AVS Forum banner

1 - 20 of 43 Posts

·
Registered
Joined
·
9,191 Posts
Discussion Starter · #1 ·
I am starting this to discuss the recent Class D amp offereings. technology, listening observations, whatever. And you can also discuss Class A,B,G,AB designs as well as they stack up to Class D.
 

·
Registered
Joined
·
527 Posts
If you are going to discuss class D you should distinguis before/after Hypex.


I have had Jeff Rowland 301 class D. I think MSRP at that time was $33k. Now I have Theta Prometheus with Hypex Ncore NC1200 module which retail for $12k. They are in a different league and the best amps I have heard ever. Does not remind me at all about class D from the past.


Unlike a class-AB amplifier a conventional class-D amplifier's output stage must employ a low-pass filter to block the noise generated by the 400kHz switching frequency. This filter interacts with the speaker load and is the reason for the less organic sound. But with the Hypex technology this filter is removed. This seems like a huge leap forward in amplifier technology, at least with the implementation from Theta Digital. Very dynamic, powerful and musical.
 

·
Registered
Joined
·
1,843 Posts
Armand007:


Thank you for your gracious review. We appreciate it.


I do need to make one correction, The NC-1200 module we use in Prometheus does include a filter to remove the switching frequency. The genius part of the design is that the load is included in the filter loop. Thus the frequency response is load invariant and load independent.


Jeff
 

·
Banned
Joined
·
239 Posts

Jeff, as time goes on, and as I live with five Prometheus monoblocks in my system, I am more and more blown away, for multi-channel as well as straight two channel (whether via CB3 HD or via straight to Gen VIII Series 3 DAC). Although I had loved my pair of Citadel 1.5s for the front left and right, I do not miss them at all - the Prometheus monoblocks are even better!  Amazing! I've discussed this in more detail in my upgrade thread at this forum.
 

·
Banned
Joined
·
239 Posts

I've been asked how I like the Prometheus monoblocks! I am very, very pleased with the Prometheus monoblocks, in my system, with my speakers, etc.

 

Before, I had a pair of Citadel 1.5s for front left and right, and Enterprise monoblocks for center and left and right surrounds. In my pre 2007 system, I had five Bryston 7BST monoblocks, and using the CB3's Extreme DACs. Or used Six Shooter as multi-channel preamp with Integra preamp to decode HDMI, and later a Marantz UD9600 blu ray player. My system was "balanced" sonically in all five vectors. This gave a great sonic equality all around and sounded what I thought was outstanding holographically in all vectors.

 

Early 2011 upgrades - to CB3 HD, Gen VIII for front left and right, and Citadel 1.5s for front left and right. Sound is the best its ever been. One caveat - the front left and right are the best sounding channels with the Gen VIII and Citadel 1.5s.

 

Now 2014 - so far, not only change speakers to five Aerial 7ts (including center) (I am selling four Aerial 9s and CC5 center) and two JL-f212 subs (I have two of three Aerial subwoofers still for sale), also give Prometheus monoblocks (instead of three Enterprises and two Citadel 1.5s, all sold already). The biggest clearest improvements are multi-channel ripped SACDs from my media server AND the 7t as a center channel (you must have plenty of room so the center 7t is spaced far enough away and in back of the left and right 7ts, which I have, otherwise the center can sound compressed) is marvelous - everything is so crystal clear! Of course, the bottom end is better than ever, as the 7ts clearly have better bass, and the JL subs are much better, lower and "feelier" than the Aerial subs - caveat, I still have work to do to properly locate and set up the subs per JL Audio's "Soundoctor" sub tech. Movies and tv are better than ever, too, but again, multichannel ripped SACDs I really notice it. And yes, straight 2 channel using only the Gen VIII Series 3 DAC and Prometheus left and right monoblocks is clearly appreciably better, too. Its like in the past week or two that my new system has come of age sonically, thanks to burn in, angling of speakers, etc. Now I know what I was missing.

 

With my 2011 upgrades, yes, my system sounded clearly overall its best at that time - but with the recent upgrades, I can now "hear" that the 2011 improvements sonically greatly improved the front end but at the cost of "fronting" the sonics because the sound was that much better at the front. Now my sound seems balanced and holographic between all speakers at all angles. Sort of makes me at least consider another Gen VIII for the surrounds - though Craig/VGI says it'd be a waste of money, I'm not so sure - having the identical sonics with the surrounds, perfectly balancing my front and surround left and rights, I think would be something I would appreciate. Of course, I'm sort of tapped out financially with the upgrades, so contributions, please? (Craig may be right - but how many of his customers have such a "balanced" system with full range rear speakers the same all around like I do?)

 

I can only tell you that I luv the Prometheus even more than I luved the Citadel 1.5s in my system! I think they are that good! The Prometheus are a hit for me. No heat (hot Az). Half the weight of the Citadel 1.5s and a good amount smaller than the Enterprises (yea - with my buggie right hip). I am very pleased.
 

·
Registered
Joined
·
1,585 Posts
The big step forward was Puzneys' self-oscillating loop including the required low pass output filters. The instability associated with the output filter, previously a real problem for Class D, was put to use.
Quote:
The stroke of genius here is that applying an input signal to an otherwise oscillating class D amplifier will result in the appropriate PWM duty cycle signal corresponding to the audio waveform.
Bob Cordell, Designing Audio Power Amplifiers


Now, as longtimelurker mentions, the more interesting question is whether a self-oscillating switching-mode amplifier circuit with a low pass filter and PSRR of 80db (typical) can at all benefit from a linear PSU.
 

·
Registered
Joined
·
9,191 Posts
Discussion Starter · #9 ·

Quote:
Originally Posted by RUR  /t/1524078/class-d-high-end-amp-thread#post_24524968


Both of the nCores - Theta's implementation and the DIY NC400 - are load invariant. Other OEM implementations e.g. Mola Mola and Merrill (sp?) probably are, too.

http://www.hypex.nl/docs/papers/ncore%20wp.pdf

I'm going to save this reading until Sunday morning while I listen quietly to my tube amps.



I am very curious about this new technology as my knowledge of Class D is in fact dated.
 

·
Registered
Joined
·
9,191 Posts
Discussion Starter · #11 ·

Quote:
Originally Posted by longtimelurker  /t/1524078/class-d-high-end-amp-thread#post_24524953


probably makes more sense to discuss class-d power supply strategies. the restbhas been solved.

Just as a point of reference, QSC is exclusively switch mode power supplies in their current models and these are mostly class AB designs. I believe Crown has a few models with switch mode power supplies as well. And there are others as well. But QSC is more marketed towards PA systems then high end home use although they are quite popular in commercial movie theaters.


Now I just built some 20watt class A amps for my LCR tweeters. I cautiously used big switch mode power supplies, 300W x 2 as I had no desire to build a 600 watt linear supply and that fact that true single ended class A has poor PSRR, it would have to be regulated or at least actively filtered. While there is some noise visible on the scope from the switchmode power supplies, about 20mvpp, the tweeters are dead quite.


So the long standing rule that switch mode power supplies are not for high quality audio amplifiers may be more urban legend than fact. Then we also need to consider the switching frequency is much higher than years ago and thus the harmonics in the audio band are far lower.


As for class D I would expect that a switchmode PS is a no brainer. That is the usually the case for classic class D but I need to read the technology behind the Hypex first.
 

·
Registered
Joined
·
1,585 Posts
Far easier to build a linear PSU than a good switch mode PSU. The question is whether - fun aside - it makes sense to do so or simply use a Hypex SMPS. A look at the NC1200 or NC400 datasheet statistics for which the amp modules were coupled with Hypex SMPS, it would appear quite a challenge to better the SMPS. But I'd really like to hear from a specialist in this particular area rather than the peanut gallery.
 

·
Registered
Joined
·
1,585 Posts

Quote:
Originally Posted by Glimmie  /t/1524078/class-d-high-end-amp-thread#post_24527442




Now I just built some 20watt class A amps for my LCR tweeters. I cautiously used big switch mode power supplies, 300W x 2 as I had no desire to build a 600 watt linear supply and that fact that true single ended class A has poor PSRR, it would have to be regulated or at least actively filtered.

Why? This wouldn't be too hard and especially not for a trained EE. I don't think you will have wanted the main PSU to be regulated. Cap and some sort of PI or resistor filtering, yes, but not regulated in the sense one would normally mean.


Anyway, I guess we should stray off the topic of class D.
 

·
Registered
Joined
·
9,191 Posts
Discussion Starter · #14 ·

Quote:
Originally Posted by stevekale  /t/1524078/class-d-high-end-amp-thread#post_24527542


Why? This wouldn't be too hard and especially not for a trained EE. I don't think you will have wanted the main PSU to be regulated. Cap and some sort of PI or resistor filtering, yes, but not regulated in the sense one would normally mean.


Anyway, I guess we should stray off the topic of class D.

This is a case of a hobby project, not a manufactured product. Yes, a linear supply is simple and even a regulated version is not that difficult. But at 600 watts, the components are large and expensive. A regulator or active filter will require multiple pass transistors with associated heat sinks. It's true that regulation is not required but an active power filter is almost as complex as a simple open loop regulator.


My plan was in fact to move to a 60hz active filtered linear power supply if the SMPS was too noisy. That would have cost me at least $300 in parts. But fortunately the $20 surplus SMPS's did OK.


And above all, class A amps are not really practical these days. I built them because I wanted to try Class A and yet mine are only for tweeter duty at 20 watts each. While a class A amp in the hundreds of watts range is possible and even buildable by a hobbyist, it just doesn't make sense especially in an HT setting - JMO.


And I'm not so sure the Krell iBias technology is really true class A. But it does seem to work well. The first generation Krell amps were true class A monsters.


Anyway it would now seem I need to scrap this month old class A amp and get some Hypex modules to play with
. However since I just got my butt chewed after last month's Visa statement came in with pages of Digikey charges, I better lay low for a while!


P.S. Pi filtering using resistors is limited to low currents of less than 100ma. Not even practical in a 60 watt tube amp although Dynaco did do it for their Model 50. Chokes ruled in tube amp power supplies and still do. But here we are talking hundreds of volts with only a few hundred ma of current max. For a solid state amp a choke filter is very difficult if not impossible at high currents and low voltages. To get enough Henrys of inductance at the required currents, the series resistance would be far too high*. I have seen some class A amp supplies that do use micro henry chokes. But I'm not sure how effective that really is. Legacy high end HiFi SS class A amps that did not use some form of active filtering used vast amounts of filter capacitance, some up around 100,000uf.


*at 50 or 60hz. High current chokes do work very well in SMPS's at 50khz plus.
 

·
Registered
Joined
·
563 Posts
ding ding ding


the hypex smps per thier own papers is perfectly sufficient.


but you can sell it for 10x the price with an in house, possibly inferior, power supply.


i
 

·
Registered
Joined
·
1,585 Posts
And the supporting evidence for the above statement is....what exactly? Please elaborate and therefore contribute.
 

·
Registered
Joined
·
169 Posts
Hypex are good, but there are a lot more companies out there making excellent class-d.

Like Anaview, which in my system sounded better than Hypex nCore.

They are also available in the Marten Statement amps, costing megabucks.

So Theta are not alone..
 

·
Registered
Joined
·
339 Posts
I built my own Hypex three channel amp. I ended up selling. I think one will find the power numbers are like calculating hp between diesel and gas engines, not comparable. The amp was good, even great. Just not something that I would live with forever.


If you want to know all the technical in's and outs, then you should start the journey of reading the thread on Diyaudio.com. Here you will see comments from the manufacturer as people build their various Hypex Ncore amps..

http://www.diyaudio.com/forums/vendors-bazaar/190434-hypex-ncore.html
 

·
Registered
Joined
·
1,585 Posts
Let me know when someone has summarised what I would expect to be a relatively few good posts in amongst 747 pages of other stuff.



"Power" is a well-defined concept.
 
1 - 20 of 43 Posts
Top