headphone like imagery at all frequency ranges which would otherwise be lost, or difficult to engineer away elsewhere in the mechanics.  (I did not know this until after purchasing the speakers, a fluke damaged resistor in the crossover which led to repair showed me this...)  As for the woofer being a small 7 inch driver, with a super strong solid magnet, and a case with a huge tunnel tube with VLF whistle type of structure gives a flat 25Hz, or -3db at 20Hz response which will knock you out of your chair even with the loss of power.  Too bad the speakers are no longer made since full range single woofer driver speakers today usually skip out on the mechanics needed for full range in trade for a dedicated sup-woofer which usually is mono and doesn't do good quality recordings any justice unless it is a really expensive tuned setup.

[BrianHG]

Hi,

I normally avoid posting on audio/HIFI/audiophool threads.

I am fairly literate in electronics, I do appreciate music and I have a wide range of musical taste. Yet, I consider myself a normal person, a bit on the skeptical and realist side.

While I do agree that there is a lot of audiophooling in the market, from special cables to power filters, etc., I challenge anyone to actually spend a few hours listining to a propper high end audiophile class setup.

I have had the pleasure to do so and actually have a friend who designs and manufactures such audiophile amplifiers, mostly valve based.

Never had I listened to music like that. The sound can be described as totally transparent, like when you look through a fogged glass (bad audio system) and then suddenly this glass is replaced by a crystal clear and clean glass. Also, it is amazing to close your eyes and feel the musicians located around you. It's like they are playing right in front of you, you can accustically locate them, feel them.

All this is a totally subjective feeling and it definitly does not happen with:

- cheap/bad quality setup (amplifier, speakers and source)
- bad arranged speakers
- room with bad accoustics
- source with bad recording/mastering
- ...

Regarding "cheap setup", mind you, the best cheap amplifiers can be bought on eBay: look for premium amplifiers from the early 80ies! This is what I have now...

I dare to say, that since the 70/80ies, people have lost the drive to hear HIFI quality music - they prefer to listen to their favorite songs through convinient means (phone, mp3, internet streaming), at the cost of sound quality. It's an option.

So, before engineers start to claim EVERYTHING as audiophooling, please do ask some friend or store to give you a propper listening session. Take your favorite CD's with you and just listen to them. Be open minded to listen to their suggested CD's, which have been selected based on the quality they were produced. One great CD for this is the Dark Side of The Moon by Pink Floyd. Another popular CD to evaluate a HIFI setup is "Arne Domnerus Group - Jazz at the Pawnshop".

I am pretty sure, you cannot just scientifically measure the quality of an amplifier, when it comes to reproducing music.

For example, I had a comparitive test, where the same amplifier was driving three different pairs of speaker. All of them premium audiophile ones. And I definitly like the sound of one particular set of speakers more than the other two. But the host prefered a different speaker and he was far more literate than I am. Conclusion: in the end, it is the individual taste.

After such audiophile listening session, I went back to my regular stereo set and noticed how bad it actually sounds...
After some months, I don't notice that anymore, as I am used to it again.

Chers,
Vitor

This is a difficult point to illustrate.  Quality and transparent definition as well as other factors means different things to different people, so, I will share a little story from my past (20 years ago).

A friend in North Carolina, (I'm from Montreal, Quebec), was having a big birthday and he heard my audio equipment which he told his wife about.  She wanted to get him a small high end sound system for his office/man-room and since I was in town, we visited the only closest local HiFi audiophile grade store.  In the budget of 1500$ for a good CD player, amp, and speakers & cable, nothing the store had sounded good, even vacuum tube amps and speakers sounded right to me and I recommended not to buy anything from there.  Just because you have an 'Audiophile' store, if any still exist in your area, doesn't mean they necessarily have anything good to hear.  The store owner may just be just as fooled by Audiophile Audiophoolery to have chosen the cherries from the overpriced phony stuff, or, the store is in the business of profiting on said overpriced stuff.

Anyways, we ended ordering a setup from Montreal from a friend's store and got a good entry level Jolida vacuum tube amp setup which really enjoyable and pleasing quality sound which outperformed what we heard with the way overpriced setup in North Carolina.

From my own personal experience, I got close to this quality sound with my own Class B amp BJT amp design back in the late 80's.  But, to do so, I had it biased so heavily that it's idle power consumption was over 150 watts and it's heat sinks make a nice space heater with I drove the full 100wattsx2 out.  It is possible to make a good BJT of MOSFet amp which can deliver most of the transparency of a vacuum tube amp except for 2 aspects unless you engineer them in.

1.  Handling of signal levels outside the power rails of your amp.  A tube amp softly distorts as you approach the rails instead of a harsh clip and perhaps a nasty inversion spike.
2.  The air feeling of some vacuum tube designs.  This where minuscule parts of your audio source get a little extra signal gain even in the presence of the major large signal tones when driving a dynamic load like a speaker.  An effect some tube designs are purposely designed to enhance.

#2 can be an additional + as it may help compensate for the shortcomings of a physical speaker trying to generate all the signals it is being given.

Now, designing an amp to replicate #1 means you will be limiting the maximum potential power of your amp to allow for that soft distortion curve at the edge of your power supply, so, why dot it unless it is your goal to correct for someone who turns up the volume too loud on a pre-amp which has a +x db gain over-driving your amp and you want to hide that problem.  (Or, just get a remastered CD version which has the volume boosted with soft clipping in the source...)

Designing a circuit to replicate #2 is much more tricky.  To do it right, you need to respond to a speaker load in real-time making a mess of how negative feedback supposed to operate, and in attempting to do so will just mess something else up in the process.  The vacuum tube is such a simple direct device that exhibit this behavior that it is probably the best way to achieve it.  It's just not worth doing it any other way, if someone wants that sound, just buy a tube amp.

Now, what I've said here is not Audio Phoolery like unidirectional speaker cable, special audio grade network cables, special audio cable burn in kits.  There is a price and quality range which does exist.  Some of the better and more expensive (not super expensive) equipment tend to be simpler in design.  No AMP all in 1 ICs, in fact, the all round best I've experienced only have a few transistor/mosfets in the audio path with many paralleled output transistors, large heatsinks and oversized toroid with huge power filter caps.

When it comes to peoples tastes, some people just want loud solid boom and noise as long as it doesn't distort.  Most sturdy negative feedback designs will make them happy with an EQ on occasion.  They also tend to prefer those cheap remastered versions of original music which has just the volume pumped up and at times some cheap anti-hiss filter with occasional treble and bass enhancement.

Those who look for that transparency, who listen to recordings with that quality of mastering, will tend closer to Tube or Class A amps, where the output isn't as linear as a negative feedback design, but, the fine breath of air still makes it through even with some massive fundamental tones included with the sound.

Those that want the best of both worlds, well, you can find affordable equipment to truly perform this feat, but, the louder you want to go, say above 100 watts per channel, the price does sky-rocket.  If you are lucky to have a friend with an audio store who has great ears, gone through the gamut of amps and speakers equipment out there, and is willing to direct you in the right direction for your needs and sell you everything at cost, you are lucky indeed.  I know I paid more than what many here would consider for 2 (stereo) speakers and an amp back in 1992, but what I have is still with me and cremates anything you can buy today at 1/4 the price.  You know when one of my SciFi buddy friends finally came over to watch the new StarTrek movie on BluRay, and halfway through the movie, after being blasted into his chair, gets up an touches the case of the speakers and says these 7inch woofer things case just don't vibrate at all as the house shakes, and everyone sounds like they are actually there right in front of you, not like my cheap surround center channel speaker.  With speakers and amp at over 20 years old, unless I'm willing to fork over 15K$US today, nothing will come close to replicating this experience, especially if you forget the Hollywood movie recorded junk and go for the Telarc CD recordings, even the first from the Telarc classical recordings from the 80s (their technical quality and equipment used is on a completely different level, their original normal CDs surpass most modern hidef recordings), this is where I get immersed into the music.

[Fungus]

When it comes to peoples tastes, some people just want loud solid boom and noise as long as it doesn't distort.

"Wind and piss", as Paul used to call it. 

(Paul was a HiFi salesman I used to know. The "wind and piss" crowd were a large chunk of his customer base)

[Fungus]

All of that is to say – somewhat to your point and the point of others here – amplifier technology has not been on a huge slope of upward improvement since the 1970s, and if you look at price performance it’s possibly declined.

Yep. 1970s was full of small companies trying to make good HiFi. 2017 is full of large companies trying to maximize profits.

Amplifiers peaked around 1980, IMHO.

Having said all of that, I’m still not personally convinced that amplifiers contribute nothing to overall hifi system performance. 

I don't think anybody here is arguing that they contribute nothing at all. It's obvious that bad amplifiers can be built.

[Bicurico]

Hi,

Returning a little to the original topic: can you build a high end audio amp for little money and why are the audiophile ones so expensive?

From the experience I whitnessed with my friend, there are some factors:

1) A good transformer is expensive, especially if custom made for the choosen design
2) Good valves don't come cheap and may hard to source (especially older russian ones)
3) The case is probably one of the most expensive items, as it has to look good, which means it needs to be produced professionally - not cheap at all.
4) PCB needs to be sourced.

Of course you could just copy an existing good design, get yourself a good deal on transformer and valves and then just build everything up in a crude case. Still, you are probably spending more than a cheap mass-produced amplifier.

The basic principle of manufacturing any product is to provide the customer with the worst product he is willing to pay for, as any improvement will not see a higher return.

Because most people are happy with loud bass, as long as there is not a too big distortion and you can disturb your neighbours. Such a one chip amplifier can be sold really cheap and is basically crap. Interestingly, it is not that much of a crap anymore, as the designs and transistors evolved.

Now let's look at an audiophile valve amplifier:

1) The store selling it to you has a price marging of at least 50%. No joke!
2) How many of such amplifiers does a manufacturer sell? Not many, for sure. Still he has to produce, source the parts, give support, etc.

Also, if someone is paying a few US$1000 for an amplifier it really has to look nice: anodized aluminium, wood, granith, whatever. Gold plated connectors and so on.

This does not come cheap and has nothing to do with the amplifier performance, well except for the design of the casing, as it acts as a shield and will have impact on the sound - this requires some expereimenting, which costs money.

Finally, for a commercial product, there is some basic certification that has to be carried out, too.

Also, in case of valves, there is the issue with valve matching. No two valves are equal in performance, so you need a trained ear to select pairs from your batch of valves.

Regards,
Vitor

[Beamin]

Whats makes my Marantz amp sound so much cleaner then the sony amp?

Marketing. I've long suspected that SONY engineers usually mess with the sound (mostly to give it more bass for their consumers).

They why spend lots more money on parts that the person will never see and the advertisers will never advertise? Its just bad business practice to add to cost putting something in your product and not telling anyone about. Look at how CD players have stickers for "1BIT ADC!!!" of course it has an ADC it wouldn't work without one. They advertise stuff that's nt even "stuff" like bragging that it runs on "BOTH voltage AND current" The other brand doesn't have sticker for that therefore it must not be as good.

If your statement is true then EVERY single person that ever paid more then $79.99 and bought an amp other than the junk "Sherwood 100 watt stereo amp" or other low/bottom end brand at every electronics store is getting ripped off. And every manufacturer is needlessly adding costs to their product to see no return on their efforts. That sounds like phoolery.

Go out and buy the shitiest low end 4 channel radio shack brand amp for your car turn the volume up and tell me it sounds the same as the big heavy ones that cost more. First thing is your speakers will sound muddy then clip then it will blow a fuse. But the exact same wattage amp that costs more just won't do that. No.

[BrianHG]

1) The store selling it to you has a price marging of at least 50%. No joke!

Things are not what they used to be, especially here in North America, but when it comes to the higher end specialist equipment, yes there still is a larger margin compared to selling equipment like Sony, Pioneer, Panasonic, Samsung, JVC...

If you want some of the better stuff, it pay's to have a good friend who owns such a store who would give you a good price.

[Bicurico]

On the premium audiophile equipment, the discount is often made at the cost of the manufacturer.

The manufacturer:
a) Provides i.e. 1-2 amplifiers for the "show room"
b) The store tries to sell whatever amplifier the customer likes
c) The discounted sales price is split between store and manufacturer
d) The manufacturer gets paid when the amplifier is finally sold

The problem of the manufacturer is that nobody buys an audiophile amplifier without listening to it first. So there are four options:

a) Work with audio stores, that rip of the manufacturer
b) Be that exquisite, that the potential customers fly in to the manufacturer and listen in his show room
c) Be a manufacturer that ships sample amplifiers to potential customers or, better, fly to the customer to set up a listening session at the customer's home -> this can be quite expensive, especially if the sales ration is low
d) Be famous enough to have potential customers listen at their friends home, who already own that particular equipment

From this difficulty in selling high end audio devices, it is clear that the marketing/commercial price is probably as high or higher than the manufacturing costs.

Cheers,
Vitor

[Zero999]

Just want to point out, that I don't think any one here meant to say that all amplifiers sound the same, full stop. The truth is, that when an amplifier is poorly designed or operated out of specification, i.e. clipping, then distortion and noticeable differences between designs, will occur.

The point is, the difference between well-designed amplifiers, operated within their specifications (no clipping), will be inaudible.

As far as amplifiers peaking in the 1970s or early 80s is concerned, well that's not true at all. Improvements have been made since then in efficiency, cost reduction and size. All with no loss of perceivable audio quality. This is enabled longer battery life in portable speakers and mobile phones, which I know don't sound very good, but that's due to the physical size of the speakers, rather than the amplifier.

I think what has happened since the early 80s is, lots of cheap, poorly designed audio crap has flooded the market and most consumers are unable to tell which is good and bad. Competition has been tough and audio amplifier companies have realised that marketing will generate more profit, than good build quality and design alone, hence the rise of all this audiophool nonsense.

Want to build a reasonably priced, audibly transparent amplifier? DO NOT USE VALVES/TUBES! They're expensive and have high distortion. AVOID PASSING THE AUDIO SIGNAL THROUGH A TRANSFORMER! Audio transformers are expensive and bugger up the signal more than any decent amplifier.

The easiest way to build an audibly transparent amplifier, is with a decent IC with a complete specification. If you just want 60W per channel, then the LM3886 will be more than adequate: much lower distortion than any valve design and most discrete designs. Use a good PCB design, with low inductance power traces and plenty of decoupling. Add a basic clipping indicator, so the volume can be reduced, before noticeable distortion occurs. That's it. You now have an amplifier which outperforms all that BS audiophool stuff, at a fraction of the price, but will still cost much more than a mass produced unit. At least you have the full specification though.

[Alex Nikitin]

1) The store selling it to you has a price marging of at least 50%. No joke!

In my time it was closer to 100% margin. £300 retail price amp was sold to trade at about £150 mark. And it was completely made in the UK ( including the casework, pcb and transformer). And we've managed to make a profit on it.

Cheers

Alex

[BrianHG]

LM3886 = 50watts peak into 8 ohm according to the data sheet, or, only around a true 35 watts RMS.
You can run the IC right to it's absolute recommended limits of +/- 42v, but you might be playing with fire if your power supply isn't regulated, just a transformer with diodes and cap.

Within 35 watts, yes, SNR of 92db + 0.03% THD (THD RATING ONLY AT 30 WATTS PEAK INTO 8 OHMS, OR AROUND 21 WATTS RMS according to data sheet), Yes, the first 20 watts on this chip should sound really good, probably excellent.

Wait a minute, this chip doesn't actually sound too good at all...  It's kinda weak crap, according to the data sheet.
Actually, using 2 of these in a bridge configuration might expand this 21 watt sweet spot to 80 watts, but now you need 4 chips for a stereo amp and an additional input differential amp.

I need an order of magnitude more power with 0.03% THD and with that power, 200watts, I now need a SNR of at least 112db so that when at low volume the first 20 watts is as good as the full volume & while at full volume, even the fine soft sounds below a watt still make it through with the higher power low frequencies mixed in.  Making above 50 watts sound good under all circumstances begins to get difficult.  The higher the power, the more difficult it is to maintain good fine control for the quieter moments yet not begin to distort as the audio power goes above 40% of the output rating.

Don't get me wrong, the LM3886 has it's place and sadly, there are those who will use it claiming over 50 watts (like Chinese single board Amp manufacturers) super clear when the data in the sheet shows it begins it's breakdown after 21 watts RMS.  But if you want to fill and shake a room with controlled refined sound, this chip is definitely not the way to go.

You now have an amplifier which outperforms all that BS audiophool stuff, at a fraction of the price, but will still cost much more than a mass produced unit. At least you have the full specification though.

Except for your word 'outperforms', the rest I cant argue here.  If I want an excellent 20 watt per channel stereo amp, with an extra functionally OK 15 watts extra headroom, but still good quality, and I now know all the specs, I would use the LM3886 and not waste my time designing my own circuit.

As for the price, It would still cost me a little bit for a good chassis for heatsink, toroid, decoupling caps needed to keep this chip fed with a power supply which wont sag too much under a potential momentary peak total 100 watt load.  Not counting the 2 chips, PCB, speaker posts and RCA Jacks and machining the case, AC plug, fuse and all my time, this would still cost me around 150$.  This would be a 200$-300$ project for a nice finished usable super clean 20 watt per channel stereo amp.  I'm still not counting my assembly time, PCB cad time and debugging potential problems.

[Zero999]

LM3886 = 50watts peak into 8 ohm according to the data sheet, or, only around a true 35 watts RMS.
You can run the IC right to it's absolute limits of +/- 42v, but you might be playing with fire if your power supply isn't regulated, just a transformer with diodes and cap.

Within 35 watts, yes, SNR of 92db + 0.03% THD (THD RATING ONLY AT 30 WATTS PEAK INTO 8 OHMS, OR AROUND 21 WATTS according to data sheet), Yes, the first 20 watts on this chip should sound really good, probably excellent.

Wait a minute, this chip doesn't actually sound too good at all...  It's kinda weak crap, according to the data sheet.

It's not that bad. At +/-28V, 60W is specified in to 4R with a THD of 0.03% which will be undetectable. 50W should be possible, in to 8R, with similar THD figures but +/-35V is required.

I need an order of magnitude more power with 0.03% THD and with that power, 200watts, I now need a SNR of at least 112db so that when at low volume the first 20 watts is as good as the full volume & while at full volume, even the fine soft sounds below a watt still make it through with the higher power low frequencies mixed in.  Making above 50 watts sound good under all circumstances begins to get difficult.  The higher the power, the more difficult it is to maintain good control.

My post wasn't a response to you personally. Up until now, no one has specified the power output or any specific designs/components, so I thought I'd throw on in there. My assumption was the amplifier is to be used at moderate volume levels, for home listening, not at a party (less critical anyway, since the amount of people talking and their bodies messing up the acoustics, mask distortion by the speakers/amplifier to some extent) or night club, and 20W or so should be adequate.

Another thing to bear in mind is that, even with a perfect amplifier and speakers, the air is inherently non-linear and distortion will increase, at higher power levels, due to physics, irrespective of how good the amplifier, speakers and room acoustics are.
https://en.wikipedia.org/wiki/Nonlinear_acoustics

Don't get me wrong, the LM3886 has it's place and sadly, there are those who will use it claiming over 50 watts (like Chinese single board Amp manufacturers) super clear when the data in the sheet shows it begins it's breakdown at 21 watts RMS.  But if you want to fill and shake a room with controlled refined sound, this chip is definitely not the way to go.

I agree that the LM3886 is no good in designs above 50W or 60W (unless external pass transistors are added to the output stage, which will be tricky to stabilise)  but refute the claim it will noticeably distort at 21W, unless the power supply voltage or load impedance is too low. The graphs on page 11 the data sheet clearly show, the THD improving, until the power exceeds the clipping level, causing it to deteriorate, from then onwards. The clipping thresholds are clearly 70W into 4R, V_S = +/-28V and 60W  into 8R, V_S = +/-35V. Unlike cheaper chips, the figures given are RMS, not peak power, which can be confirmed by calculations, using the saturation voltage and applying Ohm's law.
http://www.ti.com/lit/ds/symlink/lm3886.pdf

[Alex Nikitin]

Hmm, component cost-wise my power amplifier circuit from 1997 (both channels, omitting the power supply parts) was cheaper than two LM3886 chips with a similar power output into 8 Ohm (40W RMS). I've used N-channel Logic Level MOSFETs for the output in a very linear Class AB follower configuration, and a simple (7 BJT) voltage amplifier to drive it.

Cheers

Alex

[BrianHG]

The graphs on page 11 the data sheet clearly show, the THD improving, until the power exceeds the clipping level, causing it to deteriorate, from then onwards.

At 20Hz only.  Even I can achieve that with a small op-amp feeding the base of a darlington transistor in emitter follower setup.

Even on page 1, under description, paragraph 1, the claim is 0.1% from 20Hz to 20Khz only at 38 watts into 8 ohms.  This does not jive with the graph on page 11 where they are only showing you the distortion at the bottom 20hz which shows 0.005% THD at 38 watts.  We are not being told the whole story here by TI and you are ignoring something fundamentally important.  (I'll wait and see if you can figure this one out for yourself (it's a doo-seee  )...)

Ok, I know how you feel about audio phoolery stuff as there is a vast cloud of such BS crap, but, there does exist quality 50watt amps you can buy, where when they say 50watts 0.01%THD, they mean that at all frequencies and loads.  Most of the time, these amps are basic and only have a single volume control.  But, such devices are not in demand and they do cost $$$.

But like I said with this chip, and TI's pdf on page 4 states:
30W, RL = 8?, 0.03 %
20 Hz ? f ? 20 kHz
AV = 26 dB

With this spec, I take as their sweet spot for this IC, and looking at the built in current limiter charts and how they may react to a speaker's inductive load, I will still say that you should power this IC for 50 watts peak, expect a super clean 21 watts RMS. And my assessment of building a nice amp chassis with heatsink, good toroid and nice decoupling caps to support 2 channels with 50 watts peak, PCBs, and all other parts, this is still a good 200$-300$ project to create something respectable enough and safe to put in my living room.

[paulca]

First of all, there are no legal standards against which audio amps are measured. So, every manufacturer make up its own specifications and limits. The "shit sony" gives a THD rating of 0.8% not because it actually produces that much distortion most of the time (it definitely does not), but because that THD rating allows higher power ratings. Most amps produce least distortion at a couple watts output, and increase from there. If that sony was rated at 0.08%, then maybe it could only manage 50 W before hitting that mark. But move the limit to 0.8% and it can push out 80 W at that limit. So do they market it as a 50W @ 0.08% or 80W @ 0.8% ? Sony chooses one way, others like HK go the opposite way. Most people buy based on power numbers, not THD numbers. The fact is that even a "shit sony" produces much less than 0.05% most of the time. (I'm not just waving my hands around, I have measured plenty of them).

I bought a Sony HiFi way back in the 90s.  It was in the era when Awia were pumping out HiFi's claiming things like 300W RMS.  Utter tosh.  Sony countered with an HiFi that could actually meet it's rating of 80W RMS / 160W peak a side.  They also matched it with speakers that could take that (and probably more).   I don't have the THD figures handy, but having owned an Awia(sp?) it distorted long before it go anywhere near it's half way mark.

Word on the market was it was very much under priced just to piss Awia off.

I still have that amp.  I can't find any reason to buy a different amp.  It will have my neighbours at the door when the pictures start rattling on the walls and it still has room to go.

I have tortured it.  I once connected a really cheap karaoke mic to it to provide sound re-enforcement for a presentation.  Due to the low gain of the mic on the podium and the distance from the speaker I had all the gains at MAX.  Some bright spark clown grapped the mic and shouted "OWW!" into it.  CLICK.  The amp just went into protect for 2 seconds and came back on without a hint of being upset.

I used it at a large party to provide music in the kitchen/dining room while the main party speakers (an 800W professional hired rig) did the front room.  I simply left the volume at max and let the DJ run it however hard he wanted from the mixer (I know... one does not do this!)  Not only did it sound amazing, survived, but most people said the sound was better in the kitchen than in the main room.

Still not getting rid of it.  The CD unit died about 10 years ago, the tape decks (yes it had those) were removed and binned about 8 years ago, so I just have the top half of the HiFi remaining.  The rotatry encoded volume knob and EQ adjust don't work anymore, but the remote does.

It sits left on 100% of the time in my living room set to high levels of 24 of 32 on the knob.  You simply can't here it, not a single hiss, buzz or humm, but drop a movie Dolby intro and things start vibrating at some serious power.

Yes, there are probably better amps, but I kill amps pretty regularly.  This one seems to survive me so I'm keeping it.

[paulca]

On modern amps what peeves me is the "Nanny state" limiters.  Particularly headphone amps.

I used to use a TRIO 1980s amp to drive headphones.  If I so choose I could turn the volume up enough to melt the drivers in seconds.   Believe me, I did.  I think that amp ate about 4 pairs of panasonic and techniq headphones in the 5 or 6 years I had it.  Until one day it just stopped working.

These days you'll find it hard to get that.  Most headphone amps are limited to "sensible and health listening levels".  What is the point in listening to rock music at sensible levels?  Seriously.

If want to destroy my ears then I should be allowed to.  If I want to turn it up far enough to set my headphones on fire, then I should be allowed to.

The other thing that makes me chortle is PC Speaker systems.  Like these:
https://www.ebay.co.uk/itm/Genius-SW-F2-1-500-speaker-system-for-PC-Laptop-MP3-12W-R-M-S-500W-P-M-P-O/282616179220?hash=item41cd3caa14:g:d0MAAOSwhGlZl0H4

12W RMS... possibly.  500W PMPO, LOL  I have seen speaker sets advertised as 1500W but the power pack supplied delivers 10W.  They simply make stuff up!

As to what makes a good amp. I think it comes down to quality of the components used, not just the amps and transformers but the support components too.  99% of domestic goods HiFis will use the cheapest white label components they can get.  The bean counters will strip anything they can back.  Then there are the support features, such as sensible limiters, multi band limiters, sensibly multi-staging with sensible gain structures, proper protect, rugged-ness, thermal protection, clean PSUs, RF shielding, gold plated connectors etc. etc. etc. 

The price of high end professional amps isn't really due to the overall price of their components and design, it's more about what audiophiles will or expect to pay for it.  This is not limited to Audio gear.  Take any market and search ebay for a price ranges and you will find a wide range of prices for the exact same item.  People will log in and say, "I want to buy a thingie", but..., "I expect to get a good quality one it will cost at least £100".  So they set a minimum price of £100.  Chances are the very same seller is selling an item for £10 and £100 and maybe £200 and people will pay that, because it's what they expect to pay.  They believe they always get what they pay for.

High end gear also suffers from low demand.  However if you need a guaranteed quality amp for a studio you will have to pay for it.  So they can charge, sometimes, insane amounts of money to get something that is 2% better across the spec range than an amp costing a 1/10th as much.

[BrianHG]

On modern amps what peeves me is the "Nanny state" limiters.  Particularly headphone amps.

I used to use a TRIO 1980s amp to drive headphones.  If I so choose I could turn the volume up enough to melt the drivers in seconds.   Believe me, I did.  I think that amp ate about 4 pairs of panasonic and techniq headphones in the 5 or 6 years I had it.  Until one day it just stopped working.

These days you'll find it hard to get that.  Most headphone amps are limited to "sensible and health listening levels".  What is the point in listening to rock music at sensible levels?  Seriously.

If want to destroy my ears then I should be allowed to.  If I want to turn it up far enough to set my headphones on fire, then I should be allowed to.

Separate from this topic, start a new headphone amp project thread.  You may find an affordable solution here among us which might be affordable and please you......

[cdev]

What are the best amp modules for use with external line level audio, like from a computer?

I would like to upgrade my computer's sound with a clean powerful quiet audio module that can use my power supply and drive good external speakers.. like the small Dayton Audio ones without distortion.

[paulca]

Separate from this topic, start a new headphone amp project thread.  You may find an affordable solution here among us which might be affordable and please you......

Thanks.  I got a digital Projekt Box headphone amp which I love, massive range, quiet, clear, the only downside is, I wish it would go just a little bit hotter beyond it's nanny state limiter.  I'm not entirely convinced I could do better with my electronics skills.

[BrianHG]

What are the best amp modules for use with external line level audio, like from a computer?

I would like to upgrade my computer's sound with a clean powerful quiet audio module that can use my power supply and drive good external speakers.. like the small Dayton Audio ones without distortion.

Need more info...  Are you talking about just using your PC's 12v?  If you want a total of 50 watts out, is it good for you PC's +12v rail's load to jump around 2-52watts with the sound?
What kind of speakers?  8 Ohm speakers need a respectable voltage to give you a reasonable power.  Like +/- 24v for a unbalanced output amp, or +/-12v for a bridged balanced output amp.
A bridged DC 12v class D amp module from China should give you around 15 watts peak per channel into an 8 ohm speaker, or, 35 watts peak per channel into a 4 ohm speaker.  (This is a respectable power output for PC speakers which will fill a small room really good)

[alexxa]

I have been professionally in the field for many years.
The simple and only (scientific) correct answer is cialis
 marketing and placebo.
It doesn't cost much at all to make a perfect good performing amplifier.
Although, you have to realise that it all depends on the context. Pro audio customers have different needs than tiny boom box producers, as well as guitar amplifiers.
So that means the approuch and compromises will be (very) different as well.

The QUAD 303, from 50 years ago, was one of the first all-transistor audio power amplifiers to be essentially 'perfect' (within its bandwidth, voltage swing, and load impedance limits, of course). Any more recent design that isn't 'perfect' has been done wrong!

[Zero999]

The graphs on page 11 the data sheet clearly show, the THD improving, until the power exceeds the clipping level, causing it to deteriorate, from then onwards.

At 20Hz only.  Even I can achieve that with a small op-amp feeding the base of a darlington transistor in emitter follower setup.

Even on page 1, under description, paragraph 1, the claim is 0.1% from 20Hz to 20Khz only at 38 watts into 8 ohms.  This does not jive with the graph on page 11 where they are only showing you the distortion at the bottom 20hz which shows 0.005% THD at 38 watts.  We are not being told the whole story here by TI and you are ignoring something fundamentally important.  (I'll wait and see if you can figure this one out for yourself (it's a doo-seee  )...)

Ok, I know how you feel about audio phoolery stuff as there is a vast cloud of such BS crap, but, there does exist quality 50watt amps you can buy, where when they say 50watts 0.01%THD, they mean that at all frequencies and loads.  Most of the time, these amps are basic and only have a single volume control.  But, such devices are not in demand and they do cost $$$.

But like I said with this chip, and TI's pdf on page 4 states:
30W, RL = 8?, 0.03 %
20 Hz ? f ? 20 kHz
AV = 26 dB

With this spec, I take as their sweet spot for this IC, and looking at the built in current limiter charts and how they may react to a speaker's inductive load, I will still say that you should power this IC for 50 watts peak, expect a super clean 21 watts RMS. And my assessment of building a nice amp chassis with heatsink, good toroid and nice decoupling caps to support 2 channels with 50 watts peak, PCBs, and all other parts, this is still a good 200$-300$ project to create something respectable enough and safe to put in my living room.

The distortion does rise at higher frequencies, but I don't see what all the fuss is about? TI don't seem to have manipulated the figures at all!

I admit I missed the fact the figures I quoted were only for 20Hz, but plenty of data is provided for higher frequencies.  Refer to figure 18: +/-35V, 50W into an 8 Ohm load, THD+N <0.01% <10kHz, rising to slightly above 0.013% at 20kHz, but no one will notice the harmonic distortion above 10kHz because it will be ultrasonic. The situation (figure 16) is worse with a 4R load but still <0.03% THD+N to 20kHz. Graphs are also given at 20kHz, see figures 25 to 27, which is helpful, but beyond the audio band.

If anything TI have made the figures look worse than any one can hear. The THD+N tests were conducted with a bandwidth of 80kHz, which is excessive, because it will show a lot of ultrasonic harmonics, which no one will hear. If the bandwidth were lowered to actual audio range of 20kHz, it will more accurately reflect what can be heard, rather than what can be easily measured.

[paulca]

What are the best amp modules for use with external line level audio, like from a computer?

I would like to upgrade my computer's sound with a clean powerful quiet audio module that can use my power supply and drive good external speakers.. like the small Dayton Audio ones without distortion.

I would recommend an external digital USB amp (ideally optical rather than USB, but you will be lucky if your Mobo supports it and the drivers do, so might need a PCI sound card).  From £50 to £5000 it just depends on what you want to spend.  I wouldn't recommend trying to use the PC power supply.  They are incredibly noisy devices, lots of HF RF around, this is why internal sound cards and mobo built in audio chips are so noisy.  I expect a lot of that noise will end up on the high power 12V rails too.  I wouldn't put the amplifier inside the PC case either.  I'd put it as far away from the PC as is practically possible.

For speakers I have something similar to these for gaming and desktop use.  Cheap, cheerful, great sounding, but they are designed for close range listening, they lack bottom end and roll off completely low down (not suprising for 4-5 inch speakers) .  They were a present. 
https://www.djstore.com/numark-nwave360?gclid=Cj0KCQiArYDQBRDoARIsAMR8s_SwIrg5LckfvhsgCEBmLrFdxkW5nxiHAU-Da_n7Ki78HgjDfTxdmjoaAlzhEALw_wcB

If I had of bought them for myself I would looked at similar but with a USB digital or optical input and a 2.1 configuration.  Nurmark make "okay" hobby/semi-professional DJ equipment.  Denon make higher end stuff, either will probably do a digital input set of active speakers.  If you've got the money there are much higher end stuff, but what they really give you over the £100-£300 range, I can't say.