I have no idea what you're talking about, Mike, as in, you don't really construct meaningful sentences that reflect a thought.
Nelson Pass relayed a story at BA of a rep who came by the shop and tried to sell him on building with class D modules. He kinda chuckled, told the guy that's not what we do here, and showed him the door. Ain't nobody trying to build a class A amp that sounds like a class D amp. Nobody has ever said "That sounds almost as good as class D!".
If anything the trend in better audio has been veering towards more and more simple, linear topologies. Class D isn't considered by anybody to be a superior way to amplify sound. Class D is the solution to a set of problems where sound isn't the only concern. For instance, if you need to meet government regulations like those coming down in Europe, or if you need to stuff a very small amp someplace with very little air flow.
Getting a real nice signal with an untortured distortion signature is tricky enough to do with conventional linear topologies. I think the best class D will ever be is almost as good as class AB or A. |
@aberyclark
No, not really. The ONLY downside to class A amps is that they're obscenely inefficient and as a result make ungodly amounts of heat. Other than that they're technically superior in every way.
Transistors make a lot of garbage when you go switching them fully off. That's why you apply a little bit of bias to a class AB amp. Otherwise it's just 2 class B amps with a ton of distortion. Class D doesn't bias the transistors at all and switches them at very high frequencies and that creates huge amounts of noise and distortion. The output of a class D module then needs to be filtered extensively to block the huge amount of high frequency noise and smooth the notchy sort of output into something resembling a waveform.
The major advance in class D is stitching the transistors on and off even faster to achieve finer granularity in the pulsed output. That makes it easier to filter. Filtering isn't really advancing all that much. That technology has existed for a long time. It's just a debate of how to best apply it to this problem.
I, and many others, think simple, linear signal chains from the source to the speaker sound best. It's really hard to beat a class A amp with a nice, clean, low level, low and even order distortion character. That's very hard to do with class D.
|
@mapman
What does anything I said have to do with what I've heard? I was just roughly explaining the technology to a guy who asked a question.
I'll go listen to a pair of horns and give my honest opinion. I'm not going to waste my time on class D. Unless you've got priorities more important than listening to music, why would you want to reduce the signal to garbage, inject it with ungodly amounts of noise and distortion, then try to filter out something that sorta resembles what you started with? I'm not even remotely interested in that process. You'll never catch me bragging about how efficient my stereo is either. The best class D could ever hope to be is almost as good as class A. Some would say that about class AB too. |
Really Mikey? That's why Pass has no problem shipping big XA series amps that keep him comfortably house on the coast of CA? Unlike you. That's why it's becoming a viable venture for some to develop tubes again? Like the one Korg came out with a little while ago? Like the guys starting to make newer and improved versions of the typical types? That's why you can buy tube buffers and tube headphone amps all over the place? That's why the really top end amps are NOT class D amps, right?
I could just as truthfully say that the only people who buy hifi stereos are fools or those with some quirky nostalgia. Everyone else is going with soundbars with DSP and bluetooth speakers... with class D amps!!! Unlike your statement, mine has been the trend for several years now. I don't think yours will ever be the trend except in those places where the government mandates it, which is really the only reason that technology is being pursued in hifi. |
@aolmrd1241
Nobody has actually ever seen a subatomic particle, but people earn graduate degrees studying them. Having heard a class D amp, which I'm certain I have, is hardly a prerequisite to understanding how the technology works. We don't require car mechanics to be engineers or formula 1 drivers, do we? A comprehension of CMOS and material science or advanced driving techniques isn't required to replace an alternator or transmission. They merely need to grasp the basics of how the technology works. It's a basic logical inconsistency to presume that one must have advanced driving skills to be a competent mechanic, and that's the same logical inconsistency you're trying to assert here.
I'm pretty sure the technology was all I was talking about except for the last bit where I injected my personal opinion on preferring simple, linear topologies with simple, low order distortion. I am right in saying that class D doesn't do that very well. Almost nothing I said is up for debate or rooted in my opinion or experience. Those are technical facts that have nothing to do with listening. I think you're just attacking me because you don't like how I described what class D amps do. It certainly doesn't sound pretty. Lots of people like sausage, but making it isn't a pretty process. Grilling up a filet minion is a much more pleasant experience, like class A.
If you like class D, cool. I'm personally not interested in it on the sonic or technological level. I think the job of an amplifier is more than behaving like a straight wire with gain. |
I doubt their time will ever come. We're not talking about cars here where the real limit of their acceptance isn't the technology you directly interact with. It's the technology you don't directly interact with; the batteries with their limited power density and charge time. Technologically, no class D amp will ever compete with a class A amp. It's simply impossible to build a class D amp that reaches out to 500,000Hz and doesn't burp driving DC into a 2 ohm load. Complicated 6, 7, and 8 stage circuits had their heyday back in the 70's and have given way to more and more simple topologies and I think class D will be a fad we look back on in the 30's like we do the early 70's gear now. |
@cleeds
No, there's no NEED for an amp to drive DC or just a few Hz, nor reaching out to .5MHz. On the other hand, I like the sound of an amp that drives DC and is -3dB at 1MHz. It's also nice not having a massive pile of filters trying to turn garbage into a signal again.
Nobody has apparently noticed the massive popularity of vintage gear that's functionally very simple stuff. I watch that stuff fly off the shelves of the shops near me. Mikey's assertion that young people are all into tiny class D garbage is pure nonsense. Millennials are the ones driving the massive price spikes in vintage gear and they're buying it because it looks good, sounds good, and isn't digital. |
@stevecham
I'm talking low level distortion. .5, 1, 2, 5 watts... Very, very few class AB amps perform as well as a class A amp at single digits or fractions of a watt. Looking at a few class D examples measured on Stereophile, the low power distortion is horrendous. Looking at the Mytek Brooklyn, it makes a nifty .002% at 50 watts, but at .1 watts it's almost .02%, and the distortion waveform looks like a dumpster fire. None of the other measurements look much different for any other class D amp. That's an ungodly amount of distortion for such a low power level. Nobody can say with a straight face that kind of distortion isn't significantly coloring the sound. Most of what we call treble is defined by the behavior of amps below half a watt.
Sorry, but even the most rudimentary 2 stage single ended class A amp can best that kind of performance. |
@stevecham
Yes, .02% at 1 watt is completely insane when it's all high order garbage like class D makes. Go look at the measurements. It's a very different story when .02% is this pretty, negative phase, entirely second order waveform. The human ear doesn't perceive that as distortion or dynamics like it does with high, odd order distortion. It tends to perceive it as depth. It's not technically difficult to build an amp that gets .001% distortion across the power band. It's generally thought those amps sound cold and analytical. The things that really define the character of an amp are the response curves of the gain devices and the character of the distortion they produce when driving a load. Folks aren't going nuts about triodes for their efficiency, gain, and low distortion. People love those things for their distortion and that awesome depth and spaciousness you only get from triodes and single ended JFET amps, and you only get that with that sexy, negative phase, low even order distortion.
|
@mr_m
I think it's pretty safe to say that an amp that makes 10% distortion at 2 watts isn't even hi-fi by the definition of the term. You can get better distortion than that just feeding voltage onto the grid of a triode connected to a power supply and a load with just a couple of resistors. You'll also get this really nice second order harmonic, not some brutally offensive high order harmonic. |
|
@stevecham
I'm certain that I could tell the difference between .02% and .005%. I could tell you if it's higher or lower order, and I could tell you the phase as well if it's the low order type. Phase is irrelevant beyond 4th order. I spent most of a year voicing my amp for my speakers, trying this and trying that, and I definitely know what different kinds and amounts of distortion sound like. |
|
I'd say at .1% at 3.2 watts your amp is acting like a tone control. At a whole 1%... That's gratuitous. .0X% distortion is where most people start hearing things. Depending on distortion character, you get hardness, harshness, depth, and timbre. Amps with distortion in the low .00X% tend to sound very crisp and precise, but cold and analytical. When you compare listening impressions to measurements you find a coorelation along those lines.
I think a little 2nd order distortion is a good thing. I think the same amount of 5th or 7th order distortion is a horrible thing. I think most would agree. |
@mr_m
I don't think so. Measurements reflect it. If you listen to half way sensitive speakers, 90dB/watt or better, .02% high order distortion is going to be problematic. |
@mr_m
The kind of distortion is extremely important. 1% being composed of a nice, clean second harmonic isn't good in my opinion, not because it doesn't sound good, but rather because you're basically using the amp as a tone control. That's significant color to add to a signal. That's not to say it sounds bad. A lot of people want that kind and level of coloration. Nobody likes complicated distortion. Some fuggly mix of 5th, 6th, and 7th order distortion at .1% is going to be a lot more offensive than a big, pretty 2nd order at 1%. The kind of distortion is more important than the amount.
The problems you see in class D are the same you have with class B. The lower the power output, the more the distortion dominates. Class D just tries to minimize the distortion by reducing the time the transistors transition from off to on by pulsing them very rapidly. But no matter what, transistors switching off and on will always make huge amounts of high order distortion. It's unavoidable. It's just imppossible to fully filter it out. If it were possible, it would have been applied to class B amps by now. |
Hi Mikey!
Your statements about nobody understanding distortion or it's effects are completely unfounded. Your post seems to suggest you like cheaply built, poorly designed products. I think I'm starting to understand why I can't find you credited for producing anything. I don't think many musicians want their work to be used simply to listen to dismembered, poorly designed, malfunctioning equipment. |
For me, personally, I'll never again use an amp I can't voice to my speakers. Anything else is like buying suits off the rack. That's why I won't bother with class D. |
SMPS's are notoriously hard on caps. The fast switching speeds require very low resistance parts and still create a lot of heat inside them. That's the typical point of failure, and then many things go after that. |
@georgehifi
Thanks for posting that!
Foggythink,
In a Class A or AB amp, the filter caps need only respond to the ripple demands resulting from an audio signal. In an SMP for a class D amp, caps need to respond to switching action up in the ball park of a good fraction of a megahertz. They need to exhibit very high ripple current ratings and extremely low equivalent series resistance. Electrolytic caps aren't famous for having exceptionally low ESR, but they're the only technology that's compact enough to provide the current capacity required. The sub-optimal ESR leads to internal heating of the capacitor. The heat dries them out, the ESR rises, that causes them to make more heat, they begin leaking electrically, and they eventually just short out and start frying silicon.
Nobody will be calling any class D amp "legendary" 10 years after it was sold. They'll all be in landfills. The ONLY amps sold today that will still be around in 20 years will be class A and AB amps, and they'll still sound very good. |
There's a huge difference between a pre-amp and a power amp. SMPS's fail all the time in applications that require real current. Laptop chargers, battery chargers, TV's, you name it. These things failing isn't a rare occurrence. It's an everyday thing. There are businesses out there that do nothing but repair SMPS's for all kinds of machinery, and it's generally the caps that go to hell. |
|
@atmasphere
How do you figure? A laptop charger outputs several amp. It has to charge 6 or 8 3000mAh batteries in a couple of hours. Ain't no solid state pre-amp on the planet that has those kinds of power demands. |
@atmasphere
That's a pretty small laptop charger. My Acer has a charger that does twice that. 19V@6.32A.
No need to argue with me though. The guys you really need to argue with are the thousands of people replacing SMPS's every day and every guy who's ever worked on CNC machines. There's nothing cheaply or poorly engineered about Mazak or Haas machine tools and they can't seem to make these power supplies last. These are massive industrial hardened beasts the size of a large audio amplifier installed in filter ventilated control cabs. They're good for about 10 years. |
@erik_squires
Sorry, but we wouldn't even be having this discussion if great class D amps had been around for ages.
Judging by the measurements I keep seeing, great class D amps still haven't arrived. Most of them have issues with ringing. All of them have ultrasonic noise that sits on the outputs like standing DC. Upwards of .8 volts of noise!?!?!? That's hifi, is it? A 1kHz sinewave looks horrible through those things. For all intents and purposes, it's made entirely out of high frequency noise. Some around here have said that the HF is easily filtered out. If so, why is JA always running low ass filters on his AP to get a sine the doesn't look like a scared porcupine?
I'm with George 110%. Class D isn't bringing anything new or great to hifi. Efficiency, yes, but they're technically inferior to class A. |
Look at the measurements! It's more than just the noise. I haven't seen one yet that doesn't ring. They all have a lot of horrible things going on and it's a process of compromise and trade off to get something that sorta measures ok and doesn't sound simply bad.
What's more, I reject the wholesale dismissal of measurements. It's a very sad state of affairs that most listeners are clueless about what's in the boxes they buy. "Just buy what you like listening to" is how Bose got so big. Most people LIKE compressed dynamics and overblown bass. Is that what hifi is now? That is why Bose has used class D amps for ages now. That's what class D amps do very well.
Measurements are extremely useful for estimating what an amplifier will do with a pair of speakers. They are good indicators of how the thing will sound. How are you supposed to "buy what you like" if you don't even know what it is you like? "I like amps with ungodly noise and ringing" said nobody ever. "I like quiet, simple, linear, wide bandwidth amps with sexy negative phase second harmonics" says this guy. Which class D does that? I'm not interested in anything else. |
@erik_squires
Go to Stereophile and look at the last 5 or 6 class D amps they've measured. Unless JA applies a low pass filter on the AP the output signals look horrendous. Even with the low pass filter, they all have ringing. I was just looking at the Below Canto Black and that thing isn't horrible, but .05% of garbage distortion at 200mW!?!?!? That's ridiculous.
It's pretty widely accepted that ,01% is where most people start hearing distortion characteristics; most definitely if it's high order. Low order characteristics in the .0X range large defiine the sound of an amp. |
I'm not at all befuddled by George. He's looking at amps that overshoot and have horrible low level distortion characteristics, just like I am, and is rightly calling that a bunch of crap. JA wouldn't call any other kind of amp "good" if it had that much overshoot or low level distortion apart from some tube amps. I read the review on that Cary with the 1% distortion at 3 watts. That was a VERY forgiving review in my estimation, and the listening evaluation made it abundantly clear that the thing was most definitely acting as a tone control of sorts and departed wildly from neutral. It's ONLY saving grace is that it was entirely second order distortion. That's not what class D amps are doing. They're making a complicated .05% distortion at a fraction of a watt.
The nature of the market place makes it obvious that true high fidelity isn't what sells. Nobody is standing in the isle of a Best Buy listening for harshness on that $400 pair of Klipsch. Most people LIKE punchy, sterile sound with insanely overblown bass that sounds extremely loud on account of the distortion.
I'm much more concerned with .05% distortion at .5 watts than I am with .005% at 150 watts. For my listening, nothing going on at those levels of power make any difference. Unless you're listening to 82dB/W/M speakers at 20 feet, those levels of power don't make any difference to you either. |
20kHz is hard for any amp to do, much less well??? http://www.firstwatt.com/pdf/prod_f5_man.pdfLook at page 17 where the FW F5 absolutely nails a 200KHz square wave at 1 watt. A lot of amps are designed to roll off at or past 20KHz more for purposes of self preservation than technical limitations. You’re always going to have that guy who wants to try Litz wire ICs and create a high power oscillator. @erik I don’t recall the amp model, but Pass installed a pot on the front of the thing to dial in the distortion character ranging from a few hundredths % positive phase second harmonic to a few hundredths negative phase. Anybody looking for proof that .0X% distortion does make an audible difference, there it is. The exact same effect can be achieved with the F5 and I’ve done it. Nobody is going to tell me .05% means nothing because I can hear it and I’m definitely not the only one. Nobody is going to tell me noisy, overshooting, sloppy amps are great when I'm looking at measurements like the F5 and hundreds of other well regarded amps which have been designed by guys correlating measurements with perceptions through thousands of hours of listening trials by skilled listeners. Sorry folks, but I’m very unimpressed with an amp that has 75mV of overshoot on a 1V 1KHz square wave. 800mV of high frequency noise on the outputs isn’t real sexy either. I’m not expecting every amp to drive 200 watts DC into a 1 ohm load, make text book square waves at 200KHz, roll off 3dB at 1MHz, have less than 60mV of noise on the outputs, or make .002% 3rd order distortion, but it’s nice. |
@tweak1
That's exactly what I'm saying. Above 1 KHz the power requirements of the amp drop off sharply and the sound quality is entirely dominated by the distortion character below 2 watts.
@atmasphere
Nelson specifically warns against using Litz wire with the F5 without putting some caps across the feedback resistors to truncate the high frequency response. He's also said that he makes it a point to limit the high frequency extension of his amps after watching one with F5-like capabilities release the magic smoke on stage when a demonstrator took hold of an improperly grounded volume control box and formed an oscillator. |
