Starting to have second thoughts on Class D amps - Page 5 - AVS Forum
Forum Jump: 
 
Thread Tools
Old 01-31-2014, 01:24 PM
AVS Special Member
 
notnyt's Avatar
 
Join Date: Dec 2008
Location: Long Island, NY
Posts: 4,616
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 345 Post(s)
Liked: 522
Quote:
Originally Posted by WVSyd View Post

Notnyt:
So because your opaque pop reference falls flat - then your direct insults starts.
How many times does one have to post to be granted such Special Member insult privileges...

It wasn't an insult, or meant to be insulting in any way, it was a joke. Lighten up.

The original joke, which I'm fairly certain most here got, was just about getting popcorn to watch the main feature (everyone arguing). It's kind of a common internet joke that most are aware of.
notnyt is online now  
Sponsored Links
Advertisement
 
Old 01-31-2014, 01:26 PM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,475
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 277 Post(s)
Liked: 103
Quote:
Originally Posted by notnyt View Post

Quote:
Originally Posted by WVSyd View Post

Notnyt:
So because your opaque pop reference falls flat - then your direct insults starts.
How many times does one have to post to be granted such Special Member insult privileges...

It wasn't an insult, or meant to be insulting in any way, it was a joke. Lighten up.







biggrin.gif
notnyt likes this.
J_Palmer_Cass is offline  
Old 01-31-2014, 01:28 PM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 42
Ah Didn't see the jokey emoticon.
notnyt likes this.

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
Old 01-31-2014, 01:28 PM
AVS Special Member
 
notnyt's Avatar
 
Join Date: Dec 2008
Location: Long Island, NY
Posts: 4,616
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 345 Post(s)
Liked: 522
notnyt is online now  
Old 01-31-2014, 01:30 PM
Senior Member
 
WVSyd's Avatar
 
Join Date: Feb 2013
Posts: 333
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 42
Don't hold it against people for not being in the 18 - 49 demo
biggrin.gif

"Beware of Salesmen: They are the modern Svengali, immune to Science and Reality"
WVSyd is offline  
Old 01-31-2014, 01:33 PM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by WVSyd View Post

Don't hold it against people for not being in the 18 - 49 demo
biggrin.gif

I don't hold it against people for not being in the 18-49 demographic. I hold it against when they forget they aren't. Now get off my lawn! tongue.gif

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 01-31-2014, 01:33 PM
Administrator
 
Mike Lang's Avatar
 
Join Date: Nov 1999
Posts: 12,534
Mentioned: 14 Post(s)
Tagged: 0 Thread(s)
Quoted: 416 Post(s)
Liked: 357

Find your way back to the topic guys...


Mike Lang
Administrator
Please use the report post button to alert staff to problematic posts. Never quote or respond to them yourself.
Join the AVS Club and help support the site. Help Support AVS Forum Sponsors.
Mike Lang is offline  
Old 01-31-2014, 01:35 PM
AVS Special Member
 
notnyt's Avatar
 
Join Date: Dec 2008
Location: Long Island, NY
Posts: 4,616
Mentioned: 8 Post(s)
Tagged: 0 Thread(s)
Quoted: 345 Post(s)
Liked: 522
Anyway, back on topic and some fuel for the fire. LG clearly thinks their design is not class H, as they graph class H along with class TD. I'd tend to side with LG here.
notnyt is online now  
Old 01-31-2014, 01:46 PM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by notnyt View Post

Anyway, back on topic and some fuel for the fire. LG clearly thinks their design is not class H, as they graph class H along with class TD. I'd tend to side with LG here.

Graphs are useless without source of data. I can make a class D amp look less efficient than class A with a flick of a few keys. My guess is modern day class D is more efficient than class "TD". Remember when that document was made.

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 01-31-2014, 03:44 PM
Advanced Member
 
MBentz's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 950
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 41
Quote:
Originally Posted by Trepidati0n View Post

As for efficiency. It matters. Efficiency for amplifiers is THEE # cost sliding variable. There is a measurable and verifiable financial benefit for every watt you don't burn up as heat in the box. In a competitive market..it matters.

The true cost associated with heat in the box (to which I believe you're referring) is the condition where the most power is being dissipated in the box. At low output powers, a low efficiency often results in less total power being dissipated than high efficiency at high output power.

For example, 91% efficiency with 1000W output means 100 Watts are being dissipated inside the amplifier. If that efficiency drops to 20% at 10W output, then you're looking at 50W being dissipated inside the amplifier at lower output levels. The cooling design for such an amplifier should be capable of dissipating 100W, and is unaffected by the lower efficiency at lower output levels.

-Mike Bentz
~It's all about compromise~
MBentz is offline  
Old 01-31-2014, 04:06 PM
Advanced Member
 
MBentz's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 950
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 41
Quote:
Originally Posted by notnyt View Post

Anyway, back on topic and some fuel for the fire. LG clearly thinks their design is not class H, as they graph class H along with class TD. I'd tend to side with LG here.

That reasoning is absurd. Of course they're going to use their marketing buzz word to differentiate their specific product on a chart like that.

There is no denying that the marketing definition of "TD" fits within the technical definition of "Class H". I have no doubt that LG is modulating their rail voltage differently than your typical pro audio Class H amplifier - There are all sorts of ways to implement Class H: some are better than others. Clearly LG feels their implementation is worthy of a special marketing buzzword. An engineer would realize that they're adding cost/complexity for something that would have marginal benefit over the old school Class H.

Old School Class H: AC/DC Converter modulated by audio -> AB Output Stage
LG's Class H (TD): AC/DC Converter -> Switching Push/Pull Buck Converter (Class D) modulated by audio -> AB Output Stage

-Mike Bentz
~It's all about compromise~
MBentz is offline  
Old 01-31-2014, 04:25 PM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by MBentz View Post

The true cost associated with heat in the box (to which I believe you're referring) is the condition where the most power is being dissipated in the box. At low output powers, a low efficiency often results in less total power being dissipated than high efficiency at high output power.

For example, 91% efficiency with 1000W output means 100 Watts are being dissipated inside the amplifier. If that efficiency drops to 20% at 10W output, then you're looking at 50W being dissipated inside the amplifier at lower output levels. The cooling design for such an amplifier should be capable of dissipating 100W, and is unaffected by the lower efficiency at lower output levels.

Yes and no. You are right that if it can handle the 100W it can handle the 50W. The problem is average temperature and temperature cycling is what kills silicon. It is an integrated and non-linear failure mechanism and it takes time. It is a wonderfully rich and fun topic. But to simplify it this way.

Here is a fair analogy. Numbers have been scaled to keep disparity close. Imagine you have a can full pennies 1000 pennies. When you are dissipating 100W you take you 100 pennies per hour. When you are dissipating 50W you take out 10 pennies per hour. When you dissipating 25W you take out 1 penny per hour. When you run out of pennies, the amplifier dies. Now lets say you spend 1 minutes @ 100W of dissipation, 5 mintues @ 25W, and 54 minutes @ 25W. This means you using 3.4 pennies so you amplifier will die in 294 hours. Now what if I could change that 50W "dissipation" down to 25W by being more efficient at the non-100%-power case. I'm now only using 2.65 pennies per hour or now I have an amplifier of 377 hours. I increased the life of my amp by 28% by making the bottom end more efficient versus making the top end more efficient.

The fact of the matter is, the scaling is much more intense than that. In some cases you get 20x more life for a 10 degree C reduction in junction temperature! So in short, sometimes a lot of amps have a ton more silicon more than they need not to produce peak power but to actually be reliable.

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 01-31-2014, 04:49 PM
Advanced Member
 
MBentz's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 950
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 41
Quote:
Originally Posted by Trepidati0n View Post

Yes and no. You are right that if it can handle the 100W it can handle the 50W. The problem is average temperature and temperature cycling is what kills silicon. It is an integrated and non-linear failure mechanism and it takes time. It is a wonderfully rich and fun topic. But to simplify it this way.

Here is a fair analogy. Numbers have been scaled to keep disparity close. Imagine you have a can full pennies 1000 pennies. When you are dissipating 100W you take you 100 pennies per hour. When you are dissipating 50W you take out 10 pennies per hour. When you dissipating 25W you take out 1 penny per hour. When you run out of pennies, the amplifier dies. Now lets say you spend 1 minutes @ 100W of dissipation, 5 mintues @ 25W, and 54 minutes @ 25W. This means you using 3.4 pennies so you amplifier will die in 294 hours. Now what if I could change that 50W "dissipation" down to 25W by being more efficient at the non-100%-power case. I'm now only using 2.65 pennies per hour or now I have an amplifier of 377 hours. I increased the life of my amp by 28% by making the bottom end more efficient versus making the top end more efficient.

The fact of the matter is, the scaling is much more intense than that. In some cases you get 20x more life for a 10 degree C reduction in junction temperature! So in short, sometimes a lot of amps have a ton more silicon more than they need not to produce peak power but to actually be reliable.

Interesting perspective on the average temperature in the box. Getting into the crest factor of music really increases the complexity - and then you need to look at how much power each application will actually be using...

I think your point has way more merit when talking about capacitor lifetimes rather than the actual silicon though. Sure, silicon wears out faster at higher temperatures, but you gotta start with the actual expected lifetime of the parts first. My experience is that the caps are the weak link - and they are even more greatly affected by the ambient temperature in the box.

Looking at the total system cost: do you improve the quality / quantity of your capacitors by a few dollars, or do you spend a few dollars improving your amplifier's efficiency or cooling ability?

-Mike Bentz
~It's all about compromise~
MBentz is offline  
Old 02-01-2014, 02:00 PM
AVS Special Member
 
J_Palmer_Cass's Avatar
 
Join Date: May 2005
Posts: 6,475
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 277 Post(s)
Liked: 103
Quote:
Originally Posted by MBentz View Post

Quote:
Originally Posted by notnyt View Post

Anyway, back on topic and some fuel for the fire. LG clearly thinks their design is not class H, as they graph class H along with class TD. I'd tend to side with LG here.

That reasoning is absurd. Of course they're going to use their marketing buzz word to differentiate their specific product on a chart like that.

There is no denying that the marketing definition of "TD" fits within the technical definition of "Class H". I have no doubt that LG is modulating their rail voltage differently than your typical pro audio Class H amplifier - There are all sorts of ways to implement Class H: some are better than others. Clearly LG feels their implementation is worthy of a special marketing buzzword. An engineer would realize that they're adding cost/complexity for something that would have marginal benefit over the old school Class H.

Old School Class H: AC/DC Converter modulated by audio -> AB Output Stage
LG's Class H (TD): AC/DC Converter -> Switching Push/Pull Buck Converter (Class D) modulated by audio -> AB Output Stage



LG uses a Tracking Class D type amplifier to regulate/control the amplifier power source rail voltage, so I assume that is what they mean with use of the Class TD term.


Still, the basic design is Class H.


J_Palmer_Cass is offline  
Old 02-01-2014, 02:16 PM - Thread Starter
Advanced Member
 
Swolephile's Avatar
 
Join Date: Sep 2013
Posts: 542
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 22 Post(s)
Liked: 55
Thank you to everyone posting great info in this thread. It is helping many of us learn from different viewpoints on this topic.smile.gif
Swolephile is offline  
Old 02-02-2014, 05:06 AM
Senior Member
 
Antripodean's Avatar
 
Join Date: Apr 2009
Location: The Antripodes (OZ)
Posts: 328
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 12
+1 to the great posts

Special thanks to Bosso for his insight on these amps. I really appreciate you taking your time to explain the design of the LG amps.

explore the music
Antripodean is offline  
Old 02-03-2014, 08:25 AM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by MBentz View Post

Interesting perspective on the average temperature in the box. Getting into the crest factor of music really increases the complexity - and then you need to look at how much power each application will actually be using...

I think your point has way more merit when talking about capacitor lifetimes rather than the actual silicon though. Sure, silicon wears out faster at higher temperatures, but you gotta start with the actual expected lifetime of the parts first. My experience is that the caps are the weak link - and they are even more greatly affected by the ambient temperature in the box.

Looking at the total system cost: do you improve the quality / quantity of your capacitors by a few dollars, or do you spend a few dollars improving your amplifier's efficiency or cooling ability?

Hate to say this, but no. Heat has little to deal with why capacitors fail in audio amplifiers. The main reason they fail is because of the voltage impressed upon them. A capacitors "energy" is a function of voltage squared but the their life reduction is a power of 3.4. Therefore engineers will try and play a game of "voltage regulation vs life". If they get a bad batch then may god have mercy on your soul; remember the dell motherboard recall. eek.gif The reason why a capacitor failures seems to be common is because there are a lot of them. The reason why there are a lot of them is not because of current ripple but for voltage regulation. In older amplifiers with passive front ends your voltage regulation frequency is ~120 Hz (2x line cycle when rectified for a single phase system) . This means to keep a "stiff DC bus" you need a lot of capacitance. However, if you have an active front end your regulation frequency is nearly your switching frequency. This means the size of the tank you need goes down by nearly that ratio until you hit the limit of current ripple on the device size capability.

This is why class D amps are so damn light. They need a fraction of the capacitance to provide the DC bus voltage regulation and they need a fraction of silicon to provide the same life. Therefore they also need a fraction of metal for cooling. If you were a business that could increase your profits by a very large percentage for a few % reduction of sound quality, how could you not do it! This is the draw of class D. Its long term potential passes that of any linear output stage in terms of an engineering quality factor. The problem is that to do it right, it is an order of magnitude harder than a linear stage. This is why class D was such a cluster-DUCK a few decades ago. The technology and engineering was just not ready. Digital hand-held music players gave us a solid leap forward because they are all class D. It put a ton of engineers on the job and thus a lot more was learned. With that knowledge they were able to scale it back up to larger PA amps which was a "good thing".

Now...do I wish I could run 7x1000W class A amps...yeah...I do. Can't beat a class A amp for distortion minimization. But since that isn't reasonable I'll take the next best thing at a fraction of the price that will get me as close as I can get.

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 02-03-2014, 09:56 AM
Advanced Member
 
MBentz's Avatar
 
Join Date: Jun 2006
Location: Chicago
Posts: 950
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 24 Post(s)
Liked: 41
Quote:
Originally Posted by Trepidati0n View Post

Hate to say this, but no. Heat has little to deal with why capacitors fail in audio amplifiers.

You should read this:
http://www.lintronicstech.com/index%20pdf/reliability_of_capacitors_general.pdf

A rule of thumb that "us" engineers (yes, I'm a senior audio circuitry design engineer for Shure and calculate this frequently) is that lifetime of a capacitor doubles for every 10C reduction in temperature from its maximum rating - which you will even see mentioned in this paper. In practice, "we engineers" use more advanced techniques for calculating MTBF because we take into account several other factors and we want to make sure our products meet their specifications for the life of our products - which is typically longer than 15 years for Shure products.

Granted I don't design power amplifiers in a professional scenario, but I've never seen MTBF dominated by silicon in any of my designs. And even for the hobbyist amplifiers I've designed, the caps are always the limiting factor. If you go with the best transistor available for any given application, I would be extremely surprised to see it fail - especially in a well thought out topology.

Btw, your comments about the refresh rate of a capacitor, although true, is not the complete story. Your discharge rate and bandwidth of your power supply are far larger factors for dealing with power supply related sonic artifacts. You better have a crap ton of rail capacitance if the bandwidth of your power supply isn't much greater than your audio signal...


At the end of the day, I believe Class D should be the best amplifier topology when looking at the total system design....directly couple a good Class D amplifier to your loudspeaker voice coil and there's all sorts of tricks you could play that aren't as easy to do in the Class A realm.

-Mike Bentz
~It's all about compromise~
MBentz is offline  
Old 02-03-2014, 10:17 AM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
Quote:
Originally Posted by MBentz View Post

That reasoning is absurd. Of course they're going to use their marketing buzz word to differentiate their specific product on a chart like that.

There is no denying that the marketing definition of "TD" fits within the technical definition of "Class H". I have no doubt that LG is modulating their rail voltage differently than your typical pro audio Class H amplifier - There are all sorts of ways to implement Class H: some are better than others. Clearly LG feels their implementation is worthy of a special marketing buzzword. An engineer would realize that they're adding cost/complexity for something that would have marginal benefit over the old school Class H.

Old School Class H: AC/DC Converter modulated by audio -> AB Output Stage
LG's Class H (TD): AC/DC Converter -> Switching Push/Pull Buck Converter (Class D) modulated by audio -> AB Output Stage

Agree 100%. Every Class H pro amp differs slightly, but they are all basically implementations of Class H. Has Lab Gruppen come up with a superior flavor of Class H? I'm not qualiified to say, but their explanation is just high level marketing BS. I'm not sure it is radically different than QSC's PowerLight supplied Class H amps which is not a bad thing.
coctostan is offline  
Old 02-03-2014, 11:08 AM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by MBentz View Post

Granted I don't design power amplifiers in a professional scenario, but I've never seen MTBF dominated by silicon in any of my designs.

Amplifier is a good term in terms of understanding what is going on..but it makes me laugh a little. To me, any amplifier is nothing more than a power converter with a reference waveform. It is converting 50/60Hz AC into something else. The converters I work with are in the megawatt class (MW). In terms of topology..there isn't much difference between 10W and 10MW. More stuff to worry about at higher power...but at the end of the day there is no real difference. However, in terms of raw material cost, the cost of silicon starts taking a larger % of the total cost. Things like magnetics and capacitors scale really well in general. In short...a 10MVA transformer is not twice as much as a 5MVA transformer. In reality, it is a LOT better than that. Same goes for a 10VA inductor vs a 5VA inductor. However, silicon doesn't scale well. Reason being is the world manufactures a TON of it. Therefore cost optimization is near its peak already. Therefore if you have 100VA of silicon and want to go to 200Va of silicon, expect to pay nearly twice as much. In the case of a 1MVA converter...silicon is probably approaching 40% of the cost.

Now back to the topic here of audio power converters, it is a bit relative. I wouldn't expect in a class D that silicon as a MTBF driver anymore. It just isn't that big a piece of the pie. This is a GOOD THING! I wouldn't mind seeing a Pareto chart of a modern class D amp to see what sticks up. If in small amps I wouldn't expect the silicon to stick up to high..but I would expect it to be on the top ten. However, I do know for a fact that in old big amps it was capacitors and silicon. I read some old literature from a guy who worked on receivers and PA's systems in the eighties (I was just a teenage then). He stocked parts based upon those charts and they worked out pretty well for him in terms of managing his inventory.

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 02-03-2014, 11:16 AM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by coctostan View Post

Agree 100%. Every Class H pro amp differs slightly, but they are all basically implementations of Class H. Has Lab Gruppen come up with a superior flavor of Class H? I'm not qualiified to say, but their explanation is just high level marketing BS. I'm not sure it is radically different than QSC's PowerLight supplied Class H amps which is not a bad thing.

Half marketing BS and half truth. At the time class H rails were done a linear circuit. Good but not amazing. When LG came out basically using a buck converter to vary the rails...it was a cool thing. These days though...nothing special. But remember, saying something wasn't "clever" when you have the hindsight of a decade or more of time is having a little to much hubris IMO. I can still remember remember when the first IR remotes came out over the "sonic remotes". These days when people say "IR remote", it doesn't even phase us. Heck most of us are probably like why isn't it RF and network controllable?

My favorite though is the original snowboard (burton) and the history of it.

https://www.burton.com/on/demandware.store/Sites-Burton_US-Site/default/Company-History

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 02-05-2014, 06:41 AM
AVS Special Member
 
coctostan's Avatar
 
Join Date: Oct 2005
Location: Indy
Posts: 1,960
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 2 Post(s)
Liked: 129
I know it isn't a basic textbook class h and I have little doubt that it is an exceptional version of the topology.

I was just saying that marketing departments try to make it sound like they are doing something proprietary and revolutionary and that isn't true. LG isn't alone in this and it isn't a knock on them, but they also haven't reinvented the wheel.

If I had the budget for Lab amps I'd buy them no question. I didn't want to mess with clones and I don't need that much power from a single amp. I'm happy with QSCs at my price point.

BTW, I first snowboarded on something not far from that original Burton.
coctostan is offline  
Old 02-05-2014, 10:44 AM
AVS Special Member
 
jpmst3's Avatar
 
Join Date: Dec 2005
Location: Davidsville, PA
Posts: 8,262
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 48 Post(s)
Liked: 215
http://www.audioholics.com/amplifier-reviews/iq-audio-m300

In case anyone wants to read...

______________________________________________
Joe = LFE Addict
Joe's LOWARHORN build
Dual LMS-5400 Ultra Endtables build
jpmst3 is online now  
Old 02-05-2014, 01:56 PM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by jpmst3 View Post

http://www.audioholics.com/amplifier-reviews/iq-audio-m300

In case anyone wants to read...

I read that and looked inside. Anybody who buys this is paying a lot of money relative to the cost of the components. Not saying this is a "bad thing"..but realize the reality of it. Furthermore...some of their "wording" really boggles my mind which my guess comes from the marketers. The guy who probably "really designed this" probably doesn't put his face in public very often or is heavily coached. wink.gif I could say a whole bunch of things, but i'l leave it at this; this amp gets rid of all the "low voltage circuitry" (e.g. the plethora of op-amps and passives) and replaces it with a processor. This is the key take away. If you look at the current class D amps out there....a lot of analog circuitry on there. This makes it very hard to coordinate certain activities in relation to electronics. By converting both the input and measuring the output they can apply some very powerful controls laws (like ZVS & ZCS) which just don't fit classical methods.

But to call this "class D" is a little on the edge. It doesn't fit the classical "design". However, there is a "hard switching" event which does occur which does fit the spirit of class D amplifier design. I wouldn't be surprised if we see more of these amps hit mainstream in another 5-10 years at the same price point as current PA amps. It takes a while for big players to abandon old designs for new ones; more commonly known as we milked this cow for all it was worth...onto the new cow.

EDIT: Thanks for sharing smile.gif

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 02-05-2014, 03:42 PM
AVS Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 20,776
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 397 Post(s)
Liked: 183
Quote:
Originally Posted by Trepidati0n View Post

But to call this "class D" is a little on the edge. It doesn't fit the classical "design".

How so?

Class D means a switching output stage, which this has.

Noah
noah katz is online now  
Old 02-06-2014, 04:30 AM
Member
 
adio's Avatar
 
Join Date: Dec 2005
Location: baltimore, md
Posts: 138
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 0 Post(s)
Liked: 10
Quote:
Originally Posted by MBentz View Post

Btw, if you're going to compare class D to A/B, then you probably shouldn't compare the same wattage ratings. A 200W class D won't go much louder than a 50W A/B. It has to do with crest factor and rail voltage limitations.

Also, the quality of the power supply is arguably more important than A/B and yet most class D amps are worse off in that regard. Mixing is an incredibly difficult thing to avoid and doesn't show up in standard measurements...

Quote:
Originally Posted by arnyk View Post

Good question. As a matter of fact a 200 watt amp isn't that much louder than a 50 watt amp all things being equal. For openers, not that many people will exhaust the capabilities of the 50 watt amp. Then the difference is only 6 dB which is audible but is probably a lukewarm enough of a difference to not warm the cockles of many hearts. Remember that 10x the power is only 10 dB or about twice as loud and most of the times people with a quest for loudness want more than that!

That all said I am unaware of any evidence and I have no experiences that suggest to me that Class D watts are any different then Class AB watts in terms of raw loudness. IME 50 watts is 50 watts, +/- some of the load-dependent frequency response issues that some Class D amps have, but only some.

Both class AB amps and Class D amps benefit from the fact that the crest factor of music is so high. The Class AB amps being less efficient probably benefit more from the reduced heating and crunching of the power supply due to the higher crest factor or music. I don't know if that is the point or not from the wording of the post we are responding to.



glad this is getting/gotten cleared up. i read the bench mark tests for the HK AVR3700 at sound and vision and seenthat it only puts out 27.9 watt at 0.1% THD and 33.8 watts at 1% THD into seven channells. i said to myself 'whos' gonna buy that weak sh!t ??!!' but after reading these i guess 27.9watts is louder that what one would think.
adio is offline  
Old 02-06-2014, 08:03 AM
AVS Special Member
 
Trepidati0n's Avatar
 
Join Date: Nov 2003
Posts: 1,127
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 23 Post(s)
Liked: 93
Quote:
Originally Posted by noah katz View Post

How so?

Class D means a switching output stage, which this has.

I never said it didn't meet the definition of class D; I said it didn't meet the "classical definition". Here is a challenge for you, go out and get a hundred of audio amplifier design books that are dated 2010 or earlier (which is ~99% of them) and find me how many talk about using ZVS/ZCS for the amplifier stage in any amount of detail as a "class D". IEEE papers on this topic are still fairly recent by academic standards and PhD papers are still being published. As a FYI, universities are the new "research think tanks". Here is another test, go to 100 electrical engineers and have them draw the output stage for a "class D" amplifier and explain how it works. The overwhelming majority will draw two switches in totem pole or DC servo style and use a fixed frequency PWM to describe it.

I must be guilty because people say I am guilty because they chose to call me guilty because they refuse to see the truth. Much easier to be part of the mob..
Trepidati0n is offline  
Old 02-06-2014, 03:25 PM
AVS Addicted Member
 
noah katz's Avatar
 
Join Date: Apr 1999
Location: Mountain View, CA USA
Posts: 20,776
Mentioned: 1 Post(s)
Tagged: 0 Thread(s)
Quoted: 397 Post(s)
Liked: 183
Quote:
Originally Posted by adio View Post

glad this is getting/gotten cleared up. i read the bench mark tests for the HK AVR3700 at sound and vision and seenthat it only puts out 27.9 watt at 0.1% THD and 33.8 watts at 1% THD into seven channells. i said to myself 'whos' gonna buy that weak sh!t ??!!' but after reading these i guess 27.9watts is louder that what one would think.

Results for continuous output testing of many receivers/amps isn't relevant because the current-limiting is invoked and wouldn't occur in real-world use.
Quote:
Originally Posted by Trepidati0n View Post

I never said it didn't meet the definition of class D; I said it didn't meet the "classical definition".

rolleyes.gif One out of two is good enough for me.

Noah
noah katz is online now  
 
Thread Tools


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off