Class T amp for 23$, a revolution?
- Thread starter Squeak
- Start date
Because it's 2300$ worth of sound quality for 23$.Guden Oden said:
You can get expensive valve amps that has the same or lower output. What matters is the sensitivity of the speakers. "Modern" speakers are very inefficient, both with the power they need and the way they radiate the sound.
6W is enough for normal to pretty loud listening if you have the right speakers.
PC-Engine said:
There are a few factors that make this unfeasible:
1. Line out/line in voltage levels are NOT standardized as xxx claims. It can _typically_ be 0.4Vrms-1Vrms. Now, say that the difference is just 6dB, that is still well enough to drive the power stage into overdrive(since 6dB is double the peak voltage) it can mean the difference between 12V peak or 24V peak on the output terminals.
2. Since manufacturers know this they have dynamic headroom in the pre-amplifier, besides even with low recordings you want to be able to use the entire dynamics that the amplifier provides. Ergo, ALL amplifiers I've ever seen _can_ be driven into overdrive.
3. We can have discrete separate pre-amps and power amps. Of course, the power amp can have just about any gain and wattage.
4. Speakers don't have a fix impedance(quoted nominal impedance is just the average), but rather it is a function of frequency(with argument, i.e. phase load). It's not uncommon for an 8Ohm speaker to dip down to 2-3Ohm, and peak at >50Ohm over the frequency range.
Example:
A _really_ nasty one:
All these considerations make xxx's never-overdrive amp a pipe dream.
_xxx_ said:
Sheesh...
Your whole premise for this argument has been the axiomic suggestion that speakers should never be matched with an overpowered amplifier which is simply *not* the accepted thinking throughout most of the audio profession.
I just don't get how you can patronisingly say "Where did you get this crap? Whoever told you that deserves a good asskick and a muzzle" when you claim to know a lot about the subject area. The nature of MPI's statement should have been incredibly obvious and not worthy of a trivial debate. I'm not an expert either, but it's a very basic idea.
He knows as much as you do, obviously. Reading some stuff and misinterpreting/twisting it is just sensless.
For example, speakers impendance ratings. It's NOT the approximate value, but the real Ohm value when you have it running on DC, so the coil and capacitor are eliminated.
But there is no sense to discuss this further, think/believe whatever you want...
For example, speakers impendance ratings. It's NOT the approximate value, but the real Ohm value when you have it running on DC, so the coil and capacitor are eliminated.
But there is no sense to discuss this further, think/believe whatever you want...
arjan de lumens
Veteran
In the field of electronics, there is a rather sharp delineation between the terms 'resistance' and 'impedance': 'impedance' denotes the reaction of a circuit when AC voltage/current is applied to it, while 'resistance' denotes the circuit's reaction at DC.
_xxx_ said:
I've seen both... Nevertheless, you don't run the speakers in DC but with a signal(AC) so it's pretty much moot anyway. The seen impedance by the amplifier are as above, i.e. uneven and unpredictable.
Having seen your debate "technique" in the "Tweel" thread I think I see a pattern here...
More like,
xxx:
Tweel! Im-POSSIBLE!!
Others:
Um, ok... why?
xxx:
HA-HAA!!! Don't you see how RIDICULOUS it is!!! It will NEVER work!!! Preposterous!!!
others:
Well... It looks kinda odd, but, well, it's michelin... Surely it can't be quite so ridiculous an idea as you claim... Please explain?
xxx:
HA-HAA!!! I know my shit, I've worked with cars for ____ years!
others:
ooooook... moving on.
Now replace Tweel with Class D amps.
Only difference though, there's proven hifi Class D amps on the market as we speak(did you even look at _one_ of the links above?). But no, they're "impossible". And nobody has ever had their speakers damaged by amplifier clipping.
xxx:
Tweel! Im-POSSIBLE!!
Others:
Um, ok... why?
xxx:
HA-HAA!!! Don't you see how RIDICULOUS it is!!! It will NEVER work!!! Preposterous!!!
others:
Well... It looks kinda odd, but, well, it's michelin... Surely it can't be quite so ridiculous an idea as you claim... Please explain?
xxx:
HA-HAA!!! I know my shit, I've worked with cars for ____ years!
others:
ooooook... moving on.
Now replace Tweel with Class D amps.
Only difference though, there's proven hifi Class D amps on the market as we speak(did you even look at _one_ of the links above?). But no, they're "impossible". And nobody has ever had their speakers damaged by amplifier clipping.
Maybe you should try looking at those links yourself instead of ranting. You're comparing apples to oranges.
From the first post:
1. Class T != Class D
2. Had you tried reading it, you'd see that the amp actually produces only 6W on 8 Ohm or 9W on 4 Ohm, not 15W as stated
2. Class D per definition IS NOT HI-FI, as good as you may build it
3. Speakers' impendance is actually the resistance value while NOT CONNECTED to anything, used just as a reference. ANY kind of load has a variable impendance over frequency, but that's got nothing to do with the resistance rating
4. Many manufacturers choose to build amps with overpowered preamps in order to be able to sell it as something with more "value". It's still just bad design, or a design decision driven by marketing. Like "people will rather buy an amplifier with 2x100W than an amp with 2x50W for the same price", even if it distorts above 70% of the max volume setting
5. No way in hell should it ever be the case that the speakers are weaker than the amplifier - they'll get blown out if you exceed their ratings
6. An underpowered amp will kill speakers only because when the output stage burns through, it takes the speakers with it. If you overdrive the amp that much that it goes into clipping, it'll eventually get too hot and burn through. That's a bad design decision, not because the amp is underpowered but because it's possible to overdrive it too much
7. Many hi-fi amps allow for a slight distortion in order to "use full dynamic range" (whatever that should be, I guess you mean using the full potential of the amp even if you have a signal which is not powerfull enough to reach the max possible output power), but WITHOUT GOING INTO CLIPPING. It's not the same thing
Been there, done that. So what are you trying to prove? That many manufacturers choose to lie in order to sell more stuff? We all know that, I think.
From the first post:
1. Class T != Class D
2. Had you tried reading it, you'd see that the amp actually produces only 6W on 8 Ohm or 9W on 4 Ohm, not 15W as stated
2. Class D per definition IS NOT HI-FI, as good as you may build it
3. Speakers' impendance is actually the resistance value while NOT CONNECTED to anything, used just as a reference. ANY kind of load has a variable impendance over frequency, but that's got nothing to do with the resistance rating
4. Many manufacturers choose to build amps with overpowered preamps in order to be able to sell it as something with more "value". It's still just bad design, or a design decision driven by marketing. Like "people will rather buy an amplifier with 2x100W than an amp with 2x50W for the same price", even if it distorts above 70% of the max volume setting
5. No way in hell should it ever be the case that the speakers are weaker than the amplifier - they'll get blown out if you exceed their ratings
6. An underpowered amp will kill speakers only because when the output stage burns through, it takes the speakers with it. If you overdrive the amp that much that it goes into clipping, it'll eventually get too hot and burn through. That's a bad design decision, not because the amp is underpowered but because it's possible to overdrive it too much
7. Many hi-fi amps allow for a slight distortion in order to "use full dynamic range" (whatever that should be, I guess you mean using the full potential of the amp even if you have a signal which is not powerfull enough to reach the max possible output power), but WITHOUT GOING INTO CLIPPING. It's not the same thing
Been there, done that. So what are you trying to prove? That many manufacturers choose to lie in order to sell more stuff? We all know that, I think.
And: in case of tweel, it was just my opinion. In this case, I know these things exactly. I designed and built at least 100 amps in my life. With bipolar transistors, tubes, OP-amps, you name it.
EDIT: Class D amps as well, BTW. But not for audio - PWM is used for pretty much every cooler and pump in a car, so I have to deal with that stuff daily.
EDIT: Class D amps as well, BTW. But not for audio - PWM is used for pretty much every cooler and pump in a car, so I have to deal with that stuff daily.
Yeah, this is tiresome, as we're obviously not seeing eye to eye on this, it's going to be my last post. Also weeding out the rehashed bits.
Haven't even mentioned it.
Pure BS. The definition of hi-fi is what the sound waves ultimately coming out of the speakers sound like compared to the recording. The bits in between and what technology used is ultimately irrelevant.
I can see where you're coming from with building switching motor control amps and everything, and the old classic Class D amps for that is certainly not suitable for that. However, what you've missed is that they have evolved beyond that with very low harmonic distortion and noise due to pretty clever compensating circuitry for the modulation. The improvement from the "classic" Class D output stages is simply enormous(many orders of magnitude). Your dismissiveness without even looking at the progress is just bullheaded stubborness. :? Oh, well...
Power rating in _good_ gear is always given within a certain distortion range, obviously within the linear range. And good gear still clips.
Sounds lika a bit of a conspiracy theory to me... you mean even the high-end gear inflate their numbers for no apparent purpose(since a few extra W hardly would improve their multi-$ pricetags)? That's just weird.
The fact is just simply that it is way 99% of commercial(and non-commercial) amplifiers are built. Clipping _is_ very much an issue.
If you have designed _your_ amplifiers so the pre-amp never can supply a signal(given your supposed "standard" line in-level) so together with the gain of the power stage the output signal never can exceed the rails, well... that's just extremely conservative. And even then you still have the very unpredictable impedance load to consider, and design even more conservatively...
EOD
_xxx_ said:
Haven't even mentioned it.
Pure BS. The definition of hi-fi is what the sound waves ultimately coming out of the speakers sound like compared to the recording. The bits in between and what technology used is ultimately irrelevant.
I can see where you're coming from with building switching motor control amps and everything, and the old classic Class D amps for that is certainly not suitable for that. However, what you've missed is that they have evolved beyond that with very low harmonic distortion and noise due to pretty clever compensating circuitry for the modulation. The improvement from the "classic" Class D output stages is simply enormous(many orders of magnitude). Your dismissiveness without even looking at the progress is just bullheaded stubborness. :? Oh, well...
Power rating in _good_ gear is always given within a certain distortion range, obviously within the linear range. And good gear still clips.
Sounds lika a bit of a conspiracy theory to me... you mean even the high-end gear inflate their numbers for no apparent purpose(since a few extra W hardly would improve their multi-$ pricetags)? That's just weird.
The fact is just simply that it is way 99% of commercial(and non-commercial) amplifiers are built. Clipping _is_ very much an issue.
If you have designed _your_ amplifiers so the pre-amp never can supply a signal(given your supposed "standard" line in-level) so together with the gain of the power stage the output signal never can exceed the rails, well... that's just extremely conservative. And even then you still have the very unpredictable impedance load to consider, and design even more conservatively...
EOD
IF you are *really* worried about clipping then you can scope (using a o-scope) your input signal and find its maximum unclipped level and then scope the amplifier output and then you know you won't be clipping if you play below that level.
Dynamic range is the difference between the quietest sound and the loudest sound.
For example, in music you could have a ultra quiet slide down a guitar string and then a huge peak from a drup hit. In a film you could have you a soft russle of leaves and then a powerfull explosion.
Speakers don't "Blow out" the usually fail from these two reasons:
1) The speakers excursion goes right past its mechanical x-max/xmag limits, damage like: cone warping, suspention tearing/warping, spider tearing, cone parting from spider, tinsel lead snap and voice coil pulled out of gap/allignment can occur.
2) Thermal limits are passed and the voice coil/s or tinsel leads are burnt to a crisp.
When amplifiers are driven into clipping they can produce over twice the regular amount of power, hence a 50 WRMS amp often can deliver twice that in clipped signal.
Some of the best sounding subwoofers in the world use Class D amplification.....
Dynamic range is the difference between the quietest sound and the loudest sound.
For example, in music you could have a ultra quiet slide down a guitar string and then a huge peak from a drup hit. In a film you could have you a soft russle of leaves and then a powerfull explosion.
Speakers don't "Blow out" the usually fail from these two reasons:
1) The speakers excursion goes right past its mechanical x-max/xmag limits, damage like: cone warping, suspention tearing/warping, spider tearing, cone parting from spider, tinsel lead snap and voice coil pulled out of gap/allignment can occur.
2) Thermal limits are passed and the voice coil/s or tinsel leads are burnt to a crisp.
When amplifiers are driven into clipping they can produce over twice the regular amount of power, hence a 50 WRMS amp often can deliver twice that in clipped signal.
Some of the best sounding subwoofers in the world use Class D amplification.....
MPI:
Didn't want to attack your opinion, it's just that we have a totally different phillosophy - it's a matter of choice. In my vocabulary, hi-fi doesn't clip nor does it get converted through PWM and a low-pass, that's all.
I always build my stuff with reliability in mind. As micb said, scope is my best friend. I blew quite a few amps myself in the past and was fed up with badly designed amps, that's where I started building my own.
I also make it possible for the signal to clip (for the guitar amps it's even a must) but there the clipping will happen in the preamp section and _not_ in the power amp. That's a small but very important difference. You reach the same goals as when overdriving the power amp, but there's no risk of damaging anything if dimensioned properly.
With tubes there's no clipping in the same sense like with transistors anyway, but a different kind of distortion (the signal is being severely changed because of the op-point reaching the non-linear area, but the characteristic is such that it doesn't all get cut off like with trannies). That's why people like tube sound that much - you can hardly ever get a rectangular signal at the output.
Didn't want to attack your opinion, it's just that we have a totally different phillosophy - it's a matter of choice. In my vocabulary, hi-fi doesn't clip nor does it get converted through PWM and a low-pass, that's all.
I always build my stuff with reliability in mind. As micb said, scope is my best friend. I blew quite a few amps myself in the past and was fed up with badly designed amps, that's where I started building my own.
I also make it possible for the signal to clip (for the guitar amps it's even a must) but there the clipping will happen in the preamp section and _not_ in the power amp. That's a small but very important difference. You reach the same goals as when overdriving the power amp, but there's no risk of damaging anything if dimensioned properly.
With tubes there's no clipping in the same sense like with transistors anyway, but a different kind of distortion (the signal is being severely changed because of the op-point reaching the non-linear area, but the characteristic is such that it doesn't all get cut off like with trannies). That's why people like tube sound that much - you can hardly ever get a rectangular signal at the output.
