The Future of Audio Amplification

FYI, Class D is NOT digital, it is analog all the way through. There are no zeros and ones.

Until recently, I've been a long time stick in the mud stalwart for class A, AB, tube or solid state guy, but recently, due to some downsizing and restructuring of my listening space, I decided to take the plunge and try PS Audio's Stellar Gain Cell DAC preamp and M700 monoblocks, connected by balanced cables. After a solid week of "break in" around 200 hours (which I am skeptical about), even at the lowest volume as recommended by the folks at the company, I can say that they are as smooth, sweet, musical, warm and deep sounding as any amp I've owned and listened extensively to previously, and, perhaps even a bit more dynamic, especially at low <70 dB SPL volumes. Other amps I've also owned and enjoyed including Manley Stingray, Snapper, and NEO 250 monoblocks, Krell 400cx and intergrateds, CJ tube and solid state integrateds, Musical Fidelity, Bryston 7BST monoblocks, Bryston 4BST stereo amp, Parasound stereo amps (200 WPC and 75 WPC) and two different NAD integrateds.

I conclude that Class D, when done well at the input and output stages (in this case augmenting ICEPower modules with their proprietary Gain Cell mosfet input stage) can rival any other class of amplification. And they run totally cool, use very little wall juice (leave em on all the time!) and have micro and macro dynamics, even at low volume that reveals the source material better than any I've experienced previously. I know that's saying alot, but I have decided to keep them. Call me converted.

The title of the thread is The Future Of Audio Amplification.

The future is here now. It includes efficient toplogies that waste little energy as heat to drive loudspeakers. The future may improve from the current 85% to 95% or higher, but will certainly not incorporate far less-efficient designs such as Class A/AB.

John Atkinson measured PS Audio's M700 monoblocks ability to produce 700 watts at 1%, and over 100 watts at 0.005%, both figures into 4 Ohms at 1K. At 30 watts into 4 Ohms, which many of us never exceed at normal listening levels, that number fell to 0.002%. From 20 to 20K at 20V, that number never rose above 0.2% into 4 Ohms; the figures were better into 8 Ohms. Not that distortion tells but a part of the story, is something not to worry about with well-designed Class D amplification.

So 0.02% distortion is an "ungodly" amount huh?

Talk about dumpster fires...

And of course we can all hear 400+kHz, right? Yeah, right...

For years I poo poo’d Class D in the same fashion as Kosst has here. But, because of curiosity and a sense that a company I trust, PS Audio, wouldn’t be spouting off about how good these amps are if it couldn’t be backed up, I decided to try a pair of their class D monoblocks.

They sound better than ANY of the MANY of non-Class D amps I have owned in the past, not only to my ears, but to several of my audio buds who have now listened and who are both tubed and solid state (non-Class D) amp owners. They ALL agree that my system sounds significantly better than before with these amps.

I took a chance and am now convinced that the technology has arrived at level consistent, and exceeding, the sonic perfomance of ANYTHING out there.

Also, this nonsense about caps burning up early due to the switching technology is a lame attempt to scare people in an era where fear reigns supreme. Good electrolytics last a long time. Enough of this fear mongering.

Horseless carriages were deeply despised and 'destined for failure' when they started out too. Any car owners out there?

Class D is not digital. 

GaN technology looks to be very promising and I read that the latest state of the art for power transistors in fast switching applications is to thin-layer GaN directly onto SiC substrates. This apparently saves big costs and allows the advantages of GaN to be realized relatively inexpensively.

Looking forward to auditioning affordable Class D amps using this technology.

Here’s a couple:

https://www.researchgate.net/profile/Marco_Fanciulli/publication/224426167_Epitaxial_growth_of_zinc_...

https://www.sciencedirect.com/science/article/pii/S0022024800008538

https://techxplore.com/news/2018-08-fujitsu-triples-output-power-gallium-nitride.html

One thing that is difficult, and I have experience with series 2 and 6 elements to make CdSe-based nanocrystals with ZnS shells, is how to deal with crystal lattice mismatch; some elements’ bond lengths just don’t align well in order to interface optimally. Typically, one had to dope in some intermediary element to "fill" the gaps created by the lattice mismatch. Looks like there’s a similar challenge/constraint in fabricating thin layer GaN onto SiC wafers. This is why, IMO, we haven’t seen the industrial "breakthrough," yet. Ideally, one would want pure GaN crystals in the form similar to that achieved with SiC in order to make large enough wafers.

Gal, I ummm, night ride.