Let's stop calling Class D amplifiers "efficient" when they are not.

[restorer-john]

Having another bad day again, huh?

Here are some Japanese amps and their idle power drawn (source: Stereoplay magazine - Germany).
Yamaha M-5000 - 56W
Yamaha A-S3200 - 50W
Yamaha A-S1200 - 50W
Accuphase E 460 - 82W
Luxman L 509u - 77W
Luxman L-505uXII - 93W
Luxman SQ-N150 - 82W
Luxman L-550AX - 168W
Luxman MQ-300 - 154W
Luxman L-595ASE - 160W

Not to mention other brands from other countries.
Octave Audio Jubilee 300B - 331W
T+A M 40 HV Anniversary Edition - 220W
T+A PA 3100 HV + PS 3000 HV - 198W
Dan D'Agostino Momentum Integrated - 210W
Dan D'Agostino Progression Integrated Stream - 120W
Pass INT-250 - 240W
Rotel Michi M8 Mono - 93W
Krell Evolution 402e - 342W

Let's not even compare efficiencies when musical power is drawn!


+1

You want to compare full blown integrated amplifiers with active preamplifiers, micros, lighting/meters- you name it, with these bare bones module-in-a-box Class D power stages and claim a victory in the idle efficiency stakes? Come on!

 

[tmtomh]

Look to car audio for cooling solutions. Direct mounting of power devices to heatsink, nice parts and high power output.

Thanks, but I'm not looking for effective cooling solutions. I'm looking for informed commentary on whether @pma's claim is empirically true or not.

 

[DonR]

Modern electronics in general, not just Class D amplifiers, suffer from a deficit of repairability and longevity engineering. Everyone has been shaped to accept frequent upgrading either through marginal improvement, obsolescence or premature failure.

 

[restorer-john]

Is this the 70 watt TA model?

TA3650. I just grabbed it due to it being within easy reach. It's been sitting on the pile of subwoofers for a few years!

These test figures are accurate and correspond very well with the unit here:

So similar to the little Fosi but continuous FTC rated.

 

[dougi]

My March Audio P502 (Hypex NP502mp based) consumes ~0.2-0.3W at standby and ~25W at idle, with Australian AC voltage. Which seems to align with the data sheet (0.25W standby, ~30W combined power losses at idle).

 

[dougi]

My NAD C3050LE integrated with all of its DSP and Hypex UcD modules, consumes:

~ 0.2W in deep standby
~ 2W in normal standby
~15W on but with speakers "off"
~ 23W on with speakers set to "A".

The last two are interesting, as it seems it is doing something a bit more than operating a relay when selecting the speakers.

 

[restorer-john]

My March Audio P502 (Hypex NP502mp based) consumes ~0.2-0.3W at standby and ~25W at idle, with Australian AC voltage. Which seems to align with the data sheet (0.25W standby, ~30W combined power losses at idle).

There you go.

30W of heat in a mostly sealed, small black alloy box, with not much ability to dissipate those losses on a continuous basis, once the outer casework comes up to temp. The internal spot temperatures would shock you.

It is a recipe for failure in Australia, especially a hot, dry summer in Canberra.

 

[NoMoFoNo]

Modern electronics in general, not just Class D amplifiers, suffer from a deficit of repairability and longevity engineering. Everyone has been shaped to accept frequent upgrading either through marginal improvement, obsolescence or premature failure.

Yes, but it's very likely that this is because consumers want it that way and/or need it to be that way. People like new things, things that are different. Nobody wants the old, well-built but unfashionable furniture their mothers want to hand down. Nobody wants to drive the jalopy their father drove twenty years ago. Nobody wants to dress in the clothes of the previous generations. Hell, people spend tens of thousands to change out perfectly functional kitchens all of the time. There are audio hobbyists who hanker for the old, used audio gear from 40+ years ago, but the vast majority of consumers don't want that old stuff.

Second, wages have stagnated for 50 years, in the US at least. Consumers need goods that are cheap but modern and effective. That's why monstrosities like Wal-Mart exist.

 

[JSmith]

components sitting up against that sort of heat

Possible planned obsolescence or just poor design (too compact) in the earlier Ncore modules?

It is a recipe for failure in Australia, especially a hot, dry summer in Canberra.

Marble Bar may well be a contender there.


JSmith

 

[pseudoid]

+1
Thanks for the post, @restorer-john.

I went with a ClassD and it never gets turned off and always has music on. Don't care.

 

[SIY]

I would be curious to hear what other of our knowledgeable members think about this claim.

I strongly disagree. And I have several actual examples in my listening room (NCx500, GaN Systems, Axign, Nilai, Purifi...), all much smaller than AB equivalents and all run barely warm.

edit: Oh, and the nCORE amps built into my NAD M10.

 

[restorer-john]

Yes, but it's very likely that this is because consumers want it that way and/or need it to be that way. People like new things, things that are different. Nobody wants the old, well-built but unfashionable furniture their mothers want to hand down. Nobody wants to drive the jalopy their father drove twenty years ago. Nobody wants to dress in the clothes of the previous generations. Hell, people spend tens of thousands to change out perfectly functional kitchens all of the time. There are audio hobbyists who hanker for the old, used audio gear from 40+ years ago, but the vast majority of consumers don't want that old stuff.

I was only thinking the same thing yesterday as I was driving to deliver some (repaired class Ds actually- not kidding) HiFi to a member of ASR.

I'm the complete opposite. I love products that simply will not die. Things built to be handed on or donated to someone else.

Having owned and ran many retail businesses across many categories, I've come to really despise the way a few generations have had their expectations of durabilty and fitness for purpose rescaled to the point they have no expectation anything is going to last. They certainly don't think it's normal to fix something when it breaks.

Efficiency covers the entire life cycle as far as I am concerned, from manufacture to disposal.

 

[restorer-john]

I strongly disagree. And I have several actual examples in my listening room (NCx500, GaN Systems, Axign, Nilai, Purifi...), all much smaller than AB equivalents and all run barely warm.

edit: Oh, and the nCORE amps built into my NAD M10.

PMAs Hypex 252MP pulls around 30W (SMPS+2ch) at idle. They get really hot. He's bolted it on a decent heatsink/case IIRC.

 

[SIY]

PMAs Hypex 252MP pulls around 30W (SMPS+2ch) at idle. They get really hot. He's bolted it on a decent heatsink/case IIRC.

These amps don't, so Pavel's thermal management challenges have zero to do with Class D. Other designers have done much better in that respect.

 

[amper42]

I strongly disagree. And I have several actual examples in my listening room (NCx500, GaN Systems, Axign, Nilai, Purifi...), all much smaller than AB equivalents and all run barely warm.

edit: Oh, and the nCORE amps built into my NAD M10.

SIY,
How many watts does the NCx500 use while idle? Is the case warm to the touch after running 30 minutes?
Thank you!

 

[tmtomh]

I strongly disagree. And I have several actual examples in my listening room (NCx500, GaN Systems, Axign, Nilai, Purifi...), all much smaller than AB equivalents and all run barely warm.

edit: Oh, and the nCORE amps built into my NAD M10.

If I'm being honest, I already suspected this was the answer - but I don't think people should take my word for it, so I appreciate you weighing in as someone who can provide both expertise and empirical examples. Thanks!

 

[SIY]

SIY,
How many watts does the NCx500 use while idle? Is the case warm to the touch after running 30 minutes?
Thank you!

It's about 17W for a pair of modules plus power supply. Case gets slightly warm over hours of use.

 

[IPunchCholla]

Efficiency covers the entire life cycle as far as I am concerned, from manufacture to disposal.

Are there any total lifetime efficiency studies for these types of differences? I imagine not. But I would be curious to see the calculation for increased energy costs for larger heatsinks etc. for AB vs smaller almost everything for D come out to be. Even if we suppose Ds have a greatly reduced life, I’m not sure AB would fair that well.

As I’ve been playing around with DIYing an amp, the thermal calculations keep pushing me to D. A heat sink for an 8 watt A is barely (if that) adequate for a 125 watt AB and would be overkill (by about 3X) for a 150 watt D (TPA3255). While this heat sink would only cost around $30 (+shipping), it would still be the single most expensive component in the amp. It also drives the size for much of the rest of the case. All that extra material has to come from somewhere and all of it has to be shipped, both of which add energy consumption to the life of the product (and cost).

All else being equal, one can take use relative cost as a proxy for relative energy used to produce and ship the item, which would indicate class Ds use less energy overall in the current accounting environment. How much that would change if we charged for disposal up front, I don’t know.

 

[sam_adams]

30W of heat in a mostly sealed, small black alloy box

And therein lies the root cause of the problem. What is the fascination with making amplifiers—and other audio electronics—in tiny form factors. When and where did this start? Take the examples that @restorer-john posted and think for a moment what the thermal performance would be if the internals were in a larger, vented enclosure where convection cooling could take place. Would the temps displayed be as high if there was a natural, cooling breeze inside the box?

I get that some of this is driven by 'lifestyle' design or possibly some 'cultural thinking'—not singling anybody out here—but, really, why is it that everything has to be crammed into the smallest form factor possible—especially if it is to the detriment of the longevity of the device. It's not like anybody needs to carry around a 250WPC, class D amp in their pocket—although some of the usual suspects might.

 

[AdamG]

It's about 17W for a pair of modules plus power supply. Case gets slightly warm over hours of use.

I’m seeing the same thing and I have 3 module Nc502mp’s mounted in 1 larger enclosure. 2 of these six channel models. I don’t stack them and they are placed in the open not in a cabinet. I do have rear vented Aircom units on all my amps. At idle the amp enclosures heat up about 10 to 14 degrees above ambient room temperature. The Aircoms keep the air temps being pulled out at 84-90 degrees without running very hard or loud. Have always used mounted fan type cooling just because it’s good engineering to do what you reasonably can to minimize heat buildup.

Long term use failure from heat is data we don’t have yet, with the exception of a few failures already reported. But the proportion sample size is unknown. Maybe some of our resident Builders will share with us their failure numbers based on total sales numbers. Until then this is mostly speculation. Speculation based on decades of experience yes. No one is arguing John’s observations about build quality as compared to Legacy Amps.

I do know this. The class D amps use a fraction of the power, have a much smaller footprint, cost a fraction of legacy gear and operate at considerably lower heat producing loads. An equivalent legacy amp design of equal performance would cost 4 to 5 times the cost of a legacy AB amp. If I only get 4 or 5 years out of each amp kit. I can replace it with a new one and get another 5 years and I can do that 4 or 5 times and probably still save money in the long run. This doesn’t even take into consideration the reduction of heat load creation and removal costs. Where I live we run our Air conditioning systems 8 to 9 months a year.

Do we have any resident class D amp builders that are willing to put some numbers out there for us to use?