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

[tmtomh]

Too me that goes without saying by now, but I guess not to you and probably others. No wonder we have 80 pages in less than a month.

Oh, I think it does go without saying to @rdenney - there are just a few other folks who won't let it go and keep insisting that full power at 20kHz for 5 minutes is required or else the amp is a toy and the vendor's specs are fraudulent.

As for Rick's point, which you and I and many others agree on, we already know that at least one amp vendor tests their amps with pink noise at full power for 30 minutes and they don't overheat, go into protection, shut down, or otherwise complain or falter.

So it seems to me that the reasonable, "goes without saying" path forward is simply to settle on a signal (or small set of signals) that is more akin to pink noise than it is to white noise or to a single sine wave, and use that signal - with the guiding principle that the signal, similarly to pink noise, is modeled upon real-world use cases. And of course the signal should be on the difficult/extreme end of use cases, to ensure some headroom/robustness/abuse tolerance.

 

[gnarly]

I think you missed the point, that many here on ASR are saying even 5 minutes as per the FTC current rule is unnecessarily long, may be a couple wants it longer, but not me, not most iirc. Edit: There are obviously others who think 5 minutes isn't that bad, or not long enough. I am not siding with either, just stating my understanding from many of the posts here so far.

For generators, any such rules will naturally have to be very different as their applications are very different, but if you think they are the same in principle, then okay. I see no point to debate on that and it is outside the scope of this thread anyway imo. So I would just agree to disagree.

In case someone wants posts to include the obvious, the 5 minutes I referred to was based on current FTC rule on test conditions:

https://www.ftc.gov/system/files/ftc_gov/pdf/amplifier_rule_-_final_rule_notice.pdf

I've made several comments in this thread saying I think the 5 min FTC rule is well...ridiculous. So can't say I missed that one

My point about generators, and I do think it's salient, is that there is a time structure to the AC power an amp needs
.....that matches the time structure of the amps output.

Here's an example of fairly efficient 4 channel AB/H amps, in each of a pair of self-powered speakers I own.

The 620W per channel is a burst mode spec footnoted as: "3. Nominal 8 Ω resistive load, pink noise, 100V peak."
So an FTC rating is maybe only somewhere between 200-300 W per channel.
Note the current draws highlighted in blue.

110V line measured sag on bursts, moved me to install 230V outlets for the speakers.
I can also tell you from doing live sound festival type work, where a genny is needed out in a field, any sound man worth his salt makes sure the generator meets both average and peak power calculations needed.

It really doesn't make sense to look at amps in a vacuum, ala FTC imo...

 

[peng]

Oh, I think it does go without saying to @rdenney - there are just a few other folks who won't let it go and keep insisting that full power at 20kHz for 5 minutes is required or else the amp is a toy and the vendor's specs are fraudulent.

As for Rick's point, which you and I and many others agree on, we already know that at least one amp vendor tests their amps with pink noise at full power for 30 minutes and they don't overheat, go into protection, shut down, or otherwise complain or falter.

So it seems to me that the reasonable, "goes without saying" path forward is simply to settle on a signal (or small set of signals) that is more akin to pink noise than it is to white noise or to a single sine wave, and use that signal - with the guiding principle that the signal, similarly to pink noise, is modeled upon real-world use cases. And of course the signal should be on the difficult/extreme end of use cases, to ensure some headroom/robustness/abuse tolerance.

Haha, to me they are indeed toys, but to many pro amp users, probably not.

 

[Sokel]

The speaker may not go that low , but the owner migth still play such tracks trough the system ? There probably stuff we never hear in our music due to speaker limitations, but its still there in the signal getting amplified .

Off topic thats what i like with active speakers they are self protecting, they dont try to do stuff they cant do even if its in the signal

Yes,take it from a mega-hit like Evanescence's "My Immortal" from the live version "Anywhere but Home" (or a remix,Band version or something,forgive me to be vague but I listen to those with my nephew,despite his classical studies ) where hell breaks loose after about half the song with full 30's and up to 100's at a constant rate with very low (I suspect) CR.
It will be played from small radios to mega-systems able to do 130dB (Z) SPL .And all the amps will try to follow,no matter desktop ones or full size or big boys.
Tests should cover all range.

 

[amirm]

Whether you or I agree with it doesn't really matter. I also totally agree, 8R is out of date, but 4R is not ubiquitous either.

I have tested hundreds of speakers now. By far, the average *minimum* impedance is around 3.X ohm. So even lower than 4 ohm.

8R requires amplifiers that can swing voltage, and that's a good thing.

As you know, if the current is not there, then the voltage is for not. And at impedances below 5 ohm, current dominates.

 

[mcdn]

…I think it's worth noting that Meyer appears to have developed m-noise for the purpose of testing loudspeakers ability to deliver uncompressed, unclipped music
Maybe it could be used for amps. But with its +18dB crest factor, I think it would be a much better test for peak voltage/wattage capacity, than for continuous thermal…

The crest factor is high because at high frequencies music has that characteristic, for example a snare drum. It is as far as I know the most accurate artificial model of a music signal that we have. There’s no reason not to use it for amp testing. In fact the high crest factor in the treble makes it much more realistic than pink noise.

 

[Old_School_Brad]

An amplifier must deliver the necessary signal, whether the speaker transforms it into sound waves or dissipates it as heat, making the speaker's theoretical roll-off point irrelevant.

As I previously mentioned, many home theater users configure their L+C mains in full-range mode, requiring them to handle deep bass. Personally, I also listen to numerous tracks with significant low-frequency content. Many EDM and related subgenres feature sustained bass throughout most of their runtime, and I’ve observed similar behavior in some orchestral pieces.

A representation like the power vs. time plot below could address everything from short bursts to long-term thermal equilibrium and continuous thermal capability. However, agreeing on a standardized test signal remains a challenge.

Adopting an existing rule, even if imperfect, might be the most practical approach. There will likely never be signals that perfectly represent real-life scenarios.
Sometimes, it’s best to accept what cannot be changed.

 

[amirm]

As I previously mentioned, many home theater users configure their L+C mains in full-range mode, requiring them to handle deep bass.

Very few do that. For them, they can choose to pay more to get beefier amps than a desktop unit like Fosi/AIYIMA. You can't possibly claim that a regulation that applies to every amplifier needs to be in place to satisfy these few.

This all goes back to wanting something for free. The thinking that if you force manufactures to design for pathological cases, you get better amps for free. That is not going to happen.

As I said above, when you have unique applications for power, like people who deploy large towers for LCR, you get powerful amplification to go with it. Don't look to the government to give you that in a $200 amplifier. The $200 amp is going to struggle to give you its full power at 20 Hz. Heck, most amplifiers will produce less power there for short periods let alone sustained for 5 minutes.

 

[amirm]

Sometimes, it’s best to accept what cannot be changed.

As i said at the outset, this regulation is going nowhere. So what is left is opinion of membership and pressure on reviewers to run such tests. Those opinions, can and certainly have been changed. Few like you excluded.

 

[Old_School_Brad]

Very few do that. For them, they can choose to pay more to get beefier amps than a desktop unit like Fosi/AIYIMA. You can't possibly claim that a regulation that applies to every amplifier needs to be in place to satisfy these few.

This all goes back to wanting something for free. The thinking that if you force manufactures to design for pathological cases, you get better amps for free. That is not going to happen.

As I said above, when you have unique applications for power, like people who deploy large towers for LCR, you get powerful amplification to go with it. Don't look to the government to give you that in a $200 amplifier. The $200 amp is going to struggle to give you its full power at 20 Hz. Heck, most amplifiers will produce less power there for short periods let alone sustained for 5 minutes.

I don’t get the impression that only a small number of people run full-range L+R speakers in a home theater setup. There’s a wide range of medium to large-sized tower speakers available on the market, and it’s reasonable to assume that many are indeed used in home theater systems. With and without subwoofers.

And no, this isn’t about wanting something for free. The FTC rating is simply a number on a spec sheet, not a requirement to redesign the equipment.

As i said at the outset, this regulation is going nowhere. So what is left is opinion of membership and pressure on reviewers to run such tests. Those opinions, can and certainly have been changed. Few like you excluded.

What are your thoughts on illustrating power over time in a graph? Or noting the duration of the signal across several select frequencies in the power tests within your current testing suite?

 

[amirm]

I don’t get the impression that only a small number of people run full-range L+R speakers in a home theater setup. There’s a wide range of medium to large-sized tower speakers available on the market, and it’s reasonable to assume that many are indeed used in home theater systems. With and without subwoofers.

Your impression is wrong. I have been in this space since its inception. I have tested tons of AVRs which by default come with 80 Hz high pass and their EQ will set most speakers to small. It is a massive exception to have LCRs set to large. The only people I have seen do that are people with speakers that are hundreds of pounds each and driven by appropriate amplification.

 

[amirm]

What are your thoughts on illustrating power over time in a graph? Or noting the duration of the signal across several select frequencies in the power tests within your current testing suite?

I already explained that I am not in favor of such things. The analyzer dwell time is automatic and variable. I don't even know how you do that with other analyzers.

 

[Old_School_Brad]

It is a massive exception to have LCRs set to large. The only people I have seen do that are people with speakers that are hundreds of pounds each and driven by appropriate amplification.

That’s quite different from my experience, which is closer to a 50/50 split.

Considering that there are many mid to large tower speakers available on the market, it seems unlikely that none of them will be used in full-range setups. I’d say that neither of us may be completely right or wrong.

I already explained that I am not in favor of such things. The analyzer dwell time is automatic and variable. I don't even know how you do that with other analyzers.

Well, that clears things up. Thank you for your response.
I’ll go ahead and turn off notifications for this lengthy thread now.

 

[amirm]

Considering that there are many mid to large tower speakers available on the market, it seems unlikely that none of them will be used in full-range setups. I’d say that neither of us may be completely right or wrong.

No, you are wrong. A full range speaker needs to handle down to 20 Hz. Most tower speakers won't remotely go there. Here is one of many references, this one from KEF: https://us.kef.com/blogs/news/large...' speaker is one,– bookshelf, satellites, etc.

"Obviously for home theater applications, your system should be able to reproduce 20Hz (or lower) for maximum effect. In order to get the full benefit of your AV receiver’s internal crossover, if you have a subwoofer your front (left, center, right) speakers should be set to ‘small’ – regardless of their capabilities. "

Since everyone has subs for their home theater, the above applies to almost entire population of users of such systems.

Of course you can set a lower crossover for large speakers but you won't be running them full bandwidth.

 

[mcdn]

I think a lot of people don't have a good intuition for what high continuous power dissipation looks like physically. These are pics of a real load that can sink 800W continuous with a 50C rise between ambient and heatsinks. In other words the air coming out is 70C if the air coming in is 20C. The load resistors almost certainly have hot spots up to 110C. To achieve this level of dissipation we have a 140mm fan delivering 100cfm (150m^3/hr) over a very large heatsink area in a tunnel enclosure.



98% of amp output power will be spent as heat in the speakers. Do speakers have fans?

 

[amirm]

That is indeed a major consideration for a reviewer like me. I am not about to make a sink that can dissipate 1000+ watts (stereo) for minutes.

 

[mcdn]

That is indeed a major consideration for a reviewer like me. I am not about to make a sink that can dissipate 1000+ watts (stereo) for minutes.

Luckily nobody really makes amps or speakers that can do that either The dummy load in the pics is easy to make for less than $100 and can sink 1kW continuously. There’s a build thread in the DIY forum.

 

[Old_School_Brad]

That is indeed a major consideration for a reviewer like me. I am not about to make a sink that can dissipate 1000+ watts (stereo) for minutes.

No amplifier you review would be able to achieve that, so I wouldn’t view it as a significant concern.

No, you are wrong. A full range speaker needs to handle down to 20 Hz. Most tower speakers won't remotely go there. Here is one of many references, this one from KEF: https://us.kef.com/blogs/news/large-or-small-setting-your-av-receiver-for-your-speakers#:~:text=A 'large' speaker is one,– bookshelf, satellites, etc.

"Obviously for home theater applications, your system should be able to reproduce 20Hz (or lower) for maximum effect. In order to get the full benefit of your AV receiver’s internal crossover, if you have a subwoofer your front (left, center, right) speakers should be set to ‘small’ – regardless of their capabilities. "

Since everyone has subs for their home theater, the above applies to almost entire population of users of such systems.

Of course you can set a lower crossover for large speakers but you won't be running them full bandwidth.

I noted that they will be utilized in a full-range setup, which is very different from being true full-range speakers.

Since not everyone has subwoofers for home theater or uses high-pass filters, both the speaker and the amplifier will operate as full-range, demanding the power needed to do full-range whether or not the speaker is able to make sound-waves or heat from it. Many people I know, whether they have a subwoofer or not, will occasionally run a 2.0 stereo configuration as well.

 

[SIY]

I am not about to make a sink that can dissipate 1000+ watts (stereo) for minutes.

And add in that the load must not change in properties during the test (despite an inevitable large temperature rise). This is a severe condition, and the usual power-resistors-on-heat-sink won't do.

 

[Sokel]

And add in that the load must not change in properties during the test (despite an inevitable large temperature rise). This is a severe condition, and the usual power-resistors-on-heat-sink won't do.

Just do it the old DIY way: