Class D amps

Dan Banquer · « **on:** 4 Mar 2006, 01:13 pm »

There is a new article up at audioholics on Class D amps by Bruno Putzeys.
http://www.audioholics.com/techtips/audioprinciples/amplifiers/SwitchingAmplifierBasics.php

Another article on this very same subject that I have been working on, will follow shortly.
d.b.

Occam · « **Reply #1 on:** 4 Mar 2006, 01:41 pm »

Dan - Thank you very much for the link to the very interesting article. Dunno specifically what amp showed those zero cross spikes, but I'd assume the spectrum of those massive spikes would extend into the FM band .

Sadly, I'm not at all surprised. I'd love to open up one of these highly acclaimed by audio press, killer amps to see the snubbers accross the half bridges (of their full bridge) to see what caps (if they actually used them) they used......
Then again, I'm not the objectivist you are. I do believe that subjective perceptions trump objective measures, even if it craps up your FM reception. It might be happenstance that the few times I've heard (what I assume is) the 'great radiator' (admittedly not their top-o-the-line) it was thoroughly mediorce.

Thanks again,
Paul

tianguis · « **Reply #2 on:** 4 Mar 2006, 02:35 pm »

Dan:
Thanks for the link. Very informative.
Having built and modded an ICEpower amp (dual 250ASP B & O modules) the article gave me some ideas for further tweaking.
I didn't try to totally seal it (the chassis has louvers), just used some judicial shielding and damping. Unlike the totally sealed Bel Canto S300 (which I had in my system for several weeks) mine is very quiet in the system, unlike the Bel Canto which transmitted lots of trash to IC's and speaker cables. However, I don't use a tuner so I can't comment on that aspect.
Thanks again.

Regards,
Larry Welsh

mgalusha · « **Reply #3 on:** 4 Mar 2006, 04:55 pm »

Dan, appreciate the link, that was a very interesting and well written article. I'm looking forward to your upcoming article about class D as well.

Mike

audiophile39 · « **Reply #4 on:** 5 Mar 2006, 07:14 am »

I believe Channel Islands employs Mr. Putzey's technology. My take-away from his article is that Class D amps do not offer a real sonic improvement over conventional solid state amps. Please correct me if I'm wrong

TNT · « **Reply #5 on:** 5 Mar 2006, 08:00 am »

Quote from: audiophile39

I believe Channel Islands employs Mr. Putzey's technology. My take-away from his article is that Class D amps do not offer a real sonic improvement over conventional solid state amps. Please correct me if I'm wrong

"Hypex modules" is more precise than "Putzeys's technology" who himself doesn't think 16 transistors could be called "technology"

Re. sonic improvement, it all depends on the listener. Recent generation of class-D amps are well liked by many, otherwise the number of channels of class-D amp sold wouldn't have been more than 50 millions and growing, taking into account the fact some good linear amp ICs sell for peanuts, i.e. $0.75 per channel.

djbnh · « **Reply #6 on:** 5 Mar 2006, 12:29 pm »

Quote from: TNT

it all depends on the listener. Recent generation of class-D amps are well liked by many, otherwise the number of channels of class-D amp sold wouldn't have been more than 50 millions and growing, taking into account the fact some good linear amp ICs sell for peanuts, i.e. $0.75 per channel.

The above reminded me of the # of IPODs being sold / IPOD sound quality.

ctviggen · « **Reply #7 on:** 5 Mar 2006, 01:18 pm »

You also have to realize that "class D" is a misnomer, in the sense that very different technologies are being combined under the "class D" umbrella. The UcD modules, for instance, use a triangular wave to create PWM. The NuForce amps operate using a different technology (which can be analyzed by downloading their patents). As for me, I have a NuForce amp on my center channel, and it replaced one (and two, when I was biamping) channel of a Bryston 9b. For me, I have experienced no radiative effects, though I never listen to FM or AM. Nuforce also seems to be a very revealing amplifier. Even when using triangular modulation, Class D modules can be very different, as illustrated by the article.

Also, if anyone desires, I can put a scope on the end of the Nuforce and see what I get. I have a 100Mhz digital scope with FFT, so I could also determine a frequency content in addition to a shot of the output. Please PM me, though, as I've gone DIY and don't come here often.

As I plan on building Linkwitz Orion speakers, I'm going to need 8 channels of amplification for two speakers. I'm going to order an Aksa, Greg's simple killer amp and UcD modules and build three amps. I'm going to compare which ones I think have the best sound and then choose that amp to replicate for the Orions. The second best will become the amp for my second system, and the third best will be sold. From a DIY builder's perspective, UcD modules are much easier to build than the Aksa and SKA will be -- Bruno offers all the modules you need, and you just have to plug them together and get a case. A 4 channel amp would be relatively simple with UcD but a bigger pain with class AB.

Dan Banquer · « **Reply #8 on:** 5 Mar 2006, 02:15 pm »

Hi Bob;
You're right that Class D is a misnomer. What has happened over time is that folks have lumped all the different technologies of switching amps under one common word. Personally I would prefer the term switching amps.
Since you have a scope with an FFT please feel free to share your results with us.
d.b.

JohnR · « **Reply #9 on:** 5 Mar 2006, 02:42 pm »

Um, I don't read content on audioholics, but I don't see how "Class D" can be a misnomer, any more than class A, or B, or whatever...

TNT · « **Reply #10 on:** 5 Mar 2006, 04:23 pm »

"The above reminded me of the # of IPODs being sold / IPOD sound quality"

Fur under 50W, linear amp ICs are MUCH cheaper than class-D amps, integrated or not. The Patek chip amp is high-end in audio performance - according to 6moon and some AC posters - and riduculously high priced for a hand-wired breadboard and fancy heatsinking/chassis. It electronic content is maybe < $10-15. No electronic R&D. Why, its designer is mechanical engineer.

iPod is for the mobile generation and got well deserved success, even though personally I will never buy one - listening compressed music with earphone is not for me. Eventually a version of uncompressed iPod can become a convenient music server. I bet that will happen soon, maybe they'll come from modders first.

gary · « **Reply #11 on:** 5 Mar 2006, 06:16 pm »

Quote from: TNT

iPod is for the mobile generation and got well deserved success, even though personally I will never buy one - listening compressed music with earphone is not for me. Eventually a version of uncompressed iPod can become a convenient music server. I bet that will happen soon, maybe they'll come from modders first.

It's already happened, the Red Wine Audio modded iPod running lossless format files is a formidable digital source, to put it mildly.

Gary

randytsuch · « **Reply #12 on:** 6 Mar 2006, 04:27 am »

Quote from: TNT

"The above reminded me of the # of IPODs being sold / IPOD sound quality"

Fur under 50W, linear amp ICs are MUCH cheaper than class-D amps, integrated or not. The Patek chip amp is high-end in audio performance - according to 6moon and some AC posters - and riduculously high priced for a hand-wired breadboard and fancy heatsinking/chassis. It electronic content is maybe < $10-15. No electronic R&D. Why, its designer is mechanical engineer.

iPod is for the mobile generation and got well deserv ...

Off topic, but have you ever looked in a Patek amp? The designer likes to use expensive parts, like blackgate caps, riken and caddock resistors. Your $10-15 figure is off by at least a factor of 10. And, his time is not free, so you need to factor in that these are hand built, and hand wired. I highly doubt Peter is getting rich off of amps.

Randy

TNT · « **Reply #13 on:** 6 Mar 2006, 05:30 am »

Yes I did look at Patek, as at 6moon.com

We all agree that time is not free. Therefore cottage audio is always expensive regardless of class A/B/D. But intrinsically linear amp is cheaper at below <50W, especially chip amp. Many can make linear amps using IC or transistors, few can design good switch-mode amps, especially the self-oscillating type (UcD, NuForce, IcePower etc...) due to more difficult technologies (feedback control, magnetics, EMC etc...) therefore also much higher NRE. For the power level they produce, switch-mode amps are cost competitive.

gary · « **Reply #14 on:** 6 Mar 2006, 01:33 pm »

Quote from: randytsuch

Off topic, but have you ever looked in a Patek amp? The designer likes to use expensive parts, like blackgate caps, riken and caddock resistors. Your $10-15 figure is off by at least a factor of 10. And, his time is not free, so you need to factor in that these are hand built, and hand wired. I highly doubt Peter is getting rich off of amps.

Randy

More like a factor of 50. The BG capacitors alone are $200, the Cardas binding posts & RCA jacks are close to $100, the enclosure requires about 5 or 6 hours of a machinists time (~$300 or so) plus materials, and on top of that you've got the power supply with either an expensive toroid or batteries & a charger. I know all of this because I'm building my own, and by the time I'm finished I'll probably have spent $500 cash plus cashing in favors to get the enclosure made.

Gary

skrivis · « **Reply #15 on:** 6 Mar 2006, 01:51 pm »

Quote from: JohnR

Um, I don't read content on audioholics, but I don't see how "Class D" can be a misnomer, any more than class A, or B, or whatever...

IIRC, Class D only refers to amps using PWM. It seems like most of the current "Class D" or "digital" amps are simply variants of Class D. It's certainly a better way to describe them than "digital."

Still, if an amp doesn't fit the definition of Class D, it shouldn't be called that just because there isn't a better category for it, or for marketing reasons.

I'd also like to point out a good article at:

http://sound.westhost.com/articles/pwm.htm

Kevin Haskins · « **Reply #16 on:** 6 Mar 2006, 03:13 pm »

Both are great articles and there is plenty of good points here.

In terms of subjective opinions on the amps vs. the objective measurements that topic is never going to be settled.

skrivis · « **Reply #17 on:** 6 Mar 2006, 03:33 pm »

Quote from: Kevin Haskins

Both are great articles and there is plenty of good points here.

In terms of subjective opinions on the amps vs. the objective measurements that topic is never going to be settled.

Probably not.

I tend to lean towards measurements not always being able to identify all of the benefits of a product. But they can often show when a product has problems.

I also feel that, if measurements aren't telling us what we need to know, then we need to come up with better measurements.

I'd say that something that shows major problems is not a product I want to spend further time on. So measurements are certainly useful in that case.

Beyond that I think I'll keep my opinions to myself.

Kevin Haskins · « **Reply #18 on:** 6 Mar 2006, 04:47 pm »

Quote from: skrivis

Probably not.

I tend to lean towards measurements not always being able to identify all of the benefits of a product. But they can often show when a product has problems.

I also feel that, if measurements aren't telling us what we need to know, then we need to come up with better measurements.

I'd say that something that shows major problems is not a product I want to spend further time on. So measurements are certainly useful in that case.

Beyond that I think I'll keep my opinions to myself.

The main problem with subjective opinions is that they are not reliable. They can vary a great deal with what you had for lunch, mood and other things that have nothing to do with the items you are comparing.

For making personal choices on audio equipment I see no other meaningful way to pick something. Yes... I'd look at the technical details but where the rubber meets the matt is how it sounds. You should just understand how it sounds can vary a lot with your imagination.

From a design standpoint I'm almost diametrically opposite of that approach. Why? Simple... I understand that subjective opinions are terribly unreliable and to improve something you have to use some kind of reliable metric to gauge improvement.

I spent a lot of time modifying a CD player a few years ago. I bought two units, one to remain stock and the other to modifiy for comparison sake. I spent a lot of time modifying one unit and then comparing with the stock unit. I did all of this by ear. After a couple months of twiddling with it I had my wife listen to both blind. I had thought that the improvements where very audible. She choose the stock unit over the modified one. Grrrrr....

I then did some blind testing on myself. I couldn't reliably detect which was which.

Moral of the story; be real suspicious of your subjective opinions. They don't always tell the truth.

miklorsmith · « **Reply #19 on:** 6 Mar 2006, 04:54 pm »

I recently did a half-ass blind test with a compadre where we plugged a Modwright 9.0SE into his Krell integrated. We were able to accurately level-match.

Although I am convinced (as was he) the system sounded far better with the Modwright in the loop, neither of us was consistently able to determine whether it was "in" or not.

Blind testing be damned!