Question about wattage and speakers

[Old Garage-Bander]

Underpowering is when you have to push the amp so hard to get the required spls that your output signal is a heavily clipped waveform at all times even though that power level is less than the speakers power handling capability.

And if I may, and assuming you and I are on the same page, underpowering refers to the amp not having enough power to do the job. It is not about underpowering the speakers, which the OP asked about earlier. One cannot damage a speaker by underpowering it. If that were the case, every speaker would fail everytime we turned off our amps. Sending no power to it is the the most extream version of under-powering it.

What one would do ideally is to find that balance between amplifier power and speaker SPL that will do the job.

Let's say for example, I had a 200 watt amp and a 15" speaker. I find that it's a good combination for the practice room, but not for gigging. I have two choices. I can add more power, or I can add more speaker.

What happens if I double the amplifier power to 400 watts and keep using the 15" speaker (assuming of course it will handle the power)? Does my rig get twice as loud? No. How much louder does it get? Just a barely perceptible increase. Why? Because human hearing does not relate to power in a linear manner.

Well what if I add another 15" speaker to my 200 watt amp? I am now increasing the surface area by double, and consequently moving twice as much air. Since our ears respond better to changes in Sound Pressure than they respond to power, my rig now sounds roughly twice as loud. The added side benefit is that two speakers are sharing the load and neither one us working nearly as hard as when there was just one. It is now much less likely that my speakers will hit their mechanical or electrical limits.

And how much power would I need to sound twice as loud? About ten times the power. So I would need about 2000 watts to be twice as loud as my 200 watts. And as we've already established, ain't a single speaker on the planet will take 2000 watts. At least not that you me or most anyone else here isgoing to go out and by for their bass.

My rule of thumb (YMMV) is that the RMS rating of the speaker should be about twice the rated RMS output rating for the amp. If I had a 1000 watt amp, I would want a combination of speakers that have a total power rating of 2000 watts. And even then, it may be possible to stress the speakers if I get stupid.

 

[Old Garage-Bander]

https://www.talkbass.com/members/old-garage-bander.274006/ : sometimes i copy and paste a username and it turns out OK, and sometimes not so much.


as above = i copied and pasted, but you see the result = ???

edit

No such thing as bad press.
But you are too kind. Many others contributed good stuff to this conversation.

 

[Downunderwonder]

How does distortion increase power?

Power is the integral of voltage squared over time factored to the load. Distorting the signal frequently increases the average voltage. A fully clipped ''square wave'' makes double the power of a sine of the same peak voltage. But I think you knew that already.

 

[ThisBass]

So a pure sine wave means the voice coil is only at peak current for a tiny moment twice a cycle, whereas a square wave means it's at peak current 100% of the time.

Unfortunatelly real speakers loaded to (ported) cabinets do cause lots of phase shift to the total signal which may make considerations like this ineffective.

A square wave can be roughly approximated by its Harmonics #1 #3 #5 #7 #9.

A typical ported cab may show a phase response in the form of

A 50Hz square wave would show the following "response" due to the cabinets phase response

For 100Hz square wave the square signal shape would then look like the following response

A (theoretical) steady 90 degree phase shift versus frequency would provide the following signal

That's quite impressing. The signal looks similar like a triangle waveform BUT the Harmonics content is nothing but pure square Harmonics shifted by 90 degree each Harmonic

Last but not least a 180 degree phase shift

The signal looks pretty much the same as the "original" signal except the signal now shows inverse "polarity" which is caused by 180 degree phase shift to each Harmonic.

In a summary its comprehensible that the statement

So a pure sine wave means the voice coil is only at peak current for a tiny moment twice a cycle, whereas a square wave means it's at peak current 100% of the time.

is only true for a square wave that shows no phase shift nothing at all or at least for a square that shows phase shift that equals 180 degree steady shift versus frequency.
Its not uncommon for amplifiers and cabinets as well to show particular frequency dependent phase shift which is caused by any types of filters within the signal path. Even cabinets are nothing but a filter that show lots of phase shift versus frequency.

 

[Old Garage-Bander]

Unfortunatelly real speakers loaded to (ported) cabinets do cause lots of phase shift to the total signal which may make considerations like this ineffective.

A square wave can be roughly approximated by its Harmonics #1 #3 #5 #7 #9.
View attachment 2866759
A typical ported cab may show a phase response in the form of
View attachment 2866761


A 50Hz square wave would show the following "response" due to the cabinets phase response
View attachment 2866763
For 100Hz square wave the square signal shape would then look like the following response
View attachment 2866772

A (theoretical) steady 90 degree phase shift versus frequency would provide the following signal
View attachment 2866820
That's quite impressing. The signal looks similar like a triangle waveform BUT the Harmonics content is nothing but pure square Harmonics shifted by 90 degree each Harmonic

Last but not least a 180 degree phase shift
View attachment 2866831
The signal looks pretty much the same as the "original" signal except the signal now shows inverse "polarity" which is caused by 180 degree phase shift to each Harmonic.

In a summary its comprehensible that the statement

is only true for a square wave that shows no phase shift nothing at all or at least for a square that shows phase shift that equals 180 degree steady shift versus frequency.
Its not uncommon for amplifiers and cabinets as well to show particular frequency dependent phase shift which is caused by any types of filters within the signal path. Even cabinets are nothing but a filter that show lots of phase shift versus frequency.

Would it have been more appropriate if BassUrges has used a clipped sine wave in his example rather than a square wave? A square wave is not necessarily a distorted wave, where a sine wave that is severely clipped (distorted), might look similar to a square wave?

 

[lz4005]

One more (slightly related) question; if I'm using an overdrive pedal would it be a good idea to use the active input of an amp? Until recently I always used a bass with active pickups so I never really put much thought into it, but the bass I've been using doesn't have active pickups. Since I'm using an overdrive pedal (and a Thunderbird, which isn't active but seems to have high output for a passive) I'm wondering if setting my amp to the "active" input would generally be a good idea, regardless of what cab I'm using?

Use your ears.
Try both inputs and use the one that sounds better.
Active and passive are meaningless labels for inputs. Only confuses people.

 

[BassUrges]

I appreciate the detail. It has been many years since I worried about these things in detail (physics undergrad and brief stint in grad school), now lawyer by way of LE. I was trying to give a comprehensible example to explain why distortion increases the load on speakers.

 

[ThisBass]

Would it have been more appropriate if BassUrges has used a clipped sine wave in his example rather than a square wave? A square wave is not necessarily a distorted wave, where a sine wave that is severely clipped (distorted), might look similar to a square wave?

It wouldn't change a bit cause even for a slightly clipped sine the flat top of the signal then would show same changes as shown for the square.
Its either way questionable to try to explain amplifier overload characteristic and pending power with sine waves and its (transient) crossing up to a square.
The consideration is (of course) very easy to understand and also helpful to explain some basic things but, probably also suitable to be a good soil just to let grow any kinds of myths (sorry for direct words).

 

[Old Garage-Bander]

I appreciate the detail. It has been many years since I worried about these things in detail (physics undergrad and brief stint in grad school), now lawyer by way of LE. I was trying to give a comprehensible example to explain why distortion increases the load on speakers.

I guess either way, Square, or heavily clipped sine, the resulting load on the amp/speakers will be similar when compared with the waveforms you see with music. It seems, in either case, the area under the curve relates to how hard things have to work?

 

[byacey]

This is correct; essentially the duty cycle of the waveform.

 

[ThisBass]

Let me share with you another

I guess either way, Square, or heavily clipped sine, the resulting load on the amp/speakers will be similar when compared with the waveforms you see with music. It seems, in either case, the area under the curve relates to how hard things have to work?

Probably not.
I think the area under the curve is the most misleading consideration that can ever be done with power considerations.

Let me share with you another basic conisderation which is called the "average rectified value" which actually represents the area under a given waveform.
For a sine wave the "average rectified value" factor equals 0.6366

This is very meanigful for power considerations cause it means that the area under the curve does not apply to power.

For power considerations with sine wave the pending crest number respectively peak to rms number equals 0.707

In different words it can be explained this way.
For same peak number of a sine wave and a square wave the area under the curve was
square: 100%
sine: 63.66%

But for power consideration the sine wave would push 70.7% RMS voltage versus square 100%

 

[Old Garage-Bander]

Let me share with you another

Probably not.
I think the area under the curve is the most misleading consideration that can ever be done with power considerations.

Let me share with you another basic conisderation which is called the "average rectified value" which actually represents the area under a given waveform.
For a sine wave the "average rectified value" factor equals 0.6366

This is very meanigful for power considerations cause it means that the area under the curve does not apply to power.

For power considerations with sine wave the pending crest number respectively peak to rms number equals 0.707

In different words it can be explained this way.
For same peak number of a sine wave and a square wave the area under the curve was
square: 100%
sine: 63.66%

But for power consideration the sine wave would push 70.7% versus square 100%

Yes.
But isn't the more severely the sine wave is clipped, the more it begins to look like a square wave?
I was talking about severely clipped sine waves, (which technically are no longer sine waves) not undistorted ones.

 

[byacey]

Yes.
But isn't the more severely the sine wave is clipped, the more it begins to look like a square wave?
I was talking about severely clipped sine waves, (which technically are no longer sine waves) not undistorted ones.

Chop off everything above the .707 level on a sine wave, and it's a pretty good approximation of a square wave.

 

[Old Garage-Bander]

Chop off everything above the .707 level on a sine wave, and it's a pretty good approximation of a square wave.

Daz what I'z talkin' 'bout.

 

[Drgonzonm]

And if I may, and assuming you and I are on the same page, underpowering refers to the amp not having enough power to do the job. It is not about underpowering the speakers, which the OP asked about earlier. One cannot damage a speaker by underpowering it. If that were the case, every speaker would fail everytime we turned off our amps. Sending no power to it is the the most extream version of under-powering it.

.....

at the risk of appearing to be a troll, eminance in its understanding loudspeakers sub section power handling does warn about underpoweriing speakers.
I picked up a specialized woofer that had a minimum wattage rating as well as a maximum rating. I was researching some thing I saw regarding bass speakers and T/S

 

[byacey]

I don't think under powering is a factor in this instance.

 

[ThisBass]

Yes.
But isn't the more severely the sine wave is clipped, the more it begins to look like a square wave?
I was talking about severely clipped sine waves, (which technically are no longer sine waves) not undistorted ones.

of course it does, no doubt about it.
But the additional power caused by additional Harmonics content does not (directly) relate to the area under the curve.
That's pretty much important to understand real signals like MI signals. Sometimes real signals look alike "lean" with rather poor area under the curve but at the other hand side might be more powerfull than other ones with more of "area" at same peak level number.
Clipped real signals can show pronounced square alike spikes while the RMS remains way below a none clipped sine wave of same peak number.
Of course its possible to overdrive a real signal to get just the same crest 0dB like as is for a square wave but, the clipped real signal yet will look way different (way more complex) than an ideal square!
A real MI signal does NOT clip only one 1st Harmonic. a real MI signal will clip many of its natural Harmonics (prefered the lowish ones) while higher Harmonics may remain unclipped at some content.
That's why it is so tricky to try to derive conclusions from simplyfying models with only one (the 1st) Harmonic within a total spectrum accompanied by lots of dynamics like envelope decay (which is the easiest one to consider on this complex subject).

I think its more practical to consider the lowered crest of the signal caused by clippng and the average RMS. Just for about remaing +3dB crest (the rated amplifier sine wave output power) there will be lots of audible clipping that actually sounds like overkill. So if we would now try to consider this sine to square mutation just to get another +3dB of power we would try to discuss assumptions of overdrive that lack any serious and (most of all) practical sense.

edit,
of course I have to agree that for very excessive settings of a overdrive pedal or preamp drive knob or power amp or anything like "cranked all the knobs all the way up" anything might be possible. For 0dB crest of a signal anything might be possible but, there is no need to try to explain in the time domain with current or different basic stuff just to explain the damage of a device. There are advanced science "tools" out there to describe power mechanism.

 

[byacey]

I re-iterate: Too much power dissipated by the voice coil is what leads to burnt voice coils. It doesn't matter how you get there, the end result is the same.

You can argue this and that, but it doesn't change this simple fact.

 

[Old Garage-Bander]

at the risk of appearing to be a troll, eminance in its understanding loudspeakers sub section power handling does warn about underpoweriing speakers.
I picked up a specialized woofer that had a minimum wattage rating as well as a maximum rating. I was researching some thing I saw regarding bass speakers and T/S

Tell you what. If you can cite your reference, I'll defend your non-troll status.

I did some Eminence + underpower + speaker googling and came up with this.

Can under-powering a Guitar speaker do damage? | Eminence Speaker

It's probably not what you read as they say it's... well read for yourself what they say.

If you can provide a link to the Eminence information you mentioned, I would be interested in reading it.

That being said, please note that I specifically discussed that speakers are not damaged by under-powering them.
See post #41.
That's not to say I wouldn't be interested to read subjects about under-powering in other contexts.

Also please note that woofers and bass guitar speakers are different tools intended for different uses.

Thanks

 

[Drgonzonm]

@ThisBass, I believe the case is worse than you stated. RMS watts is a misnomer. If watts is stated as a sine wave, then the volts, is a square root sine function. I believe the RMS is a voltage reference.
If you look at the power formula, it's VI, Il, or VV, which means watts is sine squared function. Which means the area factor is 0.5 not 0.707. I believe the area to peak value ratio can be called RMS.

As stated, square waves can generated by clipping sine waves, but they can also be generated by adding harmonics, particularly odd harmonics. Something, effects can be designed to do.