Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Power Amps: Headroom vs. THD+N (Read 22144 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Power Amps: Headroom vs. THD+N

Interesting thread. The power amps I have are considerably more powerful than the receivers they being used with, but my way of thinking is that headroom is always a good thing, and it's always better to have an amplifier running at a far lower capacity than a far higher capacity. Ie, running at 30% of max rated power at 80% of max rated power. The larger amp should have lower distortion running it at lower output levels compared to the smaller amp. Headroom is a great thing!

It's like a car. If you buy a Yaris that can drive to 180 kph, but drive mainly at 120-140 kph or buy a BMW that can drive to 220 kph, and also at 120-140 kph, the extra power makes the drive effortless, it's just not stressed out at the same type of speeds. More power can be a great thing.

Power Amps: Headroom vs. THD+N

Reply #1 –
The comparison with car engines is incorrect.

More power in an amplifier does not make the electrons run faster over the cable, whereas more power in a car engine should offer more acceleration and/or better performance going uphill.  In fact, It's more like comparing gas tank sizes.

Also, your concept of giving more headroom to the amplifier is grossly exaggerated. One thing is not running the amplifier near its rated power, because it will probably show non-linearities not present otherwise, and another thing is suggesting that the more headroom, the better.

Power Amps: Headroom vs. THD+N

Reply #2 –
Quote from: [JAZ] on

The comparison with car engines is incorrect.

More power in an amplifier does not make the electrons run faster over the cable, whereas more power in a car engine should offer more acceleration and/or better performance going uphill.  In fact, It's more like comparing gas tank sizes.

Also, your concept of giving more headroom to the amplifier is grossly exaggerated. One thing is not running the amplifier near its rated power, because it will probably show non-linearities not present otherwise, and another thing is suggesting that the more headroom, the better.


Come again? I don't understand what you mean.

My rather poorly worded analogy is that I'd rather drive an amp at 20% and have it cruise than drive at 80% where its labouring and starting to clip. A more powerful amp can drive speakers easier, can it not? So why would I want an amplifier that is just barely enough when a more powerful amp can handle the demands easier and with less stress?

Distortion would be less because there is more headroom available. Surely? If I take an amp rated that can deliver 100 watts, if I use 50 watts, the distortion will be lower than a far more powerful amp, like say a 250 watt amp, also using 50 watts. I'm driving the one amp at a much less stressful point before it reaches its limits which I would think is a good thing.

Power Amps: Headroom vs. THD+N

Reply #3 –
Quote
Come again? I don't understand what you mean.

My rather poorly worded analogy is that I'd rather drive an amp at 20% and have it cruise than drive at 80% where its labouring and starting to clip.
It's not "starting to clip" at 80% of it's rated power.  It starts to clip if you try to go over 100%.

Quote
the distortion will be lower than a far more powerful amp
That's not necessarily true and with any halfway decent amp there will be zero audible distortion below clipping. 

For example, if you modify an amp for more power (say by boosting the power supply voltage and beefing-up the output transistors/MOSFETs) that doesn't reduce the distortion at a given output wattage.

Like a gas tank, nothing bad happens until you run out of gas.  Do you need a 200 gallon gas tank in your car?

Power Amps: Headroom vs. THD+N

Reply #4 –
Quote from: DVDdoug on
Quote
Come again? I don't understand what you mean.

My rather poorly worded analogy is that I'd rather drive an amp at 20% and have it cruise than drive at 80% where its labouring and starting to clip.
It's not "starting to clip" at 80% of it's rated power.  It starts to clip if you try to go over 100%.

Quote
the distortion will be lower than a far more powerful amp
That's not necessarily true and with any halfway decent amp there will be zero audible distortion below clipping. 

For example, if you modify an amp for more power (say by boosting the power supply voltage and beefing-up the output transistors/MOSFETs) that doesn't reduce the distortion at a given output wattage.

Like a gas tank, nothing bad happens until you run out of gas.  Do you need a 200 gallon gas tank in your car?


So you are saying that distortion does not increase linearly with output??? Or what? How can there be zero audible distortion below clipping? 

Power Amps: Headroom vs. THD+N

Reply #5 –
Quote from: Rich B on
Interesting thread. The power amps I have are considerably more powerful than the receivers they being used with, but my way of thinking is that headroom is always a good thing, and it's always better to have an amplifier running at a far lower capacity than a far higher capacity. Ie, running at 30% of max rated power at 80% of max rated power. The larger amp should have lower distortion running it at lower output levels compared to the smaller amp. Headroom is a great thing!


Output power usually trades off against noise floor though, so a 200W amp may not be as good as a 50W unit, and almost certainly won't be at equal price.  You probably don't want to overspec your amp by all that much if you have the option of just buying an amp that performs well at the power level you need.

Quote from: Rich B on
So you are saying that distortion does not increase linearly with output???


Correct.  Distortion is the ratio of nonlinearity to linearity.  It should be nearly constant until the amp clips, at which point it will rapidly increase.

Power Amps: Headroom vs. THD+N

Reply #6 –
Quote from: Rich B on
So you are saying that distortion does not increase linearly with output??? Or what? How can there be zero audible distortion below clipping? 


Because we know how to design electronic amplifiers that have zero audible distortion below clipping.  We've known how to do that for around fifty years and the methods of doing so are well known and cheap.  This is because engineers utilize physical laws intelligently and pay little attention to "woo".
Ed Seedhouse
VA7SDH

Power Amps: Headroom vs. THD+N

Reply #7 –
If an amp is spec'd and measured correctly then it should introduce no audible distortion at its maximum rated power, which is a lot more than you could say for speakers.

Power Amps: Headroom vs. THD+N

Reply #8 –
Quote from: saratoga on
Quote from: Rich B on
Interesting thread. The power amps I have are considerably more powerful than the receivers they being used with, but my way of thinking is that headroom is always a good thing, and it's always better to have an amplifier running at a far lower capacity than a far higher capacity. Ie, running at 30% of max rated power at 80% of max rated power. The larger amp should have lower distortion running it at lower output levels compared to the smaller amp. Headroom is a great thing!


Output power usually trades off against noise floor though, so a 200W amp may not be as good as a 50W unit, and almost certainly won't be at equal price.  You probably don't want to overspec your amp by all that much if you have the option of just buying an amp that performs well at the power level you need.

Quote from: Rich B on
So you are saying that distortion does not increase linearly with output???


Correct.  Distortion is the ratio of nonlinearity to linearity.  It should be nearly constant until the amp clips, at which point it will rapidly increase.


Are there any graphs or measurements that show distortion vs output? Not that I'm doubting you, but I would like to see some evidence for myself.

Power Amps: Headroom vs. THD+N

Reply #9 –
Quote
Are there any graphs or measurements that show distortion vs output? Not that I'm doubting you, but I would like to see some evidence for myself.
Someone posted these graphs recently.



....In that thread, I also mention that I've NEVER heard ANY distortion from ANY amplifier that wasn't either defective or being over driven.  I've heard noise, and on some older tube amps I've heard poor frequency response, but never distortion.

Power Amps: Headroom vs. THD+N

Reply #10 –
Quote from: Rich B on
Interesting thread. The power amps I have are considerably more powerful than the receivers they being used with, but my way of thinking is that headroom is always a good thing,


Headroom only benefits sound quality if it is needed to avoid clipping, and then it isn't really headroom. It is then power you need to avoid clipping.

Quote
and it's always better to have an amplifier running at a far lower capacity than a far higher capacity. Ie, running at 30% of max rated power at 80% of max rated power.


Not at all.

Quote
The larger amp should have lower distortion running it at lower output levels compared to the smaller amp. Headroom is a great thing!


There is no such general rule. A smaller amp can easily have less distortion than a larger amp, it happens all of the time. I can cite dozens of examples where a more powerful amp had more distoriton at the same power levels where the smaller power amp has less distortion.  As long as you have enough power to avoid clipping, you have all the power you need.

Quote
It's like a car. If you buy a Yaris that can drive to 180 kph, but drive mainly at 120-140 kph or buy a BMW that can drive to 220 kph, and also at 120-140 kph, the extra power makes the drive effortless, it's just not stressed out at the same type of speeds. More power can be a great thing.


That works for cars sort of, but it doesn't work for audio.  In my state the highest speed limit is 70 mph  but you can drive past the enforcement boys at anything this side of 80 mph and usually skate free.  If you are driving a car with a top speed of 80 at 80 mph it is usually pretty stressed out by doing 80 mph. But if you have a car that tops out around 100 mph, its probably under no more stress at 80 than a car whose top speed is far faster.

The thing is that power amps aren't mechanical devices like cars.

Power Amps: Headroom vs. THD+N

Reply #11 –
Quote
A smaller amp can easily have less distortion than a larger amp, it happens all of the time. I can cite dozens of examples where a more powerful amp had more distoriton at the same power levels where the smaller power amp has less distortion. As long as you have enough power to avoid clipping, you have all the power you need.


How can a smaller amp have less distortion? 

Power Amps: Headroom vs. THD+N

Reply #12 –
Quote
Someone posted these graphs recently.


Thanks! So it seems like amplifiers behave with the lowest distortion around the middle of its power band, and then I see the distortion increases quite a bit after that!  Is this a behaviour of most amplifiers in general, or solid state?

Power Amps: Headroom vs. THD+N

Reply #13 –
Quote from: Rich B on
Quote
A smaller amp can easily have less distortion than a larger amp, it happens all of the time. I can cite dozens of examples where a more powerful amp had more distoriton at the same power levels where the smaller power amp has less distortion. As long as you have enough power to avoid clipping, you have all the power you need.


How can a smaller amp have less distortion? 



As long as you don't clip out either amp, the amp's distortion is due to a number of internal design parameters that tend to be largely independent of power. The following comparison should make that clear.

This is an indepdent test of a Yamaha RXV 371:



You can see that from 7 watts to 50 watts its distortion is about the same:  0.003 - 0.004 % THD+N which is really very good.  This happens to be in the power range where it is most likely to be used most of the time. It actually has the same low distortion below 7 watts as the rise shown is due to noise.  Above 50 watts it starts becoming nonlinear, and above 100 watts its distortion starts rising shortly.

This is a test by the same ragazine of a high end AVR Anthem MRX 710 that costs about 10 times as much;



You can see that from 7 watts to 90 watts its distortion remains fairly constant but is a very much higher 0.006 to 0.008 % which is actually just fine, but is about twice as much.  It is clearly a more powerful amp, where clipping starts around 90 watts and it puts out about 140 watts above 100 watts its distortion flattens out again up to 200 watts. It may put out even more power, but it looks like the testers ran out of guts to push it harder. Point being, its at least twice as powerful as the lower distortion AVR. Roll of the dice! ;-)

Now, I obviously cherry picked the tests to make the point, but my point is clear and true: A more powerful amp need not have lower distortion.

I should add that according to much audiophile wisdom, the lower powered amp should sound like crap and be unreliable because it lacks the much-vaunted discrete output stage feature. Its 5 channel power amp section is based on  2 chips: One with 3 copies of the same power amp stage, and one with 2 copies of the same power output stage. I have this AVR and it sounds fine, and has been a reliable servant.

Power Amps: Headroom vs. THD+N

Reply #14 –
Thanks for the reply! I assume the spike at the far left is noise?

So looking at these few graphs it shows the power remain more or less the same, but where the distortion becomes nonlinear on the graph, is that where clipping is taking place? Or is that just a function of the amp reaching its rated power?

Looking at your Yamaha, it shows 50 watts to be very, very clean and then the distortion just increases and increases. I assume past 50 watts it isn't clipping? I don't know, but the behaviour seems fascinating to me.

Power Amps: Headroom vs. THD+N

Reply #15 –
As you can see, the curves divide neatly into three sections.

At the left the curve is dominated by noise, which is generally a function of choosing high-quality components.

At the right the curve is dominated by limits in the supply voltage and the on resistance of the drive transistors.

The center of the curve has to do with good circuit design, and the proper use of the dreaded negative feedback.

Audiophools have been sold on the idea that negative feedback is bad, but by supposedly avoiding it the circuit design job is made much more difficult, and I would not be surprised if more than a few "high-end" amplifiers have less than perfect performance as a result.

Power Amps: Headroom vs. THD+N

Reply #16 –
Quote from: Rich B on
Thanks for the reply! I assume the spike at the far left is noise?


Distortion is really nonlinear/linear.  At the far left the signal (the linear part) is basically zero, so the ratio is higher due to the numerator being very small.  The right is the opposite, linear is fixed but nonlinear goes up (due to clipping). 

As you can see in those plots (of which there are many more online if you'd like to know more), if you run a big amp at too low a power, it may not perform as well.  Ideal is to pick an amp the correct size for your application. 

Power Amps: Headroom vs. THD+N

Reply #17 –
Quote
As you can see in those plots (of which there are many more online if you'd like to know more), if you run a big amp at too low a power, it may not perform as well. Ideal is to pick an amp the correct size for your application.


Why is that? Or rather, what causes the bigger amp to perform worse potentially at a lower power output? I always used to think that big amps had advantages over smaller ones. But now I'm not so sure.

Power Amps: Headroom vs. THD+N

Reply #18 –
Quote
Why is that? Or rather, what causes the bigger amp to perform worse potentially at a lower power output?
That's just the nature of electronics...    Designing the amp for more power doesn't automatically give you less distortion. 

Why does amp-A have less distortion than amp-B?  ...It's a different design.

Let's consider a marketing scenario... 

One customer is looking for a live performance amp.  He's going to want high-power, a mechanically rugged design that can be banged around and an electrically rugged design that won't burn-up if he does something dumb like short-circuit the outputs.  He might want light weight for portability.  He probably doesn't care much about looks.    Cost (and cost-per-watt) are probably very important, especially if it's for a big PA set-up with multiple amps.  A built-in electronic crossover for bi-ampling or tri-amping might be desirable.    He really doesn't care about distortion (as long as it's reasonable and inaudible) and he doesn't care about a little noise.  Some acoustic noise from a cooling fan is no probem either. 

Another customer wants an amplifier for his pro recording studio.    He probably doesn't need lots of power, maybe 100W or less.    His main concern is the best specs possible.*  He wants super-low distortion, very-low noise (including no fan noise and no machanical buzzing).    Maybe he wants flat frequency response from DC to 100kHz (not hard to do).    He wants it to look good and to look expensive.    Cost is not a big concern when you are paying over $1000 for a microphone...    A higher price might actually make the amp more desirable and it might be more impressive to his clients. 




* The pro audio community isn't quite as nuts as the audiophile community, but you'll still find audio professionals that think an amp with 0.001% distortion sounds better than one one with 0.1% distortion, or that one that goes to 100kHz sounds better than one that "only" goes to 20kHz.

Power Amps: Headroom vs. THD+N

Reply #19 –
Okay, so distortion is not perfectly linear with output, which is something that many people are led to believe without any knowledge about such things.

What I find fascinating is the idea that a larger amp with bigger power supplies  may produce higher distortion at a given wattage than a smaller amp at the same wattage, but then I'm going from an audiophile background, which is to say that I don't know any better.

Now what I would like to know is the following, since I own a power amp in my system. Adding a power amp to my system changed the sound of the speakers. The system sounded a bit more alive, more powerful in the bass. I don't believe I'm making it up or that its a figment, hence my question :

If adding a power amp can make non-trivial differences to the sound quality of a system, that are unrelated to placebo. Let's say I add a power amp to a system. Would adding a power amp to a system change the output levels that you hear? Like adding a power amp to a receiver. Would the change alone result in a change in volume?

If that is the case, then perhaps what I hear is a function of a louder volume. But that would piss me off, because I spent good money on my power amp.  But I want to know if what I'm describing is a possible outcome. How would the power amp sound louder if I keep everything else the same (same master volume setting, same level trims etc). Is it a input voltage thing? Have no clue!

Power Amps: Headroom vs. THD+N

Reply #20 –
A power amplifier is specified to output its full power at a specific input signal level. If two amplifiers output their respective full powers at the same input signal level then the one with the higher power rating will amplify more than the one with the lower power rating.

Power Amps: Headroom vs. THD+N

Reply #21 –
Quote from: pdq on
A power amplifier is specified to output its full power at a specific input signal level. If two amplifiers output their respective full powers at the same input signal level then the one with the higher power rating will amplify more than the one with the lower power rating.


So if I have a receiver and I have two power amps, one that is rated to deliver 150 watts and another that is rated to deliver 250 watts, if I just hook up the different power amps keeping all else equal, there may be a difference in volume?

I just received an infraction via PM based on what I experienced with my power amp, which I think is unnecessary because I am open to the possibility that the difference I heard was not intrinsically connected to the power amp sound and more related to how it was set up. I don't know, hence I'm asking questions about it. I'm open to be wrong, so I don't think a PM was necessary on those grounds.

I'm trying to exhaust all physical explanations for what I heard rather than assuming that I heard audible figments of my imagination. I'm trying to be reasonable.



Power Amps: Headroom vs. THD+N

Reply #22 –
Quote
So if I have a receiver and I have two power amps, one that is rated to deliver 150 watts and another that is rated to deliver 250 watts, if I just hook up the different power amps keeping all else equal, there may be a difference in volume?
That's correct.  There's no real standard for power amplifier amplifier gain.

These specs for the Crown CDi series of amplifiers show that all of them are driven to full power at 1.4V.  That means the higher-power versions will play louder with the same input signal.
 
In general, the amp should be able to reach full-power with a line-level signal.  But, there's no rule that says the amp can't go to full power with a slightly weaker input signal.    And, "line level" is not that well defined, especially for consumer equipment (the line-line level RCA jacks from your DVD player may put-out a higher or lower voltage than the RCA jacks on your CD player) and if the signal is volume controlled, it depends on the volume setting.



Power Amps: Headroom vs. THD+N

Reply #23 –
"There's no real standard for power amplifier amplifier gain."

If I were talking about one device I would call that the "gain factor" or perhaps "gain factor ratio", expressed in dB. If I were talking about a series of devices in a chain I would call it the "gain structure" although that also sort of means how one sets gain of each device along the way, to achieve an overall goal, such as maximum SNR. Here I'll use both in sentences:

"The unusually high > 33dB gain factor of certain Emotiva amps is so high that many people complained they were hearing hiss from their receiver's preamp outs. Emotiva reacted to that by reducing the gain factor on later incarnations."

"By setting the output of each device to the highest level possible which won't overload the next device in the chain, you can set your overall gain structure of your complete system to have the best possible SNR."

I'm not a math guy, but here: Gain Factor = ANTILOG10 ( Gain in dB / 20 ) [i.e. ANTILOG10 ( 12 / 20 ) = 4X]
Source: NwAvGuy

Power Amps: Headroom vs. THD+N

Reply #24 –
Quote from: Rich B on
I'm trying to exhaust all physical explanations for what I heard rather than assuming that I heard audible figments of my imagination. I'm trying to be reasonable.


Well I think that you are actually not being reasonable, given what we know about the human organism including the brain.  You are rather being quite unreasonable in rejecting rather robust and uncontroversial scientific findings.  To be reasonable you would assume first that these perceived differences have no physical cause outside of your own organism.  That's what the science says is most likely the case.

Ed Seedhouse
VA7SDH