2. amps-preamps
  3. 151675-68726

Time to buy a class D amp?


Will some new class D amplifiers outperforming the current ones appear soon

(the newest ones i know were released a  few years ago)?

Class D amps attract me as I consider them the most ecological ones with obvious non-auditionable benefits.

I have no doubts that they posses the maximum ratio performance/sound quality among the amplifiers of all classes.

At the same time, the sound quality the class D amplifiers that I have auditioned produce, although is quite good,

but not yet ideal (for my taste).


I use PS Audio Stellar S300 amp with PS audio Gain Cell pre/DAC with Thiel CS 3.6 speakers in one of my systems.

The sound is ok (deep bass, clear soundstage) but not perfect (a bit bright and somehow dry, lacking warmness which might be more or less ok for rock but not for jazz music).

I wonder if there are softer sounding class D amps with the same or better details and resolution. Considering two reasonable (as to the budget) choices for test, Red Dragon S500 and Digital Audio Company's

Cherry  2 (or Maraschino monoblocks), did anybody compare these two?



128x128niodari
georgehifi's avatar
georgehifi
10,312 posts
 
GaN based Class-D in that they are going for bench specs as opposed to sonic excellence.
I worry when a manufacturer says this, (trust me it sound better)

We are able to achieve better bench specs than any other Class-D we’ve seen
Then further down he says this????? (hang on he just said spec don’t matter)


The proof’s in the pudding when we see independent measured specs on.
what switching frequency used
output filter corner frequency
switching frequency residue on output
dead time performace
phase shift figures from 2khz to 20khz
rms output wattage just before clipping into 8ohm, 4ohm and 2ohm.
All of which are known Achilles Heel’s of class-d

Cheers George
klh007
621 posts
 
@georgehifi, Tommy's MEGAschino MK2 specs into 2 Ohms;   MEGAschino channel boards are capable of comfortably driving 1000W into 4Ω and up to 2000W into 2Ω
audiozenology
362 posts
 
Measurements of Tommy's old amps are online. No issues into 2 Ohm except a hit of hit on frequency response.
georgehifi's avatar
georgehifi
10,312 posts
 
comfortably driving 1000W into 4Ω and up to 2000W into 2Ω
That’s way too vague. Just like Merrill quotes.
Ask him this, it should be done by independent bench testing, but he should tell the truth.

In stereo what is the RMS wattage "both channels driven just before clipping"into 4ohms @ 1khz

In stereo what is the RMS wattage "both channels driven just before clipping"into 2ohms @1khz

See if you get a straight answer, not even Merrill would give an answer to this and I asked 2 or 3 times on the Element 118’s (true monoblocks not bridged stereos) their was a dead silence with this like he never seen the posts, even though he happily answered many others I asked, and his amps are in a higher league than these I believe.

Cheers George
atmasphere's avatar
atmasphere
11,973 posts
 
I did not understand what kind of patent is this as class D amps are widely used for already more than a decade.
2 decades actually, but that does not mean that everything to be known about them is already known.
The proof’s in the pudding when we see independent measured specs on.
what switching frequency used
output filter corner frequency
switching frequency residue on output
dead time performace
phase shift figures from 2khz to 20khz
rms output wattage just before clipping into 8ohm, 4ohm and 2ohm.
All of which are known Achilles Heel’s of class-d
This type of generalization is problematic! A self-oscillating class D amp can have a fairly low switching frequency (400KHz for example) and consequential low filter frequency (80KHz) and yet little or no phase shift at audio frequencies. This is because a self-oscillating class D amp can run so much feedback that it can correct phase shift and even distortion caused by the application of loop negative feedback. In order to do this the feedback has to be in excess of 35dB which is nearly impossible with conventional amplification due to poor gain bandwidth product. But gain is easy to create in class D circuits.

More to discover