How good are Class D amplifiers now?

infullview

Super Member
I was looking through the offerings at Parts Express, and I see all these super high powered Class D amps for almost no money. I know these use high frequency switching technology to produce a lot of power with very little heat. I was wondering how far the technology has come and if anyone uses it in audiophile circles (other than for sub woofers).
 
People seem to refer to them by the chip inside lately. Try searching for tpa3116.

I also find it much easier to do a Google site search rather than trying to use the built in forum search.

Ex:
"Tpa3116 site:audiokarma.org"
 
I looked up one of the Class D components made by TI which delivered up to 25 W/ch and found that typical THD was around 1% at 10 KHz - not stellar by any standard. Also, distortion varied widely as a function of frequency which doesn't surprise me as this is probably due to alignment of harmonic intervals on the switching frequency of the amp. I can't imagine anyone using these things for high end audio.
 
People seem to refer to them by the chip inside lately. Try searching for tpa3116.

I also find it much easier to do a Google site search rather than trying to use the built in forum search.

Ex:
"Tpa3116 site:audiokarma.org"
Thanks for the suggestion to use google. That's a good idea and how I found the site in the first place. Should have thought of that myself :) It turns out that it was the tpa3116 chip I was looking at on the TI website.
 
Damn, wonder how my ST70 sounds great?

:beerchug:
Not trying to bash anyone's equipment - the rating on an analog device generally is a reference at max rated power. What set these devices apart is that distortion is not a function of power but of frequency.
 
Not trying to bash anyone's equipment - the rating on an analog device generally is a reference at max rated power. What set these devices apart is that distortion is not a function of power but of frequency.

Didn't take it that way, I typically post lame attempts at humor as some on here take this stereo stuff and themselves way to serious. But I will add that I was a numbers chaser in my youth until I figured out they make a good guide but really are not the be all end all. I am curious how this progresses, I have been considering one for the idiot box.
 
I tend to think lower THD is better, but Pete Millett measured his starving student headphone amp at something like 0.3% at full output and it sounds stellar to me.:dunno:

What does distortion in a class D amp look like? Does it break down into the common 2nd, 3rd, 4th etc order harmonics that you'd expect from a class AB amp or is it something else?
 
I looked up one of the Class D components made by TI which delivered up to 25 W/ch and found that typical THD was around 1% at 10 KHz - not stellar by any standard. Also, distortion varied widely as a function of frequency which doesn't surprise me as this is probably due to alignment of harmonic intervals on the switching frequency of the amp. I can't imagine anyone using these things for high end audio.

1% THD isn't typical...
You might want to take a look at the various THD related graphs shown in this pdf....
They start in earnest on page 10.
http://www.ti.com/lit/ds/symlink/tpa3255.pdf
 
Last edited:
I tend to think lower THD is better, but Pete Millett measured his starving student headphone amp at something like 0.3% at full output and it sounds stellar to me.:dunno:

What does distortion in a class D amp look like? Does it break down into the common 2nd, 3rd, 4th etc order harmonics that you'd expect from a class AB amp or is it something else?
I'd say .3 sounds a lot better than 1 percent! And that's a good question... what does this distortion look like compared to analog? And what is this rating they have on class D where it says 1%+N (for noise). What noise?
 
Didn't take it that way, I typically post lame attempts at humor as some on here take this stereo stuff and themselves way to serious. But I will add that I was a numbers chaser in my youth until I figured out they make a good guide but really are not the be all end all. I am curious how this progresses, I have been considering one for the idiot box.
I was also look at trying one of these out. That's why I started this thread to get more info and find out other practical measures other people have encountered when using them.
 
Just my .2 cents being a electronics tech since 1985. Chips are cheap, I would think the rest of the "box" would determine the how "good" it is or could be. That 1% THD chip in a better ckt would probably sound better than the .3% in a lesser ckt. IMO you can't just chase numbers for one component. But hey I am often wrong :rockon:

:beerchug:

Edit: thinking these are like DAC chips, its all about the rest of the box... IMHO
 
Last edited:
I have to admit, the specs on this chip look a lot better than the tpa3116. Do you know anyone that's used it?

I've been running Texas Instruments TPA 3250/1/5 evaluation modules almost exclusively for the past 8-10 months.
They're low distortion, very dynamic, precise...and to my ears, well balanced. Also, dead quiet. Noise floor is -140db.
They're high current amps. Power ratings at 8ohms are very deceiving....which creates some confusion.
While output figures based on 10% THD are ridiculous, in the real world, no one is going to push them that far....unless they're used in a PA system [which I assume is part of the market].

I've said it before, and I'll likely say it again...to the extent that these discussions of Class D revolve around $30-$50 low-powered amps, it seriously hampers the conversation.

I'd post a link to specs for the Hypex Ncore NC400 modules, but I'm pretty sure they're downloadable only.
 
Just my .2 cents being a electronics tech since 1985. Chips are cheap, I would think the rest of the "box" would determine the how "good" it is or could be. That 1% THD chip in a better ckt would probably sound better than the .3% in a lesser ckt. IMO you can't just chase numbers for one component. But hey I am often wrong :rockon:

:beerchug:

Edit: thinking these are like DAC chips, its all about the rest of the box... IMHO
I also used to work as a tech for Data General during the 80's on switching power supplies (and yes they used PCM on those bad boys too!) I still want to know what this so-called "noise" spec is that's on the data sheets. I don't see any spec for IM on any data sheets (may be unmeasurable?). Just looked at the spec sheet again for more answers and it looks like you have the LC snubber circuit then it goes through a cap to the speaker. I'm not in love with the idea of using an electrolytic to pass audio. Much better to either be direct (class AB), or inductively coupled (tube) in my book.
 
Back
Top Bottom