Phase shift in class D amplifiers. How it affects the sound?

In Spanish there is a saying: quien calla otorga -> who keeps silent says yes.

If you want emotion, forget about hifi. Class D sound nice, smooth, I have nothing against the sound. Good class D sounds actually subjectively better than many other class A/AB amplifiers. It also is more accurate on low sensitivity speakers than many tube amps.

However for emotions in the music, there are no actual correlation.

- By far the most emotional experience I had: 5 inches paper alnico fullrange in semi open baffle, driven by a cheap cd player and 845 S.e.t. Nothing above 5khz or under 100hz, however brought everyone to tears listening to opera cd's...

-By comparison, (15 minutes ago) I was at Steinways store listening to recreation of pianists on the grand piano with actuators to recreate a live performance. It was very poor experience. It was the real piano, but no emotions. The humanity of the player didn't translate through the electrical most sophisticated actuators and electromagnets.
 
When I was a kid I wrote many letters by hand to many embassies asking for information. All answered me, all. The ones that sent me the most information were two: the old USSR and Saudi Arabia!

The recordings that excite me come from old analog masters. Usually very great R2R/vinyls/CD with masters made before 1980. Why? Tubes equipments? H2 predominant in the masters? I do not know.
 
Last edited:
Don Quijote. Icepower does not sell these modules to you, they won't offer you support, not now and not when you have a product with defective modules.

Just now:

Reiner, CSO - Richard Strauss - Don Quixotte (1960), vinyl 24/192

URL]
 
Last edited:
maty,

As below in a limited window of X and Y axis they do show a frq response in 10.1 of datasheet, blue is 4 ohm load, green is 8 ohm, and red is no load:
attachment.php


Using a free trace program we can get a rough response copy of above red trace and then again using a free program phase extractor we get below minimum phase domain curve for that response (20kHz point phase is a minus 6,8º number so far with out knowing more about low pass slope up higher):
attachment.php
 

Attachments

  • 5002.PNG
    5002.PNG
    24.3 KB · Views: 781
  • 5001.png
    5001.png
    18.2 KB · Views: 787
Last edited:
...PASCAL module has a little better frequency curve but the phase is not like your simulation.
Yes its the limited window of X and Y axis that hinders us to get the real low pass slope up higher where so far they use a 70kHz window limit.

You can use free Rephase and estimate that Pascal graph, below is a dirty quick 4th order low pass at 80kHz and looks close.
 

Attachments

  • 5000.png
    5000.png
    11.9 KB · Views: 108
I think its as good as that Pascal in that we can see frq response goes up again where window leave us uninformed above 70kHz, its probably some kind of resonance filter at 60kHz that is needed for that amp topology and its allowed cost budget as Eva kindly reminded us play a role.

If you have these modules yourself its much easier hang it inside a soundcard loop with some resistor divider network and measure performed amplitude and phase domain, if you use REW then align impulse with Controls/"Estimate IR delay" and then Controls/"Generate minimum phase" and you will see if it deviate from minimum phase domain as discussed over at ALPHA thread where xrk971 and AKSA seems to like output verse input have faster phase as frq goes up than a pure minimum phase domain device will dictate. What they like is what can be done with any FIR filter as crazy as you want it smooth or resonant like but i don't agree it can be done without measurements in acoustic domain because as far as i know real world and sound are mostly of minimum phase nature and the same are recording and reproduce chain that music material are mixed under, so in my view only thing they get doing that by ear is a sound system that stand out because their brain feel for a improved 3d soundstage that can be hard to be recreated for others because we don't have same acoustic environment and exactly same chain of playback gear as they have. And if you intend use this digital amp for bass duties only then forget about its phase turn at 20kHz point because it doesn't really matter when there is no amplitude.
 
Just to clarify:

Emotion with audio gear has 2 possible paths:
- Being alone: Requires emotional connection with someone that understands the technology, both emotionally and in numbers. Requires emotional connection with someone that understands music. A tech friend. A music friend. Or being both yourself.
- Group listening: Requires emotional connection between the listeners. Requires emotional connection with someone that understands music. 1+ techs in the group. 1+ musicians in the group.

Some companies actually provide the emotional support network for their products. Some know the formula. Some just imitate. I have to admit that I do underground parties with prototypes at nature locations, where money has little meaning.

Some designs can't resemble as clean or emotional as others. This often happens when the creative process was just "stress for money under deception". No pride other than money.
 
I was just a click away from buying it months ago, but I decided to focus on cleaning my noisy mains and I got it right. The last improvement was so spectacular that I gave up. Since then I am more easily and more emotionally excited when listening to certain very good recordings, which is what matters to me.

From analogue masters.

The idea is/was if with a new system I can get excited with very good records from digital masters. With this phase it is ilogical to buy it.

ALPHA20 is only 18.9 watts and heat (BB52 is too heat and it needs four fans). IcePower 1200AS2 is all-in-one module and cheaper than Hypex NC500 + good buffer (with very low SNR) + Hypex/Cresnet SMPS.

When in doubt, waiting the mail answer, I better forget this module and I spend money.
 
I do not have emotional connections with the very good recordings. I am a very logical man but I want very good sound with emotion too, spending as little as possible and facing the problem from a technical and logical approach. Good specs/measurements are necessary condition.

Risks must always be minimized when money is spent I think. I want to make decisions based on good information and avoid compulsive purchases.
 
Last edited:
Human emotions use quantum entanglement in their chemistry. This implies that human emotions are always shared with other living organisms. The "1/d^2" point source field intensity law no longer holds at subatomic level (quantum entanglement does not depend on distance, but on a gated stimulus in common, which establishes a "subatomic synchronization" between neuronal networks of different individuals). You may be a victim of the negation of this principle, as from some unsolved deception in ancestors life (a superb spiritual relationship was broken for material interests).

Said that, I will also issue a warning: It's always better to get relationships right in the spiritual plane, even is this implies imbalance in the material plane, because the material plane is continuously (though slowly) being replenished by nature. The spiritual plane isn't, it's stationary.
 
Last edited:
Last edited:
Listen Ashina + Taelue + Chicagodeep

Good music, hypnotic... you do not need to spend a lot of money in hard to listen the music. Only you need is power and a subwoofer. Cheap and old class D is perfect here.

Without voices and acoustic instruments you do not care about the phase.

When listening to this music, a dopamine discharge is sought.
 
Last edited:
Opera and the like are too filled with references to bogus content of ancient cultures and power/gender/money/class conflicts often not possible to solve within that performance format. The dramatized, exaggerated and flavored exposition in full length of some mistakes of the past of the human race can be forgotten, once the synthesized version is done.

Live electronic music is made with synthesizers, computers and crowd feedback. The raw function of music is to connect people and nature. The sense of music event as a religious tribal-bounding cult is restored.

The problem with Opera is that most of them "learned" by paying a prostitute (scam, can't learn anything on top of irrational experience).

"High-End" means that the technology resembles a higher purpose (one that has been slowly developing for a long time), which can only be understood by studying the technology and the creators.

Electronic music has recordings in all sorts of qualities. There is plenty of transient information at all frequencies. While electronic music can sound barely OK with junk systems, there is usually one more level of detail only appreciable under the right human and technical conditions. Some can argue "the kids do not understand that", but the kids are hooked to electronic music (like hell) seeking for that. Emotional connection is not as good when frequency ranges are missing, leading or lagging. Voice is a more ample concept than just human voices. Synthesizers can be heard as non-human voices.

Be careful with dopamine because this is the cheapest to manufacture neurotransmitter in cells. Dopamine by itself is not high end, but it is an ingredient in more complex processes requiring several messenger substances, which can be high end.
 
Last edited:
I am write about the sound.

With electronical/computer instruments if you change the phase or other parameter it will sound different but not weird. That is the problem with the modern commercial music: there are few hours in the studio and samplers and others things are used; they totally manipulate the sound and in the end they reduce the dynamic range a lot -> you do not need a very good audio system, only power and subwoofer.

With natural voices and acoustic instruments is other story. You want natural, credible sound, emotion... and not only spectacular.

I did not go into musical tastes.
 
Last edited: