2. amps-preamps
  3. 129071-76532

Class D = Trash?

So, I'm on my second class D amp. The first one, a Teac AI-301DA which claimed to use an ICE module, was unlistenable trash. I burned it in for a few weeks, it just couldn't perform, so I sent it back. Following that, I tried the new Emotiva A-300 (class A/B). It was significantly better, but lacking in too many ways for my tastes. So I changed gears, got an 845 SET from China -- and it was an immediate and massive improvement.

So, before I went further down the SET road, I wanted to try a better class D product using a modern class D module. I settled on the D-Sonic M3-800S with the Pascal module and custom input stage. I read from reviews that these things like to have big cables, so I picked up an eBay 8 gauge power cable (Maze Audio, el-cheapo Oyaide copy plugs, braided 4-wire cable) to go along with it.

Mid-range GONE.
Soundstage depth CRUSHED.
Euphonics DISAPPEARED.

Yes, resolution went up. Driver control went up, allowing me to play compressed rock/pop and orchestra with the speakers being able to render it all. But enjoyment in the sound is basically gone. Using my best power cable (LessLoss Original) improved performance, but didn't fundamentally change the amp's nature. I ran back to my headphones (Focal Utopias) to detox my ear canals.

So, how long does a class D need to burn-in? I want to give it a fair shake before writing the technology off forever. 
madavid0
seanheis1
359 posts
 
Again, one experience with ICE modules cannot possibly speak for all Class D amps, just the one amp you heard. Goodness folks. I heard a poor tube dac once so they all must have a ways to go.....crazy thought process correct?
These issues with amps that use Ice modules are well documented on this forum. Amps using NCore, Pascal, and Abletec are generally viewed as favorable. NCore is often times described as similar to Ice, but improved. Plenty of folks don't care for NCore though...None of these Class D amps are as universally agreed upon compared to Class AB offerings from McIntosh or Mark Levinson.  
erik_squires
5,268 posts
 
Kvetching does not equal documentation.

The latest generation of ICEPower (AS or ASP) are wonderful, and to my ears indistinguishable from good A/AB amps.

Any attempt to use them as examples of Class D inferiority is highly suspect to me, based on experience.

I'd be happy to evaluate vs. nCore or Pascal when some one provides me the funds... :)

Best,

E
madavid0 OP
127 posts
 
The Teac I mentioned that supposedly used an ICE module was virtually un-listenable. I will admit that I only burned it in for a few weeks, perhaps not enough for a class D.

This D-Sonic uses a modern Pascal module for which I read has technical wizardry that goes way over my head in order to reduce switching noise even further than previously accomplished. When it doesn't sound like trash, I do appreciate the extra resolution and cohesion in the face of difficult passages. I guess I'll just have to be patient.

But, damn, this is the first time I've experienced fluctuating quality during burn-in. I'm very used to burn-in, but there's always a linear progression from "horrid" to "good".
guidocorona
4,114 posts
 

Very interesting point about switching frequencies… What is the acceptable switching friequency range is targeted by the pundits?

 

Given that break-in of a class D power amp ranges from 600 hours to approximately 1500, asking a manufacturer to exercise the device for upwards of a couple of months before shipping it to a dealer may be woefully unrealistic… Of course, everything is possible…. For a commensurate list price uplift.

 

G.



madavid0 OP
127 posts
 
It's just that...jeeze...1500 hours...damn. I thought waiting for the Furutech outlet to break in (800 hours) was savage...
More to discover