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

[Old_School_Brad]

Did you see they only want THD, not THD+N?

I initially interpret it as a proposal for retrospective application, aimed at protecting older amplifiers with high noise levels from being significantly derated?

Spoiler: Snippet

 

[restorer-john]

When I went from 1 lab workstation to 2 dedicated workstations things where wayyy better and I still could use one more and make more money. You have enough gear to populate 2 or 3 workstations and you might as well build it.

Not about money Doodski. I'm winding it all down and only doing things I haven't seen before as a challenge to help someone where I can. Or absolute classics like the TOTL receivers/amplifiers in the power war days. Basically, I'm sick of cleaning up other "technician's" disasters and won't do anything where someone else has had a go. Twice as much work, twice as much time. Not worth it for anyone.

SMD is truly horrible, especially when mixed with through hole and PbFree. Like this Hypex NCxxx (from the NAD M-22 on my bench). You have to remove the through hole capacitors (double sided board) all around this SOT23 (1.8mm) dual diode to get anywhere near it. It's only 10mm between all the caps, so they have to come out, either for direct iron or hot air.

 

[Old_School_Brad]

Not about money Doodski. I'm winding it all down and only doing things I haven't seen before as a challenge to help someone where I can. Or absolute classics like the TOTL receivers/amplifiers in the power war days. Basically, I'm sick of cleaning up other "technician's" disasters and won't do anything where someone else has had a go. Twice as much work, twice as much time. Not worth it for anyone.

SMD is truly horrible, especially when mixed with through hole and PbFree. Like this Hypex NCxxx (from the NAD M-22 on my bench). You have to remove the through hole capacitors (double sided board) all around this SOT23 (1.8mm) dual diode to get anywhere near it. It's only 10mm between all the caps, so they have to come out, either for direct iron or hot air.

View attachment 424956

That seems like a total nightmare for a hobbyist repairman like me. I would never attempt to repair something with such tiny components. Its position, being so close to the capacitors, adds to the frustration even more.

 

[Doodski]

I'm sick of cleaning up other "technician's" disasters and won't do anything where someone else has had a go. Twice as much work, twice as much time. Not worth it for anyone.

I worked mostly warranty service for large brands and some small importer/wholesaler operations as a Canada wide main warranty service depot and so I serviced all the stuff that other techs gave up working on or thought where too expensive or difficult to repair. So I know the fix the other techs rubbish and then get to the original fault(s)sort of repair service. It's grueling and more stressful work but I had many parts on hand, service literature for all the gear and familiarization going for me. That and the kind of gear that I was offering Canada wide warranty repair depot service paid well when out of warranty. The deal was with the small importers and wholesalers wanting me to be their main Canada service depot was they supply the parts I state are needed to be in stock which was a fair amount and when a out of warranty repair comes along I get the parts for free to sell. Like ordering HexFETs power devices and Darlington outputs in quantity of a thousand+ per at a time. So parts was a big money maker even more than usual. I never shied away from parts replacement and that enabled me to have a very low repeat ratio. The kind of work that you do is one-offs and that can be very grueling high risk work. Other than the warranty mechatronics work I did for the major players and high end car audio gear I would take in stuff like McIntosh, Audio Research, SAE etc and big Japanese amps/receivers/integrated amps/CD players/DAT and mini-disc and the nicer tape decks for service but the other stuff I refused because I was overloaded. I sympathize with your sentiments about the kind of service gear you experience. It's a tough go.

 

[Doodski]

That seems like a total nightmare for a hobbyist repairman like me. I would never attempt to repair something with such tiny components. Its position, being so close to the capacitors, adds to the frustration even more.

I serviced Sony gear in warranty for a bunch of years and there was many SMD PCBs combined in mechatronics gear. That is when things get hairy. LoL. The thing about this NAD M22 is it is both SMD and high current power devices too and it all has at some point the need to power it on and cross the fingers and hope it starts up and no smoke.

 

[restorer-john]

I initially interpret it as a proposal for retrospective application, aimed at protecting older amplifiers with high noise levels from being significantly derated?

Spoiler: Snippet

View attachment 424954

THD alone (without +N) will give them better numbers to advertise. The Japanese started doing THD only specs in the late 1980s when spectrum analysers became mainstream.

Much gear is limited by noise, especially a lot of the current crop of amplifiers. Also at high powers, the 'noise' from conventional PSUs can swamp the THD.

It's all self serving. Might as well A-weight it all too.

 

[jomo]

Might as well A-weight it all too.

Is there a good reason they don’t?

 

[restorer-john]

Is there a good reason they don’t?

Many companies see the value in A-weighting. There are arguments both ways (UNWTD vs A-WTD). A-WTD will always be a better number. I like both. Amir's UNWTD is the toughest/best and would trip up plenty of vintage pieces with significant mains spuriae. It's just clicking a button on the analyzer software and easy to have both, but one more thing to do.

Yamaha A-S3200 integrated.

 

[rdenney]

Which of the current peak power rating methods do you think comes closest to being clear and understandable for consumers?
Remember, the manufacturer can disclose as many ratings they choose.

I don't know. But in all programming, the dynamics vary, as does the spectral content, over MUCH shorter timeframes than five minutes. That variability can be modeled for specific examples of difficult material--brickwalled pop, movie sound effects, etc. In no case is actual content simulated by rated power being applied across all ten octaves for five minutes, or even at any frequencies at the extremes of those ten octaves for five full minutes.

Even with movie sound effects, it may be loud--0 dB full digital level played at rated power output--for some seconds, but it would not be so at extreme frequencies simply because that content would destroy consumer loudspeakers. We've heard of examples of sustained loud sounds in recordings, but not for more than a few seconds and not at 20 KHz (for example). And if the quietest bits of a brickwalled pop tune are, say, 6 dB down from the loudest peaks--which is very much brickwalled--the power needed for the quiet bits would one quarter what is needed for the loud bits. And, again, there's no way that content would be that loud at 20 KHz even if it is that loud between 50 and 5000 Hz, because it would destroy tweeters that usually have power ratings in the single digits of watts or a bit more.

Creating that model profile of dynamics and spectral content would be a good task for an expert working group convened to develop an appropriate standard for rating power that would actually be relevant to consumer use cases. Again, the acceleration and speed profile of the different fuel economy reporting requirements for cars provides an example of linking product performance estimates with realistic and typical use cases.

Yes, manufacturers can report different power levels. But the level required by law (whether or not that law is enforceable in practice) to be most prominently displayed should be most closely traced from realistic consumer use cases, because the point of an FTC rule is to provide market clarity for consumers poorly educated enough to need the FTC's help. People like us don't need the FTC.

Again, if you want a durability or heat-management standard, construct an appropriate standard. I would not oppose an indefinitely continuous test that demonstrated that at rated peak power output, those standardized profiles of dynamics and spectral content did not cause the active elements of the amplifier to exceed the temperature ratings of those elements. "Indefinite" could probably be defined as long enough for those temperatures to stabilize playing that content.

And also again, experts who are going to be more demanding won't need the FTC. Equipment intended for demanding applications that exceed what typical consumers would do would be specified more fully for that demanding customer base in any case, as live sound equipment is more fully specified than consumer equipment.

Rick "has said all this before, several times" Denney

 

[rdenney]

I should add that the continuous standard established in the 1974 requirement led, I think, to the common practice of reporting (but not defining) "headroom" or "music power" or "dynamic power" or whatever. That makes comparing amps and their ability to play music loudly really confusing and difficult, especially for poorly educated consumers. Is an amp rated at 125W continuous but with 3 dB "headroom" able to play music as loudly as an amp rated at 250W continuous but without any headroom? I think we can agree that for any realistic content, an amp rated at 250W continuous and 3 dB of "headroom" will be able to play music noticeably (just) more loudly than an amp rated at 250W but with no additional headroom. Wouldn't it make more sense for the first amp to be rated at 500W peak power playing the standardized music-like test signal versus the other amp rated at 250W playing that same standardized signal? The rating would reflect what consumers actually care about when looking at a power rating.

Rick "and then require heat management to control the component temperatures at the rated power" Denney

 

[rdenney]

Not about money Doodski. I'm winding it all down and only doing things I haven't seen before as a challenge to help someone where I can. Or absolute classics like the TOTL receivers/amplifiers in the power war days. Basically, I'm sick of cleaning up other "technician's" disasters and won't do anything where someone else has had a go. Twice as much work, twice as much time. Not worth it for anyone.

SMD is truly horrible, especially when mixed with through hole and PbFree. Like this Hypex NCxxx (from the NAD M-22 on my bench). You have to remove the through hole capacitors (double sided board) all around this SOT23 (1.8mm) dual diode to get anywhere near it. It's only 10mm between all the caps, so they have to come out, either for direct iron or hot air.

View attachment 424956

How is construction quality and serviceability indicated by power level testing of any sort?

Rick "that's what media tear-down and long-term usage reviews are for" Denney

 

[restorer-john]

How is construction quality and serviceability indicated by power level testing of any sort?

Did I say that? No. I did not. But being as you brought it up...

If an amplifier blows up during your proposal for an "indefinite" power output test it sure is an indicator isn't it? If it blows up trying to hit its specified rated power, surely the "construction quality" (your term) is clearly inadequate. If it blows up doing that, is it fit for purpose if you buy it based on the advertised specifications? Definitely not.

The only HiFi (not their professional range which was even better built) amplifiers I can think of that could output their rated power on an actual continuous (not time limited in any way) basis were Perreaux in the 1980s. They soak tested for 24 hours at full rated power. Scary stuff and I'm sure it kept the QC area very warm in winter back in NZ (Napier) in the early days. Big oil-cooled loads in drums. Wouldn't want the power bill for that today.

An "indefinite" continuous power rating is probably the worst idea ever for home HiFi amplifiers. It would have to take into account absolute worst case situations and de-rate everything. From total lack of airflow in closed cabinets, to placing on carpet or rugs, leaving magazines on top of the grilles, the unit being full of pet hair and dust, right through to massive variations in ambient temperature as well as mains voltage variations. Continuous 5 minutes at any frequency you desire from 20-20k into the specified load and distortion figure you want to advertise is perfectly fine. 5 minutes will either kill the amp or it will survive undamaged. Which amplifier do you want, the one that survives or the one that cannot handle 5 minutes?

 

[jomo]

Many companies see the value in A-weighting. There are arguments both ways (UNWTD vs A-WTD). A-WTD will always be a better number. I like both. Amir's UNWTD is the toughest/best and would trip up plenty of vintage pieces with significant mains spuriae. It's just clicking a button on the analyzer software and easy to have both, but one more thing to do.

I know the numbers look better for most situations, but it is certainly more realistic to compare hearable noise between products.

Is an amp rated at 125W continuous but with 3 dB "headroom" able to play music as loudly as an amp rated at 250W continuous but without any headroom? I think we can agree that for any realistic content, an amp rated at 250W continuous and 3 dB of "headroom" will be able to play music noticeably (just) more loudly than an amp rated at 250W but with no additional headroom.

IMO, peak power has become a more important parameter than it used to be as many modern amps have zero headroom or circuitry that shuts the amp down before it clips. As long as the test for peak power is properly characterized to model repetitive music peaks, it should have as much weight as the continuous power rating as far as usable power goes.

Continuous 5 minutes at any frequency you desire from 20-20k into the specified load and distortion figure you want to advertise is perfectly fine. 5 minutes will either kill the amp or it will survive undamaged. Which amplifier do you want, the one that survives or the one that cannot handle 5 minutes?

I agree, but not all amps are just amps. The Japanese manufacturers want to limit the test to a minute. I’m sure this is for nanny rich receivers, but they have a lot of other electronics in the same box that use nannies on the amplifiers just in case all those mentioned installation issues exist. Plain amplifiers shouldn’t have issue with the 5 minutes. If they do have issue, they should be more robust.

 

[restorer-john]

The Japanese manufacturers want to limit the test to a minute

At least they are advocating for one (1) minute.

Amir tests for an indeterminate and variable time, mostly in MILLISECONDS. In fact, he doesn't know how long any of his maximum power output tests are. They are utterly useless.

 

[RickS]

Not about money Doodski. I'm winding it all down and only doing things I haven't seen before as a challenge to help someone where I can. Or absolute classics like the TOTL receivers/amplifiers in the power war days. Basically, I'm sick of cleaning up other "technician's" disasters and won't do anything where someone else has had a go. Twice as much work, twice as much time. Not worth it for anyone.

SMD is truly horrible, especially when mixed with through hole and PbFree. Like this Hypex NCxxx (from the NAD M-22 on my bench). You have to remove the through hole capacitors (double sided board) all around this SOT23 (1.8mm) dual diode to get anywhere near it. It's only 10mm between all the caps, so they have to come out, either for direct iron or hot air.

View attachment 424956

I empathize, but likely was designed for replacement rather than repair. The module cost would not allow me to try to repair.

A number of these posts seems more appropriate in the owner's NAD thread. Did you intend them here?

 

[Old_School_Brad]

At least they are advocating for one (1) minute.

Amir tests for an indeterminate and variable time, mostly in MILLISECONDS. In fact, he doesn't know how long any of his maximum power output tests are. They are utterly useless.

I recall him mentioning that for bar graphs, it's a few seconds, while for sweeps, it's milliseconds. It's mentioned somewhere in this thread, but it seems the duration varies since the analyzer waits for stability each time.

 

[NTTY]

The only HiFi (not their professional range which was even better built) amplifiers I can think of that could output their rated power on an actual continuous (not time limited in any way) basis were Perreaux in the 1980s. They soak tested for 24 hours at full rated power. Scary stuff and I'm sure it kept the QC area very warm in winter back in NZ (Napier) in the early days. Big oil-cooled loads in drums. Wouldn't want the power bill for that today.

Hi John, I can test a Perreaux E2 (1989):

Worth cooking? For how long? 5 days?

 

[amirm]

A major reason companies have been able to avoid compliance with the amplifier rule for so long is, in my opinion, the ecosystem of bad-faith reviewers, undisclosed sponsorships, and permanent loan arrangements that seriously blur the line between independent evaluation, review and testing and paid promotion. The new FTC rules directly challenge these practices, prohibiting fake reviews, company-controlled "independent" review sites, and the suppression of negative feedback through legal threats or intimidation. This means manufacturers can no longer rely on hand-picked reviewers to echo their inflated claims without clear disclosure.

What? It doesn't do any of that. And the situation has gotten 100 times worse than it was back in 1970s where reviewers were highly technical and included measurements.

 

[amirm]

If you saw my lab bench at the moment, you'd understand.

If you only saw my backlog of gear to test, you'd understand as well.

 

[amirm]

A major reason companies have been able to avoid compliance with the amplifier rule for so long is, in my opinion, the ecosystem of bad-faith reviewers, undisclosed sponsorships, and permanent loan arrangements that seriously blur the line between independent evaluation, review and testing and paid promotion. The new FTC rules directly challenge these practices, prohibiting fake reviews, company-controlled "independent" review sites, and the suppression of negative feedback through legal threats or intimidation. This means manufacturers can no longer rely on hand-picked reviewers to echo their inflated claims without clear disclosure.

The "major" reason is much more likely to be apathy of mass market when it comes to technical specs such as watts. It has no meaning to them. I posted early in a thread that a Sony AVR I looked up did not even mention the output power in specs! How would a consumer know 50 watts from 100 watts???

The people who do care are enthusiasts and for them, they better read technical reviews like mine which have a wealth of information including reliable and detailed information about power output.

Back in 1974, hi-fi was only an enthusiast market. It hasn't been that way for decades now.