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

[pjug]

And talking about time,watching the RTA at real time analyzing it it had at least a good 10 constant,continuous seconds 50Hz at -9db.
Imagine been in Europe combining that with hum or a weak amp hitting the rails

(old_man_mode: kids these days )

How did you choose that track? I would not know how to begin on that site; it is a foreign world to me (another old man)

 

[Sokel]

How did you choose that track? I would not know how to begin on that site; it is a foreign world to me (another old man)

I just followed the line you gave and searched for EDM mastering to see their trend.
It seems that they are even using/aiming for white noise up to a point.

From what I read I expect to see worst examples than that.

Look at the images at the reddit thread.

 

[pma]

Example of huge overdrive of the amplifier by white noise. Do not try it unless you have a bullet proof amplifier.

 

[Tell]

Some guys are trying all kinds of excuses for the amps that are unable to pass tests at rated power at frequencies from 20Hz to 20kHz range. Lobby pressure of manufacturers and OEM assemblers, who are unwilling to derate advertised power specs to realistic numbers. It is understandable from business and marketing point, less from the customer point, where it is usually a protection of emotionally loved component. And that ridiculous "music signal" argument, something that is totally undefined.

Would you mind telling us when we'd ever need to have an amplifier that can put out max power at 20khz for 5 full minutes? There is simply no use ever for this for audio so why should anyone ever care?

 

[tmtomh]

Some guys are trying all kinds of excuses for the amps that are unable to pass tests at rated power at frequencies from 20Hz to 20kHz range. Lobby pressure of manufacturers and OEM assemblers, who are unwilling to derate advertised power specs to realistic numbers. It is understandable from business and marketing point, less from the customer point, where it is usually a protection of emotionally loved component. And that ridiculous "music signal" argument, something that is totally undefined.

If the derated power figures are based on full power at 20kHz, how you can call them "realistic"?

This is precisely the same issue @Tell raises in the prior comment, and that numerous other members here have posed to you - and you haven't provided a coherent answer.

 

[amirm]

Example of huge overdrive of the amplifier by white noise. Do not try it unless you have a bullet proof amplifier.

It is not bullet proof unless you subject it to such testing over and over again during a period of years. Otherwise you could have just shortened its life during such testing.

 

[MaxwellsEq]

One from JEITA

Regulations.gov

www.regulations.gov

Masimo

Regulations.gov

www.regulations.gov

Harman

Regulations.gov

www.regulations.gov

Thank you - interesting and intelligent comment from vendors who can see they would face issues even trying to measure this, let alone comply. I particularly admired JEITA's thoughtful response, especially their point about any device which has steep high and low pass filter (e.g. a 100W device that's ONLY flat from 22Hz to 19.5kHz) needing to be defined as Zero Watts. Clearly any amplifier that has a filter ahead of it (e.g. as part of an active speaker, but one where the amp is not built in) will be Zero Watts.

Poor consumers, they really are not going to understand this, are they?

 

[theREALdotnet]

Lay off the aluminum cookware.

Why do you keep dragging this discussion into the gutter with comments like that? Very disappointing behaviour from a "Grand Contributor".

 

[MRC01]

When clipping occurs, it's a sudden waveform transition that creates HF energy not in the signal. This HF energy can kill tweeters. This is the origin of the notion that small amplifiers are more likely to kill tweeters than big ones.

This is one of those long prevailing audio myths, I think because folks look at the FFT of a clipped sine wave and see that the harmonic content increases.
They then assume this harmonic content goes to the tweeter in destructive fashion.
But the fact is the harmonic content is so low in level, dB relative, that it adds little to the signal passing to the tweeter. ...

Myth or not, it has a thread of truth. Play a 100 Hz sine wave and turn up the volume to a loud level close to the limits of the speaker. Now keep the volume knob where it is and replace it with a square wave. Do you think the tweeter will survive? I doubt it. Sure, actual clipping may not be a square wave and the high frequencies may be lower in amplitude. But they will be introduced. So the degree of relevance depends on the clipping behavior of any particular amp.

 

[restorer-john]

Would you mind telling us when we'd ever need to have an amplifier that can put out max power at 20khz for 5 full minutes? There is simply no use ever for this for audio so why should anyone ever care?

HiFi audio amplifiers are supposed to be linear devices.
That means they treat all input signals the same way.
If they have gain, they apply an equal amount of gain to all frequencies.
They do that, whilst not adding unrelated signals of their own.

A power amplifier's job is to produce power (by swinging voltage and delivering current).
Lots of continuous linear power.
Well designed power amplifiers don't care what frequency they are dealing with- all frequencies are treated the same.
As much power as you want, for as long as you want.

A measly 5 minutes at full rated power, at any frequency across the audible bandwidth is not unreasonable. In fact it's not much of a hurdle at all for a well designed amplifier.

Amplifiers that cannot deliver their rated power across the audible spectrum on a continuous basis are poorly designed and/or over specified. Testing to FTC conditions often uncovers this. How is that not useful to consumers? And which type do you want to buy- the poorly designed, the over specified, or the better alternative?

 

[restorer-john]

Why do you keep dragging this discussion into the gutter with comments like that? Very disappointing behaviour from a "Grand Contributor".

What is he even talking about? Explain?

 

[olieb]

Play a 100 Hz sine wave and turn up the volume to a loud level close to the limits of the speaker. Now keep the volume knob where it is and replace it with a square wave. Do you think the tweeter will survive?

Even if you had a perfect square wave - and that will not happen by a long shot - the amplitude of the 21st harmonic (2100 Hz) would be 1/21 of the base form (100 Hz). That is over 26 dB down. So it is only 1/400 of the output power of the amp in case the impedance is the same everywhere..
If you add up all harmonics that will be sent to the tweeter (21, 23, 25, ....199), this will amount to about 2-3% of amp power.
With any kind of real clipped 100 Hz signal (more trapezoidal) I would expect your tweeter to be rather safe.

Well, unless of course you use a 1000W continuous amp on a tiny bookshelf in a 24h test.

 

[gnarly]

Myth or not, it has a thread of truth. Play a 100 Hz sine wave and turn up the volume to a loud level close to the limits of the speaker. Now keep the volume knob where it is and replace it with a square wave. Do you think the tweeter will survive? I doubt it. Sure, actual clipping may not be a square wave and the high frequencies may be lower in amplitude. But they will be introduced. So the degree of relevance depends on the clipping behavior of any particular amp.

It does have a small thread of truth...we know the harmonics of a square wave will produce HF content. As will clipping, but not as severe as a square wave.
Here's a line level measurement of a 100Hz square wave sent through a DSP crossover, showing relative drive levels for a two-way speaker.
Crossover is a Linkwitz-Riley 12 dB/oct @3000Hz. So tweeter is handling 3kHz up.

The top meter is the tweeter side of the crossover, and is measuring the level of the 100Hz square harmonics going through it.....it reads -36 dB.
The bottom is the woofer, measuring the 100Hz fundamental and its harmonics below crossover....it reads -16dB.
That 20 dB difference equals about a 100 to 1 ratio, the power going to the woofer vs the power going to the tweeter.

So a 100W amp, sending a 100Hz square wave through said crossover, would pass only about 1W to the tweeter.

If crossover is dropped to 1000Hz, there is still a 15dB difference between woofer and tweeter using the 100Hz square wave . A 32 to 1 ratio.
But the lower a tweeter is designed to run, the greater its power handling capability is.

I don't think any amp can clip worse that a square wave, can it?
If not, seems to me the 'clipping can take out tweeters idea' can only happen in carefully crafted situations.

 

[theREALdotnet]

What is he even talking about? Explain?

I think he's alluding to the association between aluminium exposure and lower cognitive function.

 

[mcdn]

The FTC rule is obviously poorly written as the JEITA submission to the FTC points out. Realistically the market wants to know two things: the continuous or thermally limited output of an amp, and also the short term power capability of the amp, and in both cases with music signal because that's what they'll be used for.

When I reviewed the Fosi M03 mono sub amp, the thermally limited continuous power was something like 30W. It can absolutely deliver 230W for short bursts, and with music programme that's all it will ever need to do - short bursts of high power and relatively low continuous power. Just one look at the case and internals is enough to see that 30W is maybe generous, but not implausible.

Technically I suppose for a given amp one could create a graph showing how long (from a known state) the amp takes to be thermally limited. But I strongly suspect it's a straight line between "thermal stability" and "250ms" for most amps, so we only need those two numbers.

 

[amirm]

A measly 5 minutes at full rated power, at any frequency across the audible bandwidth is not unreasonable. In fact it's not much of a hurdle at all for a well designed amplifier.

Maybe it is. Maybe it isn't. We need the data that neither you nor FTC have on what percentage of amplifiers can pass such a test. My personal experience of blowing up a Sony AVR and one by a key designer on this forum shows that it is much harder than you state.

Now, if we had a reason to pass such a test, sure, we would sit back and have companies deliver on this. But we can't with straight face.

 

[Old_School_Brad]

Technically I suppose for a given amp one could create a graph showing how long (from a known state) the amp takes to be thermally limited. But I strongly suspect it's a straight line between "thermal stability" and "250ms" for most amps, so we only need those two numbers.

Are you suggesting a 250 ms test? -That most amps will reach thermal stability within 250 ms? What do you mean?

 

[Old_School_Brad]

Maybe it is. Maybe it isn't. We need the data that neither you nor FTC have on what percentage of amplifiers can pass such a test. My personal experience of blowing up a Sony AVR and one by a key designer on this forum shows that it is much harder than you state.

Why do you believe it's important to gather data on how many amplifiers pass this test?
Suppose the results show that only 20 out of 100 amplifiers pass. What insights does that provide?

• From the manufacturer's perspective: "The test is deeply flawed, as evidenced by the fact that so few amplifiers can pass it."
• From a different viewpoint: "This was entirely achievable decades ago. Today's manufacturers seem to have lowered standards, chasing unrealistic power ratings."

In my view, it's not right to establish criteria and thresholds based solely on how current market products perform. If we do this, we're not driving any improvement; instead, we're essentially endorsing inflated ratings. It's like sweeping everything under the rug and declaring, "This is the baseline."

And it would be a weak baseline, dominated by false claims.

 

[mcdn]

Are you suggesting a 250 ms test? -That most amps will reach thermal stability within 250 ms? What do you mean?

I’m suggesting that “peak power” should be measured with 250ms of pink noise. Continuous power should also use pink noise but obviously would need to be tested over a longer period, at least 5 minutes.

 

[NTK]

Why do you believe it's important to gather data on how many amplifiers pass this test?
Suppose the results show that only 20 out of 100 amplifiers pass. What insights does that provide?

• From the manufacturer's perspective: "The test is deeply flawed, as evidenced by the fact that so few amplifiers can pass it."
• From a different viewpoint: "This was entirely achievable decades ago. Today's manufacturers seem to have lowered standards, chasing unrealistic power ratings."

In my view, it's not right to establish criteria and thresholds based solely on how current market products perform. If we do this, we're not driving any improvement; instead, we're essentially endorsing inflated ratings. It's like sweeping everything under the rug and declaring, "This is the baseline."

And it would be a weak baseline, dominated by false claims.

If 80% of amplifiers fail the FTC test, does that equate to 80% of amplifier purchasers are unhappy with their purchases (or their purchases fail to meet their requirements)? If not, what good is the FTC test?