Purifi 1ET400A with linear PSU tested to comply with FTC rule

[JSmith]

practical sense

It's the regulation that is at fault in that regard... there should be no necessity for interpretation, as then there can't be misinterpretation. It's a very poor "regulation" in that regard and in general, what should be a well formed and structured regulation, is one of the worst I've seen.


JSmith

 

[BR52]

Thank you, @pma, you did the work to give us a real live example. It's really enlightening, and we have some facts in the discussion.

 

[MaxwellsEq]

Yes but they will argue with 5 minutes at the fixed frequency. The debate is pointless, as most debates.

It's not the debate that's pointless. This is a deeply flawed wording which requires both manufacturers and governors to try and guess what they mean. @KSTR has proposed some pragmatic possibility in post #38.

So far, I have seen no evidence including fom you, @pma of this test being completed that complies with the wording:

at all frequencies

 

[MaxwellsEq]

It's not the debate that's pointless. This is a deeply flawed wording which requires both manufacturers and governors to try and guess what they mean. @KSTR has proposed some pragmatic possibility in post #38.

So far, I have seen no evidence including fom you, @pma of this test being completed that complies with the wording:

So rather than being negative, I will be positive and propose some sensible frequencies:

• 20Hz (EDM, pipe organ and, well we always test here!)
• 40Hz (the peak of a lot of music - anywhere 35 to 60Hz would probably do)
• 1kHz (we always test here)
• 3.5kHz (point of hearing highest sensitivity)
• 6kHz (will include 3rd audible harmonic at 18kHz)
• 9kHz (will include 2nd audible harmonic at a18kHz)
• 16kHz (sensible top end test for most people's hearing - very little real music above here)
• 20kHz (probably pointless - e.g. 2nd harmonic is at 40kHz, but we always test here, so...)

That's 8 tests, at five minutes with, say, a 30 minute cool down, it's more than 4 hours testing. For real audio, we could drop 20, 1kHz and 20kHz, since they tell us very little about what most people will experience. We could also probably drop 3.5kHz. In which case, a pragmatic test would be

• 40Hz
• 6kHz
• 9kHz
• 16kHz

What do we think?

 

[Vladimir Filevski]

I don't look at the regulation from an engineering point of view but a legal one. Viewing from that perspective then the regulation means what it says all frequencies from 20hz-20khz at 5 minutes.

From legal point you are talking nonsense, just as the FTC requirements for testing literally all frequencies is nonsense. It is impossible to test at infinite number of frequencies/something - it is not engineering point of view, but a logical one. Legal point of view must be logical, do you agree?
Testing at four frequencies: 20 Hz, 1 kHz, 6.6 kHz and 20 kHz is more than enough to comply with FTC, because all eventual problems with amplifiers are happening at 20 Hz or 20 kHz (including 6.6 kHz for Class D).

 

[Sokel]

(including 6.6 kHz for Class D).

One of my buying points for icepower is the fact they always include it at their datasheets as worst case audible scenario.
It admittedly looks bad on older ones (at the Conductor chip is as good as the 1kHz one,that's huge improvement) but you know what you get,that's a great advantage.

 

[JustJones]

From legal point you are talking nonsense, just as the FTC requirements for testing literally all frequencies is nonsense. It is impossible to test at infinite number of frequencies/something - it is not engineering point of view, but a logical one. Legal point of view must be logical, do you agree?
Testing at four frequencies: 20 Hz, 1 kHz, 6.6 kHz and 20 kHz is more than enough to comply with FTC, because all eventual problems with amplifiers are happening at 20 Hz or 20 kHz (including 6.6 kHz for Class D).

From a legal perspective you need to read the regulation as is, it isn't open to my or your "interpretation " as some here want to make it. Testing 4 frequencies may be more than enough to comply to your interpretation but it no way meets the requirements. Whether the regulation of 20hz-20khz makes sense from an engineering perspective will only come into play if the regulation is challenged or the FTC makes a clarification.
Of course there's my crystal ball interpretation which is the FTC isn't going to put much effort, if any, into enforcement of it.

 

[Geert]

When max power @1% THD would be measured at 20hz and 20kHz, is there any reason to believe an amps performance would be substantially lower at any frequency inbetween? I checked 9 amplifiers measured by Stereophile, they measure THD vs frequency at somewhere between 20W/8 ohm and 80W/8ohm (so not full power unfortunately), and for all of them distortion is max. at 20kHz.

@pma did the same measurement at much higher power for different amps, for example the Purifi 1ET400A. We can se how the distortion profile changes at max. power, but we can still see max. THD @20 Hz.

PMA's measurement of the Hypex NC252MP @ 250W, THD is a bit higher at 70 Hz but doens't make much of a difference:

Second question; any reason to believe max. power would be substantially lower when playing multiple (all) frequencies at the same time versus doing seperate measurements at 20 Hz and 20 kHz?

In summary, how difficult does the FTC need to make this rule? Can they not just ask to determine max power at any frequency between 20 Hz and 20kHz where THD reaches 1% max (and specify the duration)?

 

[Vladimir Filevski]

From a legal perspective you need to read the regulation as is, it isn't open to my or your "interpretation " as some here want to make it. Testing 4 frequencies may be more than enough to comply to your interpretation but it no way meets the requirements. Whether the regulation of 20hz-20khz makes sense from an engineering perspective will only come into play if the regulation is challenged or the FTC makes a clarification.

Again, you are talking legal nonsense! It is impossible to test infinite number of something - per your literal interpretation. That is not "engineering perspective" but legal and logical one.
You will lose big time in the court if you try to accuse some amplifier manufacturer for testing only at several frequencies (including 20 Hz and 20 kHz) and not at infinite number of frequencies (while marketing their amplifier as FTC compliant). Wanna bet?

 

[HaveMeterWillTravel]

It's not the debate that's pointless. This is a deeply flawed wording which requires both manufacturers and governors to try and guess what they mean. @KSTR has proposed some pragmatic possibility in post #38.

So far, I have seen no evidence including fom you, @pma of this test being completed that complies with the wording:

Well, let's reexamine some possible assumptions of yours:
Are you thinking of only integer values for the frequencies? After all, there are an infinite number of rational numbers between 1,000 and 1,001.

OTOH: the regulation does not specify how to configure the acquisition hardware, sampling and analysis parameters. What is the minimum probability of intercept which would be acceptable? What if one were to set their sweep time to be on the order of μsec?

Most other standards specify the test conditions.

HMWT

 

[JustJones]

Again, you are talking legal nonsense! It is impossible to test infinite number of something - per your literal interpretation. That is not "engineering perspective" but legal and logical one.
You will lose big time in the court if you try to accuse some amplifier manufacturer for testing only at several frequencies (including 20 Hz and 20 kHz) and not at infinite number of frequencies (while marketing their amplifier as FTC compliant). Wanna bet?

Are you in the United States? I wouldn't lose in court since I wouldn't be involved. The regulation doesn't require infinite number of frequencies only those between 20hz-20khz at rated power for 5 minutes.

 

[JustJones]

The problem with this regulation or at least 1 problem is they didn't specify how the 5 minute test was to be done. It's ambiguous and in my experience ambiguous regulations don't fare well when challenged. I mean we have about a half dozen different interpretations on this site alone as to what meets the requirements.

 

[Vladimir Filevski]

Are you in the United States? I wouldn't lose in court since I wouldn't be involved. The regulation doesn't require infinite number of frequencies only those between 20hz-20khz at rated power for 5 minutes.

Are you in the United States? As far as I know, kids in the elementary schools in US know that there is an infinite number of frequencies between 20 Hz and 20 kHz.

 

[JustJones]

Are you in the United States? As far as I know, kids in the elementary schools in US know that there is an infinite number of frequencies between 20 Hz and 20 kHz.

Is the any type of signal that sweeps all those frequencies?

 

[JustJones]

far as I know, kids in the elementary schools in US know that there is an infinite number of frequencies between 20 Hz and 20 kHz

Yes I assume so. Do you think these elementary schools teach how to parse government regulations?

 

[Vladimir Filevski]

Is the any type of signal that sweeps all those frequencies?

No. Steady state signal at fixed frequency with duration of 5 minutes is not the same as sweep signal (20Hz-20kHz) with 5 minutes duration.

 

[JustJones]

This discussion is pathetic.

Test at the frequency extremes, 20Hz and 20kHz. Nobody cares about 1kHz, it was always a cop out frequency. The fact Amir continues to rank amplifiers on a 1kHz SINAD number only makes him look like a fool.

Can your amplifier deliver its rated power across the audible bandwidth with a distortion less than or equal to its headline number?

So that's your interpretation of the regulation. 2 tests at the extremes?

 

[tmtomh]

This discussion is pathetic.

Test at the frequency extremes, 20Hz and 20kHz. Nobody cares about 1kHz, it was always a cop out frequency. The fact Amir continues to rank amplifiers on a 1kHz SINAD number only makes him look like a fool.

Any frequency across the rated bandwidth. Measure the harmonics. Extend your measurement bandwidth.

Can your amplifier deliver its rated power across the audible bandwidth with a distortion less than or equal to its headline number?

Your constantly changing arguments (this latest one is not what FTC prescribes) and ad hominem attacks are what’s pathetic.

 

[JustJones]

No. Steady state signal at fixed frequency with duration of 5 minutes is not the same as sweep signal (20Hz-20kHz) with 5 minutes duration.

Which does the FTC want? Either of these or something else? For some reason people keep missing the point. The regulation is to VAGUE.

 

[Vladimir Filevski]

Yes I assume so. Do you think these elementary schools teach how to parse government regulations?

I think that kids do know there is an infinite number of frequencies between 20Hz and 20 kHz. Parsing government regulations start with logic... you are not there yet.