Power amplifier tests with respect to FTC: 16 CFR Part 432 (July 5, 2024) requirements on output power claims

[amirm]

Why do you think that was the case (pun intended)? You are looking at a die-cast, machined solid block of aluminium as a chassis.

Because hi-fi was very expensive then. And very profitable. Today, the same companies lose money on most of their hi-fi line. When I asked the top executives of those companies why they still sell them the answer was that of national pride. That it would be embarrassing for Japan to get out of AV business. This is the heatsink in Sony AVRs:

You think this folded steel, cheap heatsink will be happy during max power testing for 5 minutes???

 

[amirm]

I have Hypex NC-252MPs sent to me that failed in a few short years. Look at my posts here on ASR and details. They are garbage products and not fit for purpose.

Yet they have countless happy customers. It produces half a kilowatt of power (both channels combined) in a small, $570 retail package.

Meeting its advertised spec at very low distortion, not the FTC's 1% THD:

No, it can't produce 20 Hz at full power:

But that is not a scenario for anyone using it as they don't have speakers going down to 20 Hz and being mated to this amp.

EDIT: at 1% THD, it may be able to produce similar power at 20 Hz.

I don't know what is wrong with the one you have but having seen so many amplifiers of 1980s literally go up in smoke with massive areas of the power board charred beyond recognition, nothing is unique there.

 

[Sokel]

Yet they have countless happy customers. It produces half a kilowatt of power (both channels combined) in a small, $570 retail package.

Meeting its advertised spec at very low distortion, not the FTC's 1% THD:

No, it can't produce 20 Hz at full power:

But that is not a scenario for anyone using it as they don't have speakers going down to 20 Hz and being mated to this amp.

EDIT: at 1% THD, it may be able to produce similar power at 20 Hz.

I don't know what is wrong with the one you have but having seen so many amplifiers of 1980s literally go up in smoke with massive areas of the power board charred beyond recognition, nothing is unique there.

You're right,20Hz does not make sense for speakers.
But the exact same module on a plate is used for subs too.It says so at their product page :

Hypex FusionAmp FA252 with on-board SMPS and DSP

The Hypex FA252 has two 250 Wrms Ncore® amplifiers which can be bridged to form one 500 Wrms amplifier. Output power: 4Ω: 2x 250W, 8Ω: 2x 150W

www.hypex.nl

That plate can't be larger and thicker than the Buckeye case.
That's what we are saying,we never know the application,music,signals and use is too broad.

 

[amirm]

I am in favor of standardized measurements to bring more clarity. It is the darn 5 minute runtime that is the problem with FTC.

We are talking subs. But let's talk about 20 kHz. Who needs full power at 20 kHz? Or 10 kHz? Or anything over 1 kHz? The tweeter will blow up before and/or after your ears do!

 

[Sokel]

I am in favor of standardized measurements to bring more clarity. It is the darn 5 minute runtime that is the problem with FTC.

We are talking subs. But let's talk about 20 kHz. Who needs full power at 20 kHz? Or 10 kHz? Or anything over 1 kHz? The tweeter will blow up before and/or after your ears do!

It's been 2-3 days I read a post about the tactile feel of 15Hz?To feel it down there needs a gazillion of watts!
HC people are crazy

And on top of that they DSP the hell out of them filling dips!

It's the most demanding application for lows,if honesty is needed for speakers,for subs not even the raw truth is enough.
And most of them use such plates.

Edit:for 20kHz agreed,even 10-20W is too much.And good luck finding a cheap generator with no much THD+N up there to feed the test,it would need to be analog with 200kHz BW,clean all the way up.
AP can maybe do it,or something specialized,but...

 

[olieb]

It's the most demanding application for lows,if honesty is needed for speakers,for subs not even the raw truth is enough.
And most of them use such plates.

Isn't that interesting? It seems these plates work after all.
And it is not a surprise, as even the craziest action movie does not have extended explosions for hours. So the plate amp has only to provide for (fractions of) seconds once in a while.
It would be different, if it were to drive a massage chair on full power for a whole session, but that is not what audio is about. On the other hand, a close up jackhammer video might get the amp into trouble. So maybe it is a question of taste in the end. But even with a jackhammer the average power will be much less than peak power.

And for subs, it needs a VERY big driver to not bottom out at 20 Hz with 14A_peak, which is what the Hypex will deliver (at 1% THD) at ≈ 400W (=2x200W)/4Ω.

 

[Sokel]

Isn't that interesting? It seems these plates work after all.
And it is not a surprise, as even the craziest action movie does not have extended explosions for hours. So the plate amp has only to provide for (fractions of) seconds once in a while.
It would be different, if it were to drive a massage chair on full power for a whole session, but that is not what audio is about. On the other hand, a close up jackhammer video might get the amp into trouble. So maybe it is a question of taste in the end. But even with a jackhammer the average power will be much less than peak power.

And for subs, it needs a VERY big driver to not bottom out at 20 Hz with 14A_peak, which is what the Hypex will deliver (at 1% THD) at ≈ 400W (=2x200W)/4Ω.

Anecdotal,but seems that Interstellar can push things,exactly at the application you write (BTL) and we're talking about the big brother :

Power amplifier tests with respect to FTC: 16 CFR Part 432 (July 5, 2024) requirements on output power claims

The 5 minute test should be done at rated power which is continuous average power which is clearly stated on the NC252MP data sheet as 50 watts not 250W / 4 ohm that is Max output. 50W is stated, in the datasheet, as a continuous power. Other power definitions in the datasheet remain quite...

www.audiosciencereview.com

(don't know used drivers but judging by the double array set up they must be nice ones although the test was with resistive load)

 

[mdsimon2]

Quick question re ASR amp testing for Amir, or any who know for sure...
What is the duration of the sweeps used for the Power Versus Distortion @ (4 or 8 Ohm) Both Channels Driven ?
Like in this snip of the Buckeye amp...
Thx.
View attachment 413435

I don't think this thread has explicitly answered your question, but Amir gave me the answer below when I asked the same question a few years back -> https://www.audiosciencereview.com/...lifier-with-sinad-over-120.31305/post-1151445.

Michael

 

[RickS]

The regs (as I understand them) require the device to go into standby automatically after a period of non-use. I don't think working from a trigger input meets that requirement but auto power off can if compliant with the required timings.

Would be good to know as this may be more important to EU customers than FTC testing. Does having Auto on/off as an option qualify? As I indicated, not all Hypex modules provide the same interface. If I was not shipping into the EU, would be easier to use the muting function as it is common across more designs.

 

[Mnyb]

Yet they have countless happy customers. It produces half a kilowatt of power (both channels combined) in a small, $570 retail package.

Meeting its advertised spec at very low distortion, not the FTC's 1% THD:

No, it can't produce 20 Hz at full power:

But that is not a scenario for anyone using it as they don't have speakers going down to 20 Hz and being mated to this amp.

EDIT: at 1% THD, it may be able to produce similar power at 20 Hz.

I don't know what is wrong with the one you have but having seen so many amplifiers of 1980s literally go up in smoke with massive areas of the power board charred beyond recognition, nothing is unique there.

Can the 20Hz issue be a power supply thing this particular module has integrated PSU . Some other hypex modules requires a separate PSU

 

[rdenney]

Idle power consumption of the Hypex NC252MP module (amp + SMPS) is 17W, measured. This may be more or less than in case of class AB amplifier, depending on its designed idle current. The advantage of class D lower power consumption occurs only at higher power. In case of middle volume listening room conditions, there is no advantage of class D power consumption.

The idle consumption of my B&K Reference 125.2 Class AB amps, which are similar in capability to make music loud to the NC252MP, is 60 watts (according to B&K). At middle volume listening room conditions (whatever those are in my living room), the B&K amp produces substantially more heat than my NC502MP Buckeye, which has about twice the ability to make music loud when turned up. At idle, too.

Rick “would like to know what specific input signal protocol you and John are using for your tests, in terms of frequency and duration” Denney

 

[pma]

You are showing just one example of class AB amplifier, and yes, usual idle power of similarly rated amp will be like this. But it is not a rule and the idle power depends on output stage design and idle current. With CFP output stage the optimum idle current is reduced to fraction of that of 2EF or 3EF and the idle power consumption may be lower than that of the NC module. It is not an opinion, it is a fact. And even with 2EF together with non conventional FB compensation you may get about 5mA per pair and excellent distortion numbers up to THD 20kHz.

Regarding tests - my design philosophy is that the amp must able to give the rated power at any audio frequency continuously, without time limit. And I use all kind of known test signals from single tones through twin tones, tripple tones to multitones. I also use white noise generated by analog generator with BW of several hundred kHz. I also test with full swing squares.

That said, one may see that my requirements are very different from those that are accepted here as sufficient. For this reason, I resign to further arguing and explanations here, it does not make any sense. These are two different worlds, from engineering view. The worst case approach (including error state of signal sources) versus sufficient approach based on supposed average music signal properties. So, the agreement is impossible.

 

[antcollinet]

Would be good to know as this may be more important to EU customers than FTC testing. Does having Auto on/off as an option qualify? As I indicated, not all Hypex modules provide the same interface. If I was not shipping into the EU, would be easier to use muting function as it is common across amore designs.

There is a good guidance doc to the current requirements at the bottom of this page:

Standby, networked standby and Off mode

Off mode, standby and networked standby mode. Energy efficiency of products.

energy-efficient-products.ec.europa.eu

Bear in mind there has been a 2023 update which comes into force in 2025. I've not so far found a similar guideline document for the updates - but they are more stringent.

 

[restorer-john]

Regarding tests - my design philosophy is that the amp must able to give the rated power at any audio frequency continuously, without time limit.

Same for me.

Sources are analyzers, sine/square audio sig gen, DACs and Arb. Waveform Generator. For continuous tests at rated distortion, its the QA-403's generator/frequency counter for sanity and either an analogue THD analyzer (quick) or for low distortion products, the QA's analyzer.

For clipping or this new 1% THD, nothing is critical, I can use the oldest and biggest 8R 500W loads and the leader THD analyzer as it has a tunable 20-20kHz notch with a 0.002% residual THD.

The FTC has made it easy for the industry with 1% THD for continuous tests. 1% is well into clipping. And if an amplifier doesn't clip due to it's design or protection? Test it at its rated continuous power for 5 minutes at any frequency you want, sine or square waves, with or without reactive loads.

Here's a series of tests Perreaux required in the S/M for one of their baby power amplifiers, the PMF-2150B (200W+200W I listed further back in this thread) that would likely vaporize most of the little pretend watt boxes Amir likes so much.

They even provided a schematic to build their recommended switchable load box, such was the effort companies went to back in the 1980s. A different world with products built to last. If you ever get a chance, take the lid off a Perreaux PMF-xxxx series power amplifier. The output zobel inductor is bigger than most little Chinese amplifier boxes.

PMF-5150B. Rows of Hitachi MOSFETs, 12 per channel.

My best mate just picked up a couple of 8000C Perreauxes for close to free- as-is. 500wpc @8R continuous and 900wpc continuous at 4R. He asked me to look at these too. (he's got into Perreaux) I can't test these at full power as I don't have the loads to handle over 1kW per channel on a continuous basis. I'd need drums of oil! One needed a new power switch (they arc out) and the other needed a new fan. They are over 40 years old and work perfectly. After coming out of a club installed in a 19" rack for 4 decades. They are basically the same as the PMF-xxxx series, just with a pair of gain controls, LED meters and a fan for extreme use.

You don't even need the fans to be honest, unless you were running a nightclub- they have enormous heatsinks down each side.

Like this (this is an 8000B). See those silver cans at the rear? They are proper balanced line transformers.

 

[pma]

So we have maximum, worst case approach (@restorer-john , @pma ), vs. minimum just sufficient approach based on average music signal assumption (@amirm ). We thought that preconditioning + 5 min test at rated power would be the acceptable compromise for both parties, but, as evident, it is not. So, @amirm , please continue with time undefined level sweeps as always, there will be no more interventions from our side. Your readers are happy with the status quo and this is important.

 

[sernikus]

Apples vs oranges.... cheap, all-in-one module vs.... what ? And why ? Nonsense.

 

[antcollinet]

Apples vs oranges.... cheap, all-in-one module vs.... what ? And why ? Nonsense.

Please quote what you are replying to.

 

[JustJones]

Audioholics was involved in bringing awareness to the sunsetting of the amplifier regulations.

Be careful what you wish for you just might get it.(Aesop's Fables)

Now they're trying to get them to reopen the ruling to make changes.

Audioholics Petition to Reopen FTC Final Amplifier Rule of 2024​

I wonder how that's going? Comments have ended. Has enough time passed for FTC response?

Audioholics Petition to Reopen FTC Final Amplifier Rule of 2024

Our petition urges the FTC to reconsider its August 2024 Amplifier Rule, which expands power output regulations to ALL audio products. The broad, retroactive scope and blanket approach risk industry disruption.

www.audioholics.com

 

[restorer-john]

Now they're trying to get them to reopen the ruling to make changes.

Audioholics Petition to Reopen FTC Final Amplifier Rule of 2024​

I wonder how that's going? Comments have ended. Has enough time passed for FTC response?

Another clown show. The rules are clear.

If you want to advertise any advertised power output above 2W, you must adhere to FTC guidelines. End of story. They forgot to actually read the regulations.

 

[tmtomh]

So we have maximum, worst case approach (@restorer-john , @pma ), vs. minimum just sufficient approach based on average music signal assumption (@amirm ). We thought that preconditioning + 5 min test at rated power would be the acceptable compromise for both parties, but, as evident, it is not. So, @amirm , please continue with time undefined level sweeps as always, there will be no more interventions from our side. Your readers are happy with the status quo and this is important.

Such a self-serving, bull**** mischaracterization of @amirm ’s approach and testing regimen (not to mention an overly flattering characterization of your own)