[JohnR]

I've never heard a digital amp. However, just as a general comment, it seems that semiconductor fabrication advances will have impact on more areas of audio as we go along. Some version of Moore's law if you will. Look at the price/performance of CD/SACD/DVD players coming out, for instance. The weaknesses are in cost-cutting areas which can be remedied by a tweaker, but the fundamental semiconductor technology is top notch. The promise of digital amps for the big companies is being able to deliver a lot of power cheaply in high volumes -- HT receivers, cars and so on, which is why loads of companies are investing in the technology. Not so anyone can sell a few "high end" amps. However, the trickle-down effect will happen, just like with DACs and decoders, and I'd guess much less thean 10-15 years.

Anyway that's my theory

[DVV]

I've never heard a digital amp. However, just as a general comment, it seems that semiconductor fabrication advances will have impact on more areas of audio as we go along. Some version of Moore's law if you will. Look at the price/performance of CD/SACD/DVD players coming out, for instance. The weaknesses are in cost-cutting areas which can be remedied by a tweaker, but the fundamental semiconductor technology is top notch. The promise of digital amps for the big companies is being able to deliver a lot of power cheaply in high volumes -- HT receivers, cars and so on, which is why loads of companies are investing in the technology. Not so anyone can sell a few "high end" amps. However, the trickle-down effect will happen, just like with DACs and decoders, and I'd guess much less thean 10-15 years.

Anyway that's my theory

Perhaps I phrased it crudely - what I meant was that in 10-15 years, they will have the current problems more or less resolved to a point where we can call it a day.

As for the trickle down, why John, it's happening already, we don't have to wait a single day for it. Hitachi is offering mini systems with digital amplification, and before you know it, everybody and their dog will be offering it.

But that fact does not mean they will sound great, does it?

And make no mistake about it, semiconductor fabrication techniques will reduce our entire beloved audio to just another chip in the PC sooner or later, and it won't be a long wait, methinks.

Cheers,
DVV

[JohnR]

As for the trickle down, why John, it's happening already, we don't have to wait a single day for it. Hitachi is offering mini systems with digital amplification, and before you know it, everybody and their dog will be offering it.

But that fact does not mean they will sound great, does it?

Perhaps trickle-down is the wrong word I don't mean that you will be able to buy cheap amplification, I mean that you will be able to buy very good amplification relatively cheaply...

For instance, currently, many believe that the best bang for the buck in digital playback is to buy a relatively inexpensive consumer player (or DAC) and modify it. In fact there's a booming market for people who offer this service, eg Bolder and Modwright.

I'm suggesting that something similar will happen with digital amplification. The best bang for the buck will be to buy a relatively inexpensive HT receiver or multichannel amplifier, and modify it. It may not be the absolute best, but it will be very good, as good as almost anyone would want.

[JohnR]

Having said that, there's also another angle... which is that, since these digital amps are often built as complete modules, it will soon be quite easy for the DIYer to make an N-channel amplifier with loads of power at a reasonable cost. B&O already have a module that includes a switching supply, I don't know how good it is but it can only get better. (And cheaper.)

OK, enough crystal ball gazing

[DVV]

As for the trickle down, why John, it's happening already, we don't have to wait a single day for it. Hitachi is offering mini systems with digital amplification, and before you know it, everybody and their dog will be offering it.

But that fact does not mean they will sound great, does it?

Perhaps trickle-down is the wrong word I don't mean that you will be able to buy cheap amplification, I mean that you will be able to buy very good amplification relatively cheaply...

Oh, I understood what you wanted to say, John, no problems there. But exactly the same was said and expected of every new technology thus far, and all of them, including the ancient vinyl, are still seeing improvements as we speak. Just look at the CD - still evolving, and you still have to pay kilobucks for top notch performance (on an absolute scale).

For instance, currently, many believe that the best bang for the buck in digital playback is to buy a relatively inexpensive consumer player (or DAC) and modify it. In fact there's a booming market for people who offer this service, eg Bolder and Modwright.

I'm suggesting that something similar will happen with digital amplification. The best bang for the buck will be to buy a relatively inexpensive HT receiver or multichannel amplifier, and modify it. It may not be the absolute best, but it will be very good, as good as almost anyone would want.

I agree that what you say is indeed the most likely scenario. However, there are problems associated with digital amplification which have yet to be resolved, and the worst of it is that they have in absolute terms nothing at all to do with digital amplification as such.

I refer to switching power supplies. Sure you can make a good one, even a great one, but you end up paying as much, or more, than for an equivalent analog power supply. The reason is as prosaic as it is obvious - these PSUs tend to be "notchy", i.e. with a stepped rather than smooth waveform. The way to minimize the problem is to use power regeneration, but this in turn requires high quality, and hence high price, processors. And you still need a mirror image of your power amp in the power supply section, which kills most of the benefits of highly efficient switchmode power supplies.

Look at Chord, a UK company of high repute - their PSU costs as much as the power amps inside. Together, they warrant another mortgage of the house.

And our prpblems will compund elsewhere. Switchmode power supplies are notorious for polluting the grid, which means our power will just get dirtier and dirtier as time goes by.

Cheers,
DVV

[DVV]

Having said that, there's also another angle... which is that, since these digital amps are often built as complete modules, it will soon be quite easy for the DIYer to make an N-channel amplifier with loads of power at a reasonable cost. B&O already have a module that includes a switching supply, I don't know how good it is but it can only get better. (And cheaper.)

OK, enough crystal ball gazing

I beg to differ - cheaper, yes, better - no, in fact, worse.

Yes, you will be able to buy digital power amp modules for less and less money, but the savings will need to be accellerated beyond price drops of the technology itself. This means less money to spend on power supplies, and John, digital or analog, but the power supply still remains the most critical part of any power amp.

Believe me John, I don't use oversized power transformers because I get a hard on from big toroids, but because it directly affects the sound. If I want clean power delivery into anything reasonably called a speaker, I have to be able to deliver real world watts, and lots of 'em, into mean and nasty loads. And there's no other way to get that except by making your power supplies big and mean.

Switchmode can be done VERY cheaply, although there is a problem with it; these power supplies were initially designed for low voltages and high currents, as in PCs, and it's VERY hard getting both high currents and high voltages from them. It stops being cheap and simple very quickly, and turns expensive and dirty real fast.

Cheers,
DVV

[JohnR]

Well, there's no rule that says that you have to use a switchmode power supply with a switching amp.

[Dan Banquer]

Just a quick note here: 48 Volt, 10 Amp Switchmode power suplies can be had for around 200.00 to 250.00.

[DVV]

Well, there's no rule that says that you have to use a switchmode power supply with a switching amp.

Quite so - but then you are half way back where you started from. What I mean is, you will still have power supplies which are undersized, as they generally are today, because they cost a lot.

While not having tried it myself, I'm not sure what would the impact be on a digital amp being fed off classic analog power supplies - should be nothing, but one never knows until one tries.

Lastly, doing that sort of negates the convenience of an all-in-one module, doesn't it?

Cheers,
DVV

[DVV]

Just a quick note here: 48 Volt, 10 Amp Switchmode power suplies can be had for around 200.00 to 250.00.

For $250, I'll do you a +/-48V 2x15 ampere power supply using analog.

Cheers,
DVV

[Raj]

Hey DVV,

if you're in the mood for selling perhaps a power amp also!!!    

Thanks
Raj

[DVV]

Hey DVV,

if you're in the mood for selling perhaps a power amp also!!!    

Thanks
Raj

Ooooooooooooooh, you naughty man!

What is it with you guys, everybody seems to want me to make a power amp kit? Why is that such a big deal, I would have thought there were quite enough kits already available.

Actually, I might have done it already, but what's stopping me is the logistics behind it all. For a kit to be that, I'd have to offer all the parts with it. This can of course be done, but I'm not sure the end price would be convenient.

My greatest contraint are my local laws, which make exports, ANY exports anyhow, a VERY complex job with much paperwork. I am simply not equipped to handle that.

Then again, perhaps I could get away with just the circuit boards, mailed to a couple of addresses only, then redistributed from there. I think it's the printed circuit boards which are by far the largets contraint on DIY.

I guess something like 70-80 watts per channel should be enough for everybody, assuming of course outstanding load tolerance into 4 and even 2 ohms.

What would be considered a reasonable amount people would be ready to pay for a power amp? Let's assume a modular approach: in original form, just the voltage gain stages are fully electronically regulated, and in an advanced form, the current stages could be fully electronically regulated as well, either straight out, or later on. Optional balanced input. Full DC servo, no pumping. Full DC and overheat protection. Overload threshold warning indicator.

No @#$%^&* capacitors in the signal path.

Before anyone answers, go to my site at http://www.zero-distortion.com , to the download section, and check out the available projects there. For example, there's an Elektor project of an amp delivering 150W/8 ohms (and 290W/4 ohms), for which I can supply scanned printed circuit artwork, with full protection and rather easy to build. Sounds good too.

Cheers,
DVV

[Raj]

Hi DVV,

actually i was joking (only teasing, I know it's not really what you want to do!), but hey you're suggestion is pretty good.

Best regards
Raj

[JohnR]

Halfway back is still halfway forward

I believe most, if not all, of the high end class D amps presently on the market use linear supplies.

[DVV]

Halfway back is still halfway forward

I believe most, if not all, of the high end class D amps presently on the market use linear supplies.

Like I said, John, there's no doubt in my mind that digital amplification is the future, and that no matter what problems there are or have yet to be discovered, they will be resolved sooner or later.

As for power supplies, I believe most digital amps use switchmode power supplies as of this writing. Their manufacturers will tend to save wherever they can, given that the digital technology is still rather expensive as such. But we could see changes soon enough.

Cheers,
DVV

[DVV]

Hi DVV,

actually i was joking (only teasing, I know it's not really what you want to do!), but hey you're suggestion is pretty good.

Best regards
Raj

Wrong Raj, actually, I'd love to do that. For example, put my trusty 1M (that's what I call it, "1M" stands for 1 MHz) into some of your hands. The problem is that I don't have an accompanying preamp for it, I'm only just starting on that.

What I want to do is make a preamp with tone controls unlike any I have ever seen anywhere. First off, they would have to act at the extremes only, without even touching the midrange, because the extremes are where we actually need them. Next, I want exactly repeatable steps in increments of 1 dB, namely -3dB, -2dB, -1dB, 0, +1dB, +2dB and +3dB; this is all that's really required to compensate for the room/speaker interaction. Lastly, I want the tone control circuit to have as little sonic signature as possible, ideally zero, but I'll settle for very hard to hear at all.

Now, I have some ideas about how to do it, but I'll have to do a lot of work before I'm satisfied. The rest, like kicking the distortion off the meter scale, hitting 2 or 3 MHz is easy, no big deal, a lot of people can do it. Of course, 2 or 3 MHz is its internal (inherent) pass band, it's externally limited to much less, typically 500 kHz (value chosen because with it, the phase shift in the 20-20,000 Hz band is less than 0.5 degrees - practically non-existent).

I would also want a variant of it with two or three parallelled outputs, internally filtered at desired frequencies, thus combining the preamp with an electronic crossover function. Opening the door to active amplification, if you like. I happen to be a great believer in the active approach.

And I would want that preamp to cost something most people could afford without selling mom and dad down the river. Say, 500 euros max, which is about $550, or about 350 quid for you, on the condition that they do NOT have to plonk down all the dosh at once, but can do it over time.

That's why I never put any project like that in public view. It has to be oustanding in my view before I start enticing people to spend money on it.

Cheers,
DVV

[shaynet]

The problem with digital amps is that they only overcome some of the practical limitations/inconveniences associated with analog amps.  For example, they are far more effecient, thus removing the requirement for a lot of heat sinks.  They also have less problems with cross-over distortion.

Having said that however, good analog amps have the right amount of heat sinks and solved the cross-over distortion problem.  A good analog amp will also have a LOT less distortion, typically lower high frequency noise, and better damping.

As a designer of very high end digital equipment once told me - the perfect implementation of analog is perfect, the perfect implementation of digital approaches analog.

[Raj]

Hi,

What's the craziest claims you've heard in some commercial equipment, all those keyphrases that companies use to tell us about a radical 'new' concept (that turns out to be nothing more than a better component or 2, or a new name for an old topology)?

Thanks
Raj

[DVV]

Hi,

What's the craziest claims you've heard in some commercial equipment, all those keyphrases that companies use to tell us about a radical 'new' concept (that turns out to be nothing more than a better component or 2, or a new name for an old topology)?

Thanks
Raj

My favorite is Krell's "Sustained Plateau Biasing". This is in effect another variation on the sliding bias theme, although I would imagine Dan d'Agostino did a better job of it than most (he's usually very thorough, also extremely competent). But I dig that there name, man, it has a ring to it.

Reminds me of the way words are added on in German, so you end up with a military rank of "hauptgruppenoberstrumbahnfeuhrer".

Cheers,
DVV

[Raj]

'Sustained plateau biasing' , who do they pay for this kind of marketing work? I've thought about anything alse I've heard that even comes close, and I can't think of anything.  Perhaps they could do a funky name for a power supply using dual bridge rectifiers, then Dejan we could form a patnership and sell it for megabucks!

Thanks
Raj