Amp Headroom vs Efficiency (and quick question about dyno vids)

[geolemon]

Heat is a great point - there's going to be some effect, at least with some amps...
Individual electric component values shift, when warm vs cold. The power supply side is pretty isolated and/or basic enough to be pretty unaffected, but the amp side could shift. You never see those plots.

I still suspect it's below the range of hearing.

People will argue all day over a thousandth of a percent, just as they will argue the absolute superiority of 1mm extra Xmax in a sub (and - being close to my heart, I'm guilty ), without accounting for a myriad of system attributes that need to line up just right in order to even get close to that extra 1mm...
...or having a system where even a jump from 1% to 2% THD is able to be heard.

Richard Clark's amp challenge hasn't gone uncollected over the past couple decades for no reason - despite everyone's proclivity to slippery slope fallacies, false cause - and especially in today's age of class A vs D, the old "no true Scotsman" fallacy reasoning (Google it). He was just smart in making the point that things that can't be heard, can't be heard.

Sent from my LM-G820 using Tapatalk

[Stealth1]

This test seems to indicate the opposite of the chart that was posted earlier in this thread. It is for a 4 channel Pioneer class D car audio amplifier, and if I am interpreting this chart correctly, it shows that THD increases along with output. Is that correct?

If that is the case, it seems it would be difficult to generalize whether such a pattern applies to all class D amplifiers, class AB, etc. (or what kind of variation might be found even with an amplifiers of the same class). Again, back to the Pioneer chart, we see something around a 3-4db increase in THD+N between 20 and 50 watts output, which corresponds to about a 5% increase in THD+N between 20 & 50 watts.

Researching this topic I found the link below which simulates distortion, and I was definitely able to tell the difference on a 5% step. If my reasoning here is correct, this would lead to a noticeable difference in sound quality for the amp that I have posted between 20 and 50 W, and by that logic, it seems that a similarly performing amp with a lot of head room would produce notably better sound quality.

https://www.audiocheck.net/testtones...ll.php?frq=125

Just a small note: I am admittedly somewhat out of my depth on this type of a discussion, so if I have made any serious mistakes in my calculations or understanding of these topics I do apologize.

[Jdunk54nl]

This test seems to indicate the opposite of the chart that was posted earlier in this thread. It is for a 4 channel Pioneer class D car audio amplifier, and if I am interpreting this chart correctly, it shows that THD increases along with output. Is that correct?

If that is the case, it seems it would be difficult to generalize whether such a pattern applies to all class D amplifiers, class AB, etc. (or what kind of variation might be found even with an amplifiers of the same class). Again, back to the Pioneer chart, we see something around a 3-4db increase in THD+N between 20 and 50 watts output, which corresponds to about a 5% increase in THD+N between 20 & 50 watts.

Researching this topic I found the link below which simulates distortion, and I was definitely able to tell the difference on a 5% step. If my reasoning here is correct, this would lead to a noticeable difference in sound quality for the amp that I have posted between 20 and 50 W, and by that logic, it seems that a similarly performing amp with a lot of head room would produce notably better sound quality.

https://www.audiocheck.net/testtones...ll.php?frq=125

Just a small note: I am admittedly somewhat out of my depth on this type of a discussion, so if I have made any serious mistakes in my calculations or understanding of these topics I do apologize.

While that amp is decent for car audio for the price, it is nothing to really write home about as far as amps go.

Albeit, a lot of amps will have that "u" shape. They get better and better until they start slowly creeping up and then finally clip. Just the better designed don't have as much of a creep up and clip close to their lowest distortion and highest power.


Here is one of the best amps measured and it still does it, but its distortion is nowhere near that of the pioneer until well into clipping (Right side y-axis shows the db to compare)




Here is a hypex based amp:



And here is a purifi based amp:

[geolemon]

That's what I was referring to here:

...this is the measurement for this particular amp. Is that necessarily true for ALL class A/B amps? Might there be ones where the power supply falls short, and the distortion accordingly increases at some point before rated power?

And even if it WERE true for all class A/B amps... What about other amp classes? Class D in particular - it clips ugly. With class A/B amps we could get close to clipping and most people wouldn't notice if the needle got a little into the red on extremes.

There's a risk in seeing one chart, reading one thing, hearing one thing, and thinking that dictates some sort of rule.

You have to be able to understand and poke holes in generalizations.

That's exactly where my lengthy post started.

The high level point - EVEN in the case of that unusual amp, it is NOT a reason to not set your gains conservatively low.

For more mid-level and especially full range class D, the benefits are more obvious - as shown.

Sent from my LM-G820 using Tapatalk

[Jdunk54nl]

This also comes into, can anyone actually here this stuff. I know when I have taken distortion listening tests, I usually can't hear distortion that well with test tones until about 5% (-25db). Music is upwards of 10% (-20db) distortion.

Then we take speaker distortion into account, and it is WAY worse than even that pioneer amp (Even the purifi woofer). So is it significant if the amp has that much? Probably not.


From Erin's Audio Corner: For reference, that pioneer amp had THD at about -55db,

[Justin Zazzi]

Again, back to the Pioneer chart, we see something around a 3-4db increase in THD+N between 20 and 50 watts output, which corresponds to about a 5% increase in THD+N between 20 & 50 watts.

Researching this topic I found the link below which simulates distortion, and I was definitely able to tell the difference on a 5% step. If my reasoning here is correct, this would lead to a noticeable difference in sound quality for the amp that I have posted between 20 and 50 W, and by that logic, it seems that a similarly performing amp with a lot of head room would produce notably better sound quality.

https://www.audiocheck.net/testtones...ll.php?frq=125

Just a small note: I am admittedly somewhat out of my depth on this type of a discussion, so if I have made any serious mistakes in my calculations or understanding of these topics I do apologize.

That audiocheck site is wonderful! So if you can hear the presence of 5% distortion, that's great! However, if we compare that to the distortion plot of the amplifiers above then when you see a difference of roughly 5% it might actually be going from 0.100% to 0.105% distortion. So keep an eye on how far down the distortion is vs the signal, and don't concentrate too much on the relative difference in distortion.

It's a little bit like you can tell if the headlights on a car are on or off, but could you tell if the headlights were on or off if the car was on the moon and you were looking from earth? At some point, it doesn't matter if the lights are on or off or purple or flashing because you can't detect them anyways. Distortion is kinda like that.

Also, thank you for going out of your depth!

[Holmz]

This test seems to indicate the opposite of the chart that was posted earlier in this thread. It is for a 4 channel Pioneer class D car audio amplifier, and if I am interpreting this chart correctly, it shows that THD increases along with output. Is that correct?

If that is the case, it seems it would be difficult to generalize whether such a pattern applies to all class D amplifiers, class AB, etc. (or what kind of variation might be found even with an amplifiers of the same class)..
..
.

Led go from 200 to 20w with -10 db?
then 20 to 2 with another -10dB
then 2 to 200mW, then 200mw to 50mW with 6dB

Or think of it as a 4000:1 ratio.

Now imagine 16k levels and draw a sinewave... it looks like a sinewave.
Now reduce thlevels to 4 and draw that sinewave, and it looks like a Tetris convention.

The distortion at zero output should always be high on a classD, but there is also not a lot to hear with the volume off... so who cares?

[Stealth1]

Led go from 200 to 20w with -10 db?

The distortion at zero output should always be high on a classD, but there is also not a lot to hear with the volume off... so who cares?

I'm not sure of the point you are trying to make here.

[Stealth1]

I see what you mean, the amperage is different on the still shot of the test... I don’t know... ask him... he normally answers questions... infact I will ask...

Did you ever receive a reply on this?

[Holmz]

I'm not sure of the point you are trying to make here.

Well if the distortion is highest at low volumes, does it matter?
(The is no actual sound coming out at zero volume anyhow)