I've been using a 1993 vintage Adcom Amp since...1993. I moved to an apartment last year, and purchased some more efficient speakers. The Adcom runs very hot, which is a problem in the summer. Turning the AC unit in my listening room is definitely a negative.
It seems that Class D has really come into its own measurement-wise (pretty much class leading in S/N and THD), and I'm very interested in the heat and space efficiency of these new designs.
The amp I bought is based on the Hypex Nc252mp (https://www.marchaudio.net.au/product-page/p252-stereo-250-watt-power-amplifier). It seems to me these amps are going to behave like the underlying classD power module/supply unless the designer adds something weird, like a class A input stage, to make it cost more money. Or new op-amps, but that appears to have no effect.
After putting in the amp I do subjectively perceive a difference in sound quality. I suspect it is invented. Unfortunately, it is very difficult to a/b the amps due to safe connecting/powering concerns.
Has anyone run any blind testing of new Class D designs?
OK, I've had the new March Audio p252 in the system for two days.
I had a bit of a long strange trip comparing it to my old amp (Adcom 5802). When I first connected the March, I was sure I could hear a difference. A bit tighter in the bass, not as smooth in the highs. Was the image not as wide? The soloist seemed a little further back?
This evening I had a chance to A/B them, albeit under less than ideal conditions. Since I had to power them off before reconnecting, there was a bit of a lag between listens, and, of course, I'm getting up from the couch to do it. Just to complicate things, the March reverses the channels, so I had to be careful to connect right channel on my DAC to left channel on the March. I put in a pair of cables with bananas instead of spades, to make it easier. After trying to volume equalize on the music, I started doing it on a 1k test tone with a decibel meter app on my iPhone. As it happens, the two amps have the same gain, so I realized I didn't have to change the volume at all. Eventually I got the switch down to less than 30 seconds.
The better I got at making the swap, the less I could hear a difference. When I started I was almost sure I could, to my distress. Ninety minutes later I'm pretty sure I would fail a blind test between the two.
On the one hand this is oh-so-predictable. I am pretty convinced, in my head, by the volume of tests that suggest this would be the case. On the other hand, on first blush I started hearing some of the things subjectivists and kilobuck amp designers like to say about Class D. I guess my doubts rule my biases even more than my reasoning!
All in all, it was a weird experience. I'd love to get switchboxes that would allow me to make more immediate comparisons. They'd have to be 1-in to 2-out, XLR and Speaker (with some pop protection?).
It's nice to have a cool amp here in the living room, so I can leave the A/C off and enjoy the music. Also, with only the DAC/Streamer and the little p252, I could get rid of the rack and de-clutter.
Next up, I'm thinking I'll test the Orchard Audio PecanPi against my Cambridge DAC. I'm sure the sound will be fine, but I really like the Cambridge streaming features, including Chromecast built-in.
Here are the claims/theories of why Class D sounds different:
-having, effectively, an AM transmitter in your system could have audible effects -The same high frequency info can cause IM distortion in the audible band, but this measures well below audible levels in specs -" " could cause odd behavior in tweeter (this seems like BS to me, the energy level would be ridiculously low) -lots of feedback in a Class D design (not clear that high overall feedback designs have been shown to be audibly different) -PWM signal sounds qualitatively different
MMostly it is kilobuck amplifier designers who suggest these theories, requiring more than a grain of salt (although the first one is a fairly objectivist speaker designer, Alan Shaw). I just don't know if there is evidence of audibility for any of this. I've seen quite a few blind amplifier tests, but none involving Class D.
I don't think it's only cables. The main idea is that a pure distortionless signal will completely null another pure distortionless signal. It does not matter if it's a wire or an amp or whatever. If there is a difference, that difference can be measured and compared with the test input signal level. If the null difference is below a certain dB difference, say 95db, that difference would presumably be below human hearing detection and the devices under test could be considered equal in audio transparency, i.e. virtually distortionless.
That's just my 2cents. Please see Ethan's description. Sorry I don't have a link handy.
Post by Michael Lawrence on Aug 20, 2019 17:33:50 GMT
This is all nonsense. There is no audible difference between audio amplifier topologies unless the designer of a specific piece of gear is truly incompetent. Most Class D amps have an integrating low pass filter at the outputs that removes all ultrasonic components. (Some models use the loudspeaker inductance to accomplish this.) People said the same thing about digital converters for a while, until people realized how digital "works."
There's actually far LESS negative feedback in a class D topology, which is an inherent fact of how the switching mechanism works. (If it's taken from the output stage, it must be averaged in order to be compared with the input audio signal, and can't compensate for the output filter's distortion. If it's taken after the output filter, there will be some phase shift relative to the input and so less feedback can be used before instability.)
Thanks for the write up, let take this from another angle and compare cheap amplifiers one class H and one Class D or class A at the under $200 price point which topology if done correctly will be the better choice.
Actually I am looking for a well priced subwoofer amplifier not plate I will buy something powerful so i can upgrade my subwoofer in future if need be. There is so many affordable amps on the market but which one is better for the LF range and at that price point for PA power amps is one class better quality because the other amp potentially is more expensive to build? With one of the topologies being better quality due to its cheaper components while the other is worse because it more expensive to build and is at the lower price point.
It is for home use and need up about 400 watts each channel.
Post by Michael Lawrence on Sept 5, 2019 0:24:37 GMT
For a home application the topology would not be a contributing factor to my decision, especially at those relatively low power ratings ("relatively" because I work in a live sound reinforcement background where devices often require thousands of watts per channel). I would look for an amplifier from a competent manufacturer and at the power handling and price that were appropriate and call it a day. Crown is always a solid choice, XLS series is fine for home use.
Back in the '70s I we only learned Class A, B, AB and C probably 'cause that's all we we're ever expected see? So at this late date, I feel it's high time to look into this a little closer. Even though now I discover the D topology was invented in 1955, seems it really didn't "catch on" till the '80s ...read all about it here en.wikipedia.org/wiki/Class-D_amplifier.
To be honest, I will be going with Chinese made class H amp I don't know if that is sacrilege for Americans or for audiophiles in general but their price point is significantly better plus I am in a 240 volt zone and most Crowns on the parts Express website are 120 volt plus more expensive.
But your answer basically makes me understand for home use, especially for sub frequency use its not going to matter so much. Appreciate the response.