Clipping & Underpowered Speakers

[killdozzer]

I'm interested in something quiet specific; is there such a thing as speakers NOT getting enough power to perform as they should*, but amp NOT going into clipping?

(*As they should according to their own specs, not some unrealistic expectation.)


Is this possible? Or is it that if speakers lack power the one and only scenario is that the amp will start to clip?

Even more specifically; is it possible that the result of lack of power sounds like attenuation in certain freq without amp going into clipping?

 

[DVDdoug]

Even more specifically; is it possible that the result of lack of power sounds like attenuation in certain freq without amp going into clipping?

No. Low power/volume doesn't hurt frequency response or distortion. Of course it's possible to over-drive a speaker into distortion with a high-power amplifier that's not clipping.

But your ears are not linear... When you turn-down the volume it sounds like you've turned-down the bass even-more (Equal Loudness Curves). And at low levels the "little details" and some of the high frequencies simply become too-quiet to hear.

...In the 1970s most receivers had a "loudness" switch (actually a loudness compensation switch) that worked along with the volume control to boost the bass as you turned-down the volume. For some reason that's mostly gone out-of-style... Maybe because the receiver only knows the volume-control setting and not the actual acoustic loudness. Some modern AVRs do have it, along with the ability to calibrate to the acoustic loudness.

 

[Tom C]

I’m a little unclear on the question. The difference between clipping and lumpy frequency response should be readily apparent. There’s a local radio station that has a recording of their station identification that’s clipped, somewhat like the announcer was shouting into the mike. The quality is obvious, even with the volume set low.
Maybe a specific example would help. Or maybe I’m the only one who’s not getting it.

 

[Ron Texas]

This is a vague question. Missing is speaker sensitivity, room size, desired acoustic output and probably a few I missed.

 

[killdozzer]

No. Low power/volume doesn't hurt frequency response or distortion.

So, whenever a speaker needs power that the amp can't deliver, the amp will clip? That's the only possible consequence?

If you allow me to propose an imaginary situation and let's say you have a test CD with all those test tones; you play one single freq that hits the impedance region of the speaker around 8 ohms and you play it very loud, let's say 96dB SPL, but the amp can handle it easily. After that you deliberatly pick a test tone that hits the impedance dip leaving the volume at the same level and if your amp can't give you that freq at that level, it will always clip?

 

[killdozzer]

@Tom C & @Ron Texas thanks for joining in. I was leaving them as in "whatever the speaker needs, whatever its specs might be".

OK, I see it won't do, but maybe my previous post helps this out.

Still, is there at all such a situation where underpowered speakers give any other symptoms other than amp going into clipping and you hear the clipping distortion?

 

[killdozzer]

lumpy frequency response should be readily apparent

Is the amp ever a suspected perp for the lumpy frequency? Maybe it's what you wrote. How is it readily apparent?

 

[DVDdoug]

So, whenever a speaker needs power that the amp can't deliver, the amp will clip? That's the only possible consequence?

An amplifier clips when you turn it up too loud. i.e. If you try to get 110 watts out of a 100W amplifier. That could happen with inefficient* speakers, or with a large room, or if you just like loud music. ...That's probably not a good example because 100W is usually more than enough in a "living room" with any speakers. But if you like to rattle the walls with bass you might need more power and bigger woofers/subwoofers.

The speaker "needs" more power if it's not getting loud enough for you. Of course, if you feed them too-much power you can burn them out. If they are not going loud enough for your taste you may need a bigger amplifier and/or "bigger" or more efficient speakers.

Note that double the power is only +3dB and although a 3dB change is 'noticeable" it's probably not worth replacing your amplifier. +10dB is 10 times the power.



* Speakers don't usually have an "efficiency" spec. If they give you a spec it's a sensitivity spec. which gives the dB SPL level at 1W and 1 meter distance.

 

[freemansteve]

Clipping can occur at several point from file data to speakers....
The trick is isolate where it's occurring. It's pretty easy.

 

[DVDdoug]

If you allow me to propose an imaginary situation and let's say you have a test CD with all those test tones

If you do that be careful with high-frequency tones. The tweeter in a "100W" speaker can't handle anything near 100W (and the woofer can't handle 100W continuously either).

you play one single freq that hits the impedance region of the speaker around 8 ohms and you play it very loud, let's say 96dB SPL, but the amp can handle it easily

Most amplifiers are OK with "normal" speaker impedance variations. Amplifiers (at least solid state amplifiers) have low source-impedance which means the output-voltage is independent of load impedance.

Speakers are specified & tested with "constant voltage" (voltage independent of load). That means the amplifier isn't putting-out as much wattage where impedance is higher (and it's putting-out more where the impedance is low) but a speaker with flat frequency response will be flat when driven from a flat amplifier.

The speaker will "suck" more power at lower impedance and sometimes that can be an issue but it's usually not a problem with music (or voice) since the frequency and power vary from moment-to-moment and you're not stressing the amplifier with constant worst-case conditions. Under these conditions an amplifier will usually put-out more power than it's normal rating instead of clipping, especially for relatively short-term peaks.

 

[killdozzer]

@DVDdoug Thank you very much. Some interesting info. I know about those basics, the doubling of the power and ratings of the amp and ratings of the speaker, sensitivity and the rest.

So, what we usually call headroom is desirable only to prevent clipping? Speakers are not using any of it before they start to clip? At quiet volumes there's absolutely no difference between 2x40 and 2x80 wpch into 8?

Speakers don't behave any better with a stronger amp if the volume is low? Either it's sufficient at a certain SPL or it's clipping.

 

[Spkrdctr]

@DVDdoug Thank you very much. Some interesting info. I know about those basics, the doubling of the power and ratings of the amp and ratings of the speaker, sensitivity and the rest.

So, what we usually call headroom is desirable only to prevent clipping? Speakers are not using any of it before they start to clip? At quiet volumes there's absolutely no difference between 2x40 and 2x80 wpch into 8?

Speakers don't behave any better with a stronger amp if the volume is low? Either it's sufficient at a certain SPL or it's clipping.

Yes, that is correct. If the volume is low you receive no benefit from having a more powerful amp. Most (many?) people get this all confused especially audiphools. They think that they need 100watts per channel for their 70db listening. The thought being having more power available allows the speakers to have more air and better bass on and on. It is basically BS. If you are using only 5 watts from your 50 watt per channel amp, getting a 100 watt amp does nothing to improve your system. Great topic and one that is misunderstood by most regular audio customers.

 

[Prana Ferox]

Speakers do not clip and speakers do not 'need' power.

Speakers act as a reactive load (resistance, but varies by frequency.) Your amplifier takes an input signal of a given voltage amplitude and multiplies it by whatever gain it is set to, and then can either maintain that amplified voltage on the outputs or it can't. It can clip if that gain-multiplied voltage exceeds its power supply voltage limits - it 'hits the rails' - and this can show up even without output load, as Amir frequently complains about in AVR testing. Or it can clip if attempting to apply that voltage to the speakers' reactive load exceeds its power supply current delivery capacity.

When the amp clips, to simplify things, the normally sinusoidal output waveform distorts at the peaks. This produces distortion of varying order. It does not produce 'lumpy frequency response'; if an amp's FR is bad it is generally a function of the negative feedback circuit being messed up (or interacting with the load when it shouldn't, which isn't necessarily volume dependent.) That clipped signal also acts more like a DC output which ends up causing more heat generation in the speaker voice coils, which can contribute to 'frying' them.

You can talk about speakers with 'difficult' highly reactive loads but it's not really needed for this discussion.

Speakers can suffer from power compression, i.e. running them near but not necessarily at their limits can also cause voice coil heating leading to the motors become less efficient and the output volume to drop (and in the worse case causing the user to crank the volume to compensate, making it worse.) Also if you're really pushing your drivers you can get into non-linear ranges where it's not quite hitting mechanical limits but the physical resistance of the speaker is making it harder to move further - don't do this, and certainly don't go beyond it.

Basically like most machinery you want to size your amps and speakers such that you're comfortably within limits during all operating periods.

@Spkrdctr is correct above but the opposite side of the audiophoolery is that, while you may not need a huge amp to supply a tiny load to your speakers, it is also not necessarily true that a smaller wattage amp is better at it (i.e. the Nelson Pass First Watt business.)

 

[rwortman]

Yes, that is correct. If the volume is low you receive no benefit from having a more powerful amp. Most (many?) people get this all confused especially audiphools. They think that they need 100watts per channel for there 70db listening. The thought being having more power available allows the speakers to have more air and better bass on and on. It is basically BS. If you are using only 5 watts from your 50 watt per channel amp, getting a 100 watt amp does nothing to improve your system. Great topic and one that is misunderstood by most regular audio customers.

If you are listening to music at an average power of 5 watts, depending on the source material, the crest factor could easily be 15db or more. To cover a 15db crest factor at an average level of 5 watts, you need 158 watts available to avoid clipping.

 

[freemansteve]

Speakers "do not need power"? - you will become rich from free energy and perpetual motion machines!

 

[Spkrdctr]

Speakers do not clip and speakers do not 'need' power.

Speakers act as a reactive load (resistance, but varies by frequency.) Your amplifier takes an input signal of a given voltage amplitude and multiplies it by whatever gain it is set to, and then can either maintain that amplified voltage on the outputs or it can't. It can clip if that gain-multiplied voltage exceeds its power supply voltage limits - it 'hits the rails' - and this can show up even without output load, as Amir frequently complains about in AVR testing. Or it can clip if attempting to apply that voltage to the speakers' reactive load exceeds its power supply current delivery capacity.

When the amp clips, to simplify things, the normally sinusoidal output waveform distorts at the peaks. This produces distortion of varying order. It does not produce 'lumpy frequency response'; if an amp's FR is bad it is generally a function of the negative feedback circuit being messed up (or interacting with the load when it shouldn't, which isn't necessarily volume dependent.) That clipped signal also acts more like a DC output which ends up causing more heat generation in the speaker voice coils, which can contribute to 'frying' them.

You can talk about speakers with 'difficult' highly reactive loads but it's not really needed for this discussion.

Speakers can suffer from power compression, i.e. running them near but not necessarily at their limits can also cause voice coil heating leading to the motors become less efficient and the output volume to drop (and in the worse case causing the user to crank the volume to compensate, making it worse.) Also if you're really pushing your drivers you can get into non-linear ranges where it's not quite hitting mechanical limits but the physical resistance of the speaker is making it harder to move further - don't do this, and certainly don't go beyond it.

Basically like most machinery you want to size your amps and speakers such that you're comfortably within limits during all operating periods.

@Spkrdctr is correct above but the opposite side of the audiophoolery is that, while you may not need a huge amp to supply a tiny load to your speakers, it is also not necessarily true that a smaller wattage amp is better at it (i.e. the Nelson Pass First Watt business.)

Ok, you took the engineering way of responding. Having gobs of power is all for nothing unless your needs require a lot of power. Anyway, good post Prana!

 

[Spkrdctr]

If you are listening to music at an average power of 5 watts, depending on the source material, the crest factor could easily be 15db or more. To cover a 15db crest factor at an average level of 5 watts, you need 158 watts available to avoid clipping.

OK, drop it to 1 watt.

 

[Urvile]

Ok all of this is very interesting. How do you tell that your amp is clipping other than actually having moments with no sound?
I know it happens occasionally with my existing amp. (I plan to upgrade) But how much is degraded signal?
I'm really curious because my speakers are not sensitive.
R700 are

Sensitivity (2.83V/1m)

88dB

And the L400

Sensitivity (2.83V/1m)

87.5dB

Both are 4 ohm impedance.

 

[rwortman]

OK, drop it to 1 watt.

Unless we are using high sensitivity speakers, I don’t think we are spending much time listening at one watt. My living room system was once a 15wpc vintage receiver driving a pair of DIY 6.5” two ways. It would play pretty loud and sounded pretty clean. It was surely clipping on big drum hits and bass thumps but that doesn’t mean the clipping was preferable to having more power. People often use small amps on small speakers when the sensitivity of little bookshelf speakers is lower than big towers. I think 75 to 100 watts is about minimum for medium sensitivity speakers in a medium sized room if music is going to be played rather loud from time to time unless you have speakers with 95+ db sensitivity. I have systems in rooms for background listening that have less.

Clipping on percussion is not that easy to hear. I run a mixing board for a live band once a week. It is a digital board and all the effects are digital. I can’t apply compression until after the ADC’s. I let my kick and snare drum mic’s clip a little at the input so I can get better SNR on the normal hits and you can’t hear a problem. It’s a tradeoff. In my opinion there is no reason a 25wpc amp should sound better than a 250wpc amp so we should buy all the clean power we can afford. Better to have headroom and there is no drawback

 

[Spkrdctr]

Unless we are using high sensitivity speakers, I don’t think we are spending much time listening at one watt. My living room system was once a 15wpc vintage receiver driving a pair of DIY 6.5” two ways. It would play pretty loud and sounded pretty clean. It was surely clipping on big drum hits and bass thumps but that doesn’t mean the clipping was preferable to having more power. People often use small amps on small speakers when the sensitivity of little bookshelf speakers is lower than big towers. I think 75 to 100 watts is about minimum for medium sensitivity speakers in a medium sized room if music is going to be played rather loud from time to time unless you have speakers with 95+ db sensitivity. I have systems in rooms for background listening that have less.

Clipping on percussion is not that easy to hear. I run a mixing board for a live band once a week. It is a digital board and all the effects are digital. I can’t apply compression until after the ADC’s. I let my kick and snare drum mic’s clip a little at the input so I can get better SNR on the normal hits and you can’t hear a problem. It’s a tradeoff. In my opinion there is no reason a 25wpc amp should sound better than a 250wpc amp so we should buy all the clean power we can afford. Better to have headroom and there is no drawback

I agree with you in your specific special case. I'm answering in the regular general audio user situation. A generic answer for generic general use. But in your specific case you are spot on!