Class D Amps - Anyone else having second thoughts?

[R Baer]

I have designed class AB power amps with SMPS and conventional power supplies, and with the same amp design, the difference between the two supplies was not at all easily noticeable. This was with the NeoPak 3.5 and that convinced me of the future of SMPS. The SMPS was designed to emulate the characteristics of a traditional supply, but that wasn't a big deal. The point here is that the supply was the only difference.

So, identical amps with different supplies and the results were that the differences were very, very minor. Perhaps trivially so.

Well, we never compared our class A/B with with two different power supplies, but I would agree with your opinion on the power supply. In our listening test we tried four totally different power sections, two class D, a class H and a class A/B, mated to the same preamp. In that instance, there was a very noticeable difference in sound and feel between the various topologies. There were no measurements taken and no frequency analysis done. I made the final choice for our amp strictly by ear.

 

[Soverntear]

May not be the reason, but a friend of mine is in school to be an electrition. he said the main reason for using class d is its heat & power usage are much more stable

 

[Petebass]

......class d is its heat & power usage are much more stable

It sure is. My AI gets to a certain temp and stays there regardless of how loud I play or how hard I push it. It even maintains that same temp even if I'm not playing through it at all........

 

[pgk]

There was an thread last year from a CA9 believer that just didn't have the gains matched. The CA9 versus a properly designed - SMPS powered class-ab - is like comparing power chords. If a class-d with SMPS has similar measurements, beyond that of human hearing, young or old - it would be highly suspicious that a true double blind test that someone could hear or feel a difference.

If there was a "feel" difference, it can be recorded, it can be measured. You can capture the input signal and compare it to the output. The difference minus the obvious power gain is the coloration.

The protocols for the double blind is to then investigate why there was a statistical difference if there was one. Measurements can do this.

Popularity wins out over science, over and over. Musicians choice, word of mouth, win out over any real measurements. People like it when authority dismisses the science.

Many manufacturers are having great times with class-d.
There's a few variations in techniques, and along with SMPS it's leading a new round of amps that seem very popular. It will be hard to sell the old heavy stuff to anyone on top of things.

dude, you sound Exactly like Julian Hirsch from High Fidelity magazine in the 60's. he felt that if 2 things measured the same they will sound exactly the same, went to great lengths to "prove" it, only to be proven wrong time and again. when i was a kid i used to call it "listening with a slide rule". this argument is as old as the hills. older actually, and to each their own. as for me, i'd rather listen to music than look at it, and my ears don't lie

 

[okcrum]

I have designed class AB power amps with SMPS and conventional power supplies, and with the same amp design, the difference between the two supplies was not at all easily noticeable. This was with the NeoPak 3.5 and that convinced me of the future of SMPS. The SMPS was designed to emulate the characteristics of a traditional supply, but that wasn't a big deal. The point here is that the supply was the only difference.

So, identical amps with different supplies and the results were that the differences were very, very minor. Perhaps trivially so.

+1. The SMPS would need to have very poor transient response to have a noticable difference.

Edit: or just run out of current...

 

[seamonkey]

and my ears don't lie

There are optical illusions and there are audible illusions. Illusions aren't lies, but you're not seeing what you think you're seeing.

Test gear today has is far more accurate and reliable than ears, and not influenced by other parts of the brain. And we have years of historical data to show how interelated human hearing is. In "general" of course, as there will always be golden ears, just as there are going to be people hard of hearing.

 

[Bob Lee (QSC)]

dude, you sound Exactly like Julian Hirsch from High Fidelity magazine in the 60's. he felt that if 2 things measured the same they will sound exactly the same, went to great lengths to "prove" it, only to be proven wrong time and again.

Who proved him wrong?

 

[12bass]

As a proponent of the scientific method, I'd venture that, although everything is theoretically quantifiable, there are still some aspects of audio reproduction which are not easily measured. I can reliably tell the difference between op amps which are claimed to have less than 0.00004% THD. I'd also venture that all power amps do not sound exactly the same, even if volume matched.

 

[Bob Lee (QSC)]

It would seem the difference between those two op amps might not be described by that particular THD figure. Sometimes an op amp that is optimum for one circuit is not well-suited to another. I would be hesitant to use any op amp whose manufacturer provides only a THD spec, and nothing else.

All power amps don't sound the same, but the better ones do tend to.

 

[12bass]

Sure.... THD is only one specification. There are many others... IMD might be more useful. However, my point is that what makes components sound different is not always easily demonstrated in the specs. One such example is comparing an LM4562 with an OPA2211A.

 

[Bob Lee (QSC)]

Right, but there's a big difference between "specs" and "what can be measured." Usually the former is a minute subset of the latter.

 

[aborgman]

Right, but there's a big difference between "specs" and "what can be measured." Usually the former is a minute subset of the latter.

Often a tiny subset.

There is nothing we can hear that can't be measured.

There certainly are things we can hear that aren't regularly measured.

 

[pgk]

Who proved him wrong?

i think j gordon holt's theories were the other side of that coin, right?

 

[Medford Bassman]

My 2 cents...

I have owned a Shutle 6.0 for about 3 years now. I love it. I won't go into Genx Benx CS, as I think we all agree it's top notch.
As to the amp, I have been impressed with its power and tone. I've run this thing for hours prior, during and after gigs and it never heats up. Reliability? For me, 100%. I pair it with a NEOX 212 cabinet and have never had a problem.

Do tube amps sound different? Sure. I'm not convinced the difference is that great. It seems to me personal taste/bias really rules the day more so than hard facts.
I'm speaking as a guy who used Sunn 200watt amps back in the day with 2x15 and 4x12 cabinets.
For me, the sound difference is negligible, but my back and old *** sure feel better at the end of the night.

Again, just my 2 cents.

 

[Implosion]

As a proponent of the scientific method, I'd venture that, although everything is theoretically quantifiable, there are still some aspects of audio reproduction which are not easily measured. I can reliably tell the difference between op amps which are claimed to have less than 0.00004% THD. I'd also venture that all power amps do not sound exactly the same, even if volume matched.

On the topic

 

[Handyman]

On the topic

That's a pretty amusing video. Tube screamer clones are a pretty good choice for that experiment, since the op-amps are more or less still operating in their linear region (the "clipping" is caused by diodes in the feedback loop- they basically lower the gain as the output voltage increases).

Where you would likely hear a huge difference is in an application where the op-amp's output is saturating into the power rails. Different op-amp families can behave very differently when clipping, which isn't very surprising.

Of course, this doesn't have much to do with class D power amps.

 

[Bob Lee (QSC)]

i think j gordon holt's theories were the other side of that coin, right?

Did he have theories, or just hypotheses?

Did anyone actually prove Julian Hirsch wrong?

 

[12bass]

On the topic

Rather than using guitar pedals and amps, I audition parts in a high fidelity circuit and generally monitor through either K702s or monitor speakers. Although some differences may be audible through low fidelity guitar amplification, by using such a rig, a lot of the nuances are probably obscured by distortion and bandwidth limitations.

 

[fdeck]

Well, we never compared our class A/B with with two different power supplies, but I would agree with your opinion on the power supply. In our listening test we tried four totally different power sections, two class D, a class H and a class A/B, mated to the same preamp. In that instance, there was a very noticeable difference in sound and feel between the various topologies. There were no measurements taken and no frequency analysis done. I made the final choice for our amp strictly by ear.

Different strokes. If I noticed a difference, I'd be all over it with my test gear, and would probably not stop until I understood the root cause.

By day, I work with "instruments" of a different nature: Scientific instruments. If two designs for a particular type of instrument give different results, I sure as heck want to know why.

 

[pgk]

Did he have theories, or just hypotheses?

a pilosophy i think would be more accurate. if nothing else i think he probably got more people listening with their ears rather than buying something based solely on a spec sheet, for better or worse. to encourage people to really listen hard and take note of the differences one hears, which is always a good thing.

Did anyone actually prove Julian Hirsch wrong?

i don't know for certain, and so i spoke out of turn. apologies