• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

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

Status
Not open for further replies.
It is routine for class d amp to have unpredictable distortion vs frequency vs power.
In that case, additional measurement points might be required.

This regulation is so poorly written that you need to try to re-interpret words, like "obtainable" to even begin to agree as to what the regulation actually requires.
How am I reinterpreting the word?
There’s no ambiguity about its meaning in this context. I believe the way I explained it in the comment you quoted was clear.

The manufacturer can choose to measure at 1 kHz if they wish. If they believe this single measurement is sufficient, they can disclose that rating for 20 Hz - 20 kHz if they wish, and claim compliance with FTC rules. Whether this meets the standard or not is ultimately their responsibility. If they are confident in their product without conducting further tests, and third-party testing or random testing by the FTC later confirms that the performance is obtainable, then all is well.
However, if third-party testing shows it is not obtainable, then they’ll likely face consequences such as a warning, fine, or whatever action the FTC deems appropriate for the situation.
 
Government provides no specification for how that distortion is measured. I can pick very high bandwidth and invalidate most such ratings by companies.
It's worse than that, they mention THD+N without specifying the noise bandwidth or the distortion bandwidth.
 
If someone would like to rescue this thread from the realm of endless semantics please do so but otherwise I'm minded to close this thread at the weekend. We don't need another neverending debate thread.

Thanks.
 
The manufacturer can choose to measure at 1 kHz if they wish. If they believe this single measurement is sufficient, they can disclose that rating for 20 Hz - 20 kHz if they wish, and claim compliance with FTC rules. Whether this meets the standard or not is ultimately their responsibility.

I assume by "their" you mean FTC? If so, FTC can bring a lawsuit against any business found not to be in compliance. Since compliance is so poorly specified, one can never be sure that a specific product isn't going to result in a legal action. In my line of business, I have to comply with all government regulations or risk heavy fines and possible shutdown. Leaving interpretation of the rules to the government is not an option.
 
The manufacturer can choose to measure at 1 kHz if they wish. If they believe this single measurement is sufficient, they can disclose that rating for 20 Hz - 20 kHz if they wish

That's your interpretation, but what I also read in the rule before it gets into details is "The rated power output is measured in compliance with the standard test conditions" ... "Is measured" is not "believed to be" based on a more limited measurement than what's prescribed, and it's not "full validation of compliance can wait for 3th party testing". But who knows ...

If someone would like to rescue this thread from the realm of endless semantics please do so but otherwise I'm minded to close this thread at the weekend. We don't need another neverending debate thread.

The more than 90 pages of discussion in the 2 threads covering the FTC rule and the Audioholics petition are sufficient proof the rule isn't clear enough for everyone (where 'everyone' includes experienced engineers). Unless an FTC member jumps in and clarifies the spirit of the law and gives the correct interpretation, it's nothing but opinions.
 
Then we're back to the same question: how does a manufacturer comply with the rule to avoid legal action if the rule is not properly written and is open to interpretation?
It does not. The misinterpretation is made by laymen audience.

One needs to know what to test for.

Below is the SMPS limited power at LF.

smpslimited_thdfreq_53W.png
 
Then we're back to the same question: how does a manufacturer comply with the rule to avoid legal action if the rule is not properly written and is open to interpretation?
Compare and contrast a rule written by lawyers and other “laymen” with a proper standard written by technical people (e.g., ASTM standards).
 
How does one become compliant with "shall be obtainable at all frequencies" if not by actually obtaining, while testing? ;)


View attachment 419711
But it doesn't say anything about which bandwidth should be used during the test/measurement. It should, especially for class D amplifiers, have a rather decisive effect on the result, right? Regarding distortion in the higher frequencies that is.

Please note I am not a technical expert so these are questions from my side. I have read that 45 kHz bandwidth is enough, others say 80 kHz. I don't really know what to believe. :oops:
 
Gene's petition article was much more coherent that the rule itself.

If a rule needs to be followed there can be no room for assumptions about what it means... otherwise it's not a rule, but a jumbled mess of suggestions.

The fact it is like that makes it clear there is no intention by the FTC to enforce any of it... or even attempt to.


JSmith
 
It does not. The misinterpretation is made by laymen audience.

One needs to know what to test for.

Below is the SMPS limited power at LF.

View attachment 419889

Regulations don't work like that. You can't have a regulation that is open for interpretation by each manufacturer and, separately, by the government. Tell me, what did FTC mean, and how do you know from the rule, as written?

1. All frequencies between 20-20KHz must be tested, each for 5 minutes to not exceed 1% THD+N
(how does one even do that? - this is literally the wording of the rule)
2. The test must be run at 20Hz and 20KHz
3. The test should be at 20Hz, 1kHz, 20KHz
4. The test should be at 20Hz, 100Hz, 1KHz, 3.5KHz, 10KHz, 15KHz, 20KHz
5. The test should be at 32 points, linearly spaced between 20 and 20KHz
6. The test should be at 64 points logarithmically spaced between 20 and 20KHz
....???

And then, what bandwidth of measurement should be used for the 1% THD+N calculation? 20KHz? 48KHz? 1MHz? 1GHz? Where are the details needed to reproduce a reported non-compliance test should FTC come after my published specifications? As a non-layman, you already know that the exact conditions of the test need to be specified if one wants to reproduce and compare the results.

The rule is poorly written and opens up manufacturers to the risk of legal action, regardless of how they interpret it.

Compare this to any of the ITU-R recommendations, for example. Extremely detailed specifications, with all the parameters defined and spelled out, all the steps needed to reproduce listed, with detailed examples and ways to verify, etc.
 

Attachments

  • 1736430753576.png
    1736430753576.png
    47.5 KB · Views: 15
So that's your interpretation of the regulation. 2 tests at the extremes?

My 'interpretation'? I don't see anything needing much 'interpretation'. The Commission has clearly stated they regard a single 1kHz power output rating as deceptive and amplifiers are required to output their full rated power across the accepted audible bandwidth if they are to claim ANY power output rating above 2W. Whether some little website owner agrees or disagrees is irrelevant.

Specifically, 250mW to rated maximum power shall be obtainable for a period not less than 5 minutes, at any frequency within the specified bandwidth (default 20Hz-20kHz) with distortion (+N) not exceeding either 1% or the manufacturer's rated distortion and noise figure below that. A manufacturer can specify well below 1% for their FTC compliant rating if they so wish, but not above it.

Let's not forget the 250mW THD+N low power limit requirement here. It's always been there, but gets conveniently ignored by many. If you've got an amplifier with just 60uV of noise (20-20k untwd), that's already 0.0042% THD+N before you even consider the harmonic distortion component.

Like this vintage 80s amp when the manufacturers were being honest:
1736403839148.png

~0.002% is the contribution from residual noise at the low end (250mW) in that above example.

My point is 1kHz is not representative of the continuous power capabilities of amplifiers. Most roll off at each end. Power drops and distortion rises. Hence the need to test at the extremes to comply. Otherwise there's just more of the same cheating.

Through the 70s and 80s we had the half power bandwidth test. That was the frequency bandwidth at which an amplifier could achieve its rated power less 3dB (half power) under 0.1% THD. Plenty of amplifiers tested could not achieve even that. The UK based HiFi Choice tested both power bandwidth and also used an excellent overload recovery test which is still valid today.

So in answer to your question- no, not 2 tests at the extremes only.

One could create a 300 second (5 minute) linear sine sweep from 20Hz-20kHz and monitor the THD in real time, just like I have done. It's not hard, as the hard wall is just 1% THD. Maybe Amir's AP can't lock onto the fundamental and report a continuously moving THD up the spectrum in real time, I don't know, but my inexpensive QA403 can.
 
We locked this thread because it has run its course as it applies to the Subject Product.

If you all wish to continue to debate the Government Standards in a generic manner as it applies to all Amplifiers. Please feel free to open a follow on new thread. If you do let us know and we will drop a link to the new thread here.

Thank you for your understanding and support. ;)
 
Status
Not open for further replies.
Back
Top Bottom