How true? Class D amps can't sustain their maximum spec'd power

[Gorgonzola]

A few times I've heard pundits insist that class D amps can't actually sustain their advertised power, at least not without additional cooling with larger heatsinks and/or cooling fans. Exactly how true is this?

These critics seem to me to saying in effect that class D amps' are falsely rated by their makers and also that this is less often the case for A/B amps.

I can't tell from my personal experience with five different class D amps since I've never driven my speakers to levels high enough to need such sustained power as those amps claimed.

 

[Matias]

This had been debated ad nauseum already here in ASR....

 

[Doodski]

These critics seem to me to saying in effect that class D amps' are falsely rated by their makers and also that this is less often the case for A/B amps.

I have never seen a class AB amplifier that did not put out to it's specification. Most do a little better actually. I've RMS power tested hundreds of different models of home and car audio class AB amplifiers.

 

[DVDdoug]

I have never seen a class AB amplifier that did not put out to it's specification.

Interesting... I had the impression everybody was "fudging" these days.

A few times I've heard pundits insist that class D amps can't actually sustain their advertised power, at least not without additional cooling with larger heatsinks and/or cooling fans. Exactly how true is this?

There's no scientific or engineering reason that class-D amplifiers can't sustain power. And of course, it's up to the designer to build it with adequate cooling/heatsinking. And since they don't generate as much heat they don't need as much cooling.

A class-D amplifier works on similar principals as a switching power supply and most modern power supplies (not just in audio applications) are now switching designs (at any power level).

 

[MakeMineVinyl]

Class D amps (and class A/B amps) are primarily limited in their long term full output by thermal considerations. Class D amplifiers do need adequate heatsinking for sustained power, and this is clearly stated in the nCore data sheets for example. Class A/B amps obviously also need enough heat sink capability to keep up with the dissipation of the output devices. Fans are increasingly being used in amplifiers due to various regulations and realities.

Stating power output used to be more straightforward in the days of FTC regulations and stereo only amplifiers. Now with multi-channel receivers and amplifiers there are a multitude of complications and lack of standardization.

This is yet another reason I prefer tubes; there is none of that thermal-dissipation-limiting-full-power-output nonsense.

 

[MaxRockbin]

Just one example, but probably as close to an Apples to Apples as you will find. If you look at the ASR review of the Pioneer SC-61 AVR (class D), because the unit was defective, Pioneer shipped Amir a very similar newer model that happens to be class A/B - the VSX-LX504, both are rated at 130W into 8 ohms. "
I am very surprised as the nearly identical graph for the SC-61 and VSX-LX504. They have identical output power too. But of course the VSX-LX504 can only produce high power levels for 30 seconds or so after which it refuses to go higher (dashed teal line). Not the case with SC-61 which could do so forever.
...
And after completing the 8 ohm distortion testing: I should say that by this time in the test, the SC-61 was almost room temperature. It is so much more efficient.

Obviously this is not conclusive. Some Class D amps do have poor cooling. This Pioneer just isn't one of those. It says nothing about whether Class D amps as a group tend to have less adequate cooling than A/B. That might be the case because it's fun to make a 200W amp the size of a John Grisham paperback.

 

[Gorgonzola]

Thanks, everyone, for comments to this point.

So far what I'm hearing is that both class D and A/B amps are subject to thermal limitations and both might need some additional heatsinking and/or fans to achieve specified maximum outputs for sustained periods, especially since the "RMS" is no longer enforced vis-à-vis specifications.

At a high-end audiophile site I visit once in a while, somebody asked for amp recommendations to handle "very" low impedance and "very" insensitive speakers; (no examples were given of such a speaker). When I suggested that (e.g.) Purifi modules could deliver over 400 watts to 4 ohms and were stable at 2 ohms (with a supremely high damping factor). One respondent, (who implied extensive personal experience), curtly replied that class D amps were barely up to the task at 8 ohms much less at lower impedance. There was the strong insinuation that A/B amps in general are more realistically rated that class D's.

The question I also asked there was, How many consumer speakers actually have such low impedance, (<2 Ohm), combined with very low sensitivity, (say <80 dB/1watt@1 m). Nobody came up with an example.

 

[thorvat]

Thanks, everyone, for comments to this point.

So far what I'm hearing is that both class D and A/B amps are subject to thermal limitations and both might need some additional heatsinking and/or fans to achieve specified maximum outputs for sustained periods, especially since the "RMS" is no longer enforced vis-à-vis specifications.

While it is true that most of the amps indeed can't sustain delivering declared max power for extended period of time some of them certainly can. If that is of importance to you try looking for THX certified amps as that is one of the requirements to pass the certification test. I would also expect some of the non-certified amps to be able to achieve that but it is impossible to tell without testing.

P.S. The name of the cheese on your avatar pic is gorgonzola, not gorganzola.

 

[Koeitje]

Its temperature limited. In practice it doesn't matter, because no content uses full power all the time.

 

[anphex]

Thanks, everyone, for comments to this point.

So far what I'm hearing is that both class D and A/B amps are subject to thermal limitations and both might need some additional heatsinking and/or fans to achieve specified maximum outputs for sustained periods, especially since the "RMS" is no longer enforced vis-à-vis specifications.

At a high-end audiophile site I visit once in a while, somebody asked for amp recommendations to handle "very" low impedance and "very" insensitive speakers; (no examples were given of such a speaker). When I suggested that (e.g.) Purifi modules could deliver over 400 watts to 4 ohms and were stable at 2 ohms (with a supremely high damping factor). One respondent, (who implied extensive personal experience), curtly replied that class D amps were barely up to the task at 8 ohms much less at lower impedance. There was the strong insinuation that A/B amps in general are more realistically rated that class D's.

The question I also asked there was, How many consumer speakers actually have such low impedance, (<2 Ohm), combined with very low sensitivity, (say <80 dB/1watt@1 m). Nobody came up with an example.

I too still have to find one of those. But I think no manufacturer would actively go for this range since the 4 Ohm is the most widespread in amps. I guess it just happens in cases where a badly engineered speaker has an Ohm drop in some place of the frequency range to 2 Ohm or even lower and causes distortions very early. So you'd use a good, stable amp to "save" the performance of a bad speaker.

As for the long term rated output I remember reading somewhere that NC400 for example drops to 75WRMS after about 40 seconds of 400 Watt. Then again, would you consider this unrealistic scenario (running an amp on the physical edge for a long time) worthy of adjusting the rated power spec?

 

[tomtoo]

I think many consumer amps run into thermal problems driven full power a long time.
If you like to do this, better get a pro amp.

Edit says: Independent from the class

 

[MaxRockbin]

Regarding impedance - many speakers do drop below 4 ohms (usually in a narrow frequency range). The reviews on this site and most sites with measurements will give you a nice graph. Also on that graph you will typically see phase as well. In a nutshell, if you have a high phase angle when impedance is low. Say 25degrees+ with <4 ohms for example, it will be a more demanding load on the amp than 0 degrees.

Also, except for many tube amps, I haven't seen too many with an inadequate damping factor for driving most conventional speakers. There are always outliers either way. And probably the amps that aren't so great won't be bragging about it.

 

[Pdxwayne]

I have Pioneer SC95. Using all speakers for music playback in surround mode, crossing at 80hz vs 50hz have obvious difference in the highs. It sounded more open with 80hz crossover. Power related?

 

[Koeitje]

I have Pioneer SC95. Using all speakers for music playback in surround mode, crossing at 80hz vs 50hz have obvious difference in the highs. It sounded more open with 80hz crossover. Power related?

Points at signature...

 

[Pdxwayne]

Points at signature...

Care to explain more? Thanks!

 

[levimax]

Slightly off topic but related. From reading Amir's reviews it seems like most class D amps use regulated power supplies which limits their "peak" output compared to AB amps using linear supplies. Is there a reason class D amps need regulated supplies?

 

[tomtoo]

Slightly off topic but related. From reading Amir's reviews it seems like most class D amps use regulated power supplies which limits their "peak" output compared to AB amps using linear supplies. Is there a reason class D amps need regulated supplies?

There is no "need" they also work with linear power supplys. But they like regulated voltage.
If you like power without heat( efficience), you go the way with a switching PSU, thats why you see them usually as pairs.

 

[anphex]

Slightly off topic but related. From reading Amir's reviews it seems like most class D amps use regulated power supplies which limits their "peak" output compared to AB amps using linear supplies. Is there a reason class D amps need regulated supplies?

No, but it would make little sense using efficient class d with inefficient supplies.

 

[Gorgonzola]

...
P.S. The name of the cheese on your avatar pic is gorgonzola, not gorganzola.

Yes, I know. It was just my spelling mistake.

 

[MaxRockbin]

I have Pioneer SC95. Using all speakers for music playback in surround mode, crossing at 80hz vs 50hz have obvious difference in the highs. It sounded more open with 80hz crossover. Power related?

Without knowing anything about your speakers, your room, your listening distance or the nature of dark energy, I can say there is a reason 80hz is the more typical crossover to subs than 50hz. At 50 you are probably at or near the limit of your speakers' best qualities. And a 50hz (or any) crossover is not instant. It doesn't suddenly flip from main speakers to sub. So you're asking your main/satellites to play well below 50hz with diminishing volume. I cross mine at 100hz though my speakers are supposed to play down to 50 without much distortion (at my listening levels, position etc).

You may also have a speaker boundary interference thing going on between your subs & main speakers (search this site for speaker boundary interference). That will vary with position and frequency. But crossing higher will often spare you that.