Can you damage equipment if you under power a speaker?

[philw44]

The theory goes that an undersized amp will clip before the max wattage of the speaker.

 

It will do if your gain staging is incorrect. In such a case it would be operator error causing a low powered amp to clip and blow the speaker, not a problem of the actual amplifier being underpowered.

 

D-mills you make good points about using extensive DSP processing to regulate the load. That is how to do it safely, but without such precautions I can't see the point.

 

I have not heard any systems that sound noticeably better when pushed to the their operating limits (usually worse). Oh, and if we are carrying on with transport analogies - My car can do 120Mph, I'm entering a 30Mph zone. My car performs comfortably at the speed of 29mph however I could quite easily have a rush of blood to my foot and hit 31Mph. Most of the time the speed camera won't flash. Another time it might

 

Having more rig is safer and always sounds better.

[Dmills]

D-mills you make good points about using extensive DSP processing to regulate the load. That is how to do it safely, but without such precautions I can't see the point.

The point is that empirical experience is that peak/average power is almost always > 2, so an amp having twice the cont. power handling of the speaker is actually quite safe (Until you run it into clipping (which is effectively limiting and raises the average power)).

This incidentally is why most modern amps will not manage continuous rated output, think about it, most of them are plugged into 16A supplies limited to ~4KW continuous, but the short term power (~a few hundred ms maybe) could be north of 6KW. There is nothing to be gained by designing an amp that can deliver the same power as a day long sine wave as it can for a hundred ms burst, because nobody needs a day long sine wave, but having a hundred ms at a time at +3 or +6 over long term is actually very useful.

 

Having more rig is safer and always sounds better.

And weighs more, and takes up more truck pack, and combs like buggery if you don't have not just twice the rig, but the RIGHT twice the rig. An extra 6KW in the amp rack is maybe another 15Kg and 2U on a bad day, twice the number of boxes (which have just all become horn loaded mids to get the dispersion down to 30 degrees), is another trip in the transit.

 

Like everything it is a trade off, there have been gigs where I could usefully have had 4 times the continuous rating of the boxes and it would have been safe, but a factor of two is usually safe even without extensive DSP support as long as you don't bottom out the drivers in the magnet (Only generally a problem with dance music, most live stuff does not have much energy below 80hz or so).

 

Now, bringing enough rig for the gig is of course vital in whatever way you achieve it.

 

Regards, Dan.

[StevieR]

Going back to the OP, there is no direct damage caused by an amp to a loudspeaker as a result of the amp being underpowered. However, there is an increased risk of the input signal becoming clipped or distorted as a result of lack of system headroom - i.e. the operator is trying to get more level from the system and runs the gain struture into the red. This WILL result in loudspeaker damage.

 

So, in summary, underpowering a loudspeaker CAN cause damage to the loudspeaker - but not directly!

 

My 2p...

 

Steve

[Ken Hughes]

I've watched this topic from the start and Steve's question was simply ".. could any damage occur by under powering

these speakers ..".

 

Forgetting frequencies and other technical jargon as they don't matter a jot in this instance the answers are;

1) underpowering sensibly and with care - NO;

2) driving a lower power amplifier to extreme into ANY loudspeaker - YES;

 

The simple truth is this taken from more years experience than many of you have been alive.

In basic lingo if you stick too much welly up ANY loudspeaker's backside for too long it will eventually die.

If you don't overdrive your amp and therefore you don't create any nasties in the system you have a racing

chance of survival.

 

Unfortunately there are other factors involved which muddy the waters such as cheap, nasty drivers etc. but

in general if you don't go wild you'll be ok.

Bit like driving a car on an icy road.

[paulears]

I suspect most people who have worked, as I have, in the music department of colleges have seen many wrecked drivers caused by the poor handling of too small amps, cranked up to eleven! Stacking up a host of stomp boxes makes the problem even worse. Of course, it's the users who are to blame, wanting more output than the system is capable of.

 

I subscribe to the bigger amp than speakers method of system planning. Others don't - I can live with that. I fundamentally disagree with quite a bit of the stuff posted here, but it's a free world. All I can do is draw on my own experiences which have shaped my method of working. I'm quite happy with the technical function of the knob on a power amp - but the practical effect of turning it anticlockwise is that power output goes down. Jacking up the input level to the amp restores it - but watt (pun intended) is absolute is that it changes the volume. No point being pedantic about it's operational characteristics - it's placement in the gain structure of the entire system is to allow adjustment of what comes out the other end, with the system running at optimum signal to noise. Some people obviously use them for volume adjustment - and if your system has just a left and right full range loudspeaker on the end, I think this is an acceptable 'wrong' application in some systems. Everyone has their own way of working that suits them. Some operate with the controls fully clockwise, and make level changes in another device. Some people insist on powering up and down with the knobs fully anti-clockwise. Others operate with the gain structure set so that 50% works for them. No one system is right for every case.

 

What exactly are we arguing about here? Ken summed it up rather well.

1) underpowering sensibly and with care - NO;

2) driving a lower power amplifier to extreme into ANY loudspeaker - YES

 

Nobody can argue with that one, can they? All we're doing is playing with the ifs and buts.

[Brian]

So, in summary, underpowering a loudspeaker CAN cause damage to the loudspeaker - but not directly!

And to really confuse things it depends how underpowered the amp is.

[mackerr]

We've got to be talking unrealistic frequencies though haven't we?

 

What routine example can you give where x sized MF/LF driver takes twice the rated RMS?

 

Every speaker. How much power it takes to damage a speaker is dependent on both the maximum power, and the time over which it is applied. Virtually every manufacturer has a peak rating of twice the RMS rating. It is either heat, or over excursion that damages speakers, and even the peak rating should not create over excursion no matter the time, and the heat build up will be dependent on how much power, over how much time, and how conservative the manufacturer was in their rating.

 

Clipping does not create DC, it creates more high frequency content, which will raise the ratio of power to the high frequency drivers, which may in turn damage the driver.

 

Mac

[J Pearce]

Clipping does not create DC, it creates more high frequency content

 

FFT analysis of a square wave will show lots of high frequency content, due to the snap from -ve to +ve, but once at the top of the square, at low frequencies, you have a dc peak.

 

So confusingly it is not both high frequency and dc?

[mackerr]

Clipping does not create DC, it creates more high frequency content

 

FFT analysis of a square wave will show lots of high frequency content, due to the snap from -ve to +ve, but once at the top of the square, at low frequencies, you have a dc peak.

 

So confusingly it is not both high frequency and dc?

 

No, because it is still AC between +Ve and -Ve. The fact that the waveform is flat topped means there is lots of high frequency, but it is still AC. If it was DC wouldn't the plus and minus components sum to zero?

 

Mac

[mervaka]

a square wave is made up of odd harmonics. it's the waveform that contains the most energy, ie the area inside the waveform. this is one good reason why they are dangerous.

 

so if you pump a square wave into an amp at saturation level, you'd be ragging the tits off it!

[Brian]

A square wave contains twice as much power as a sine wave of the same peak-to-peak amplitude.

[mervaka]

I thought it was √2 times greater?

[Rob_Beech]

The little tag that Paul DOES leave out (though it should go without saying really) is "With the input signal as a constant". Providing this is true then the controls on the amp will adjust wattage. It's not "What they do" as such, just what happens is the input level remains the same. They ARE sensitivity controls, you can get the maximum power with them turned down (there is a limit on this you know, you can overload the input stage and cause clipping here, if you haven't already from your next device back up the line, so best to leave these up and control from elsewhere as is the norm).

 

Larger amps have better control over a driver or series of drivers. This is more important and indeed more noticeable on LF drivers. particularly those with larger diameter.

 

500 watt driver (in a suitable box that allows this power without over excursion.... remember that one) with a 500 watt amp. Everyone really should be happy. But to get the maximum continuous level out of the box, you push the amp to its limit. You have no headroom so your continuous level MUST be lower than this if your program material has any dynamic range at all. So with a 500 driver on a 500 watt amp, if you want dynamics ad don't want to make your amps sweat too much, your continuous level would expect to output maybe half of this. 3dB lower in theory.

 

This is all well and good, but 3dB is useful in places. Now lets look at a 1000 watt amp on the same driver. Again, providing the driver and enclosure are suitably matched and of a passable quality, there is no reason why the driver won't handle peaks of 1000watts, if it doesn't, its poor to say the least. (John, I don't own a box that won't comfortably handle twice its rms rating (and more) for some time, significantly longer than a second). If we give ourselves the same headroom of 3dB (it's comically low but its for the sake of making sense of it all) then we get a continuous rating of around and upto 500 watts (which the driver is happy with) and peaks up to 3dB louder which again both the driver and the amplifier are happy with. You've suddenly got 3dB more headroom in the system with the same number of boxes.

 

Where it DOES go wrong is if you forget that that 3dB is HEADROOM, for occasional peaks and drive the amp into this level more continuously. The amp will be more than happy but over time the voice coil will heat up and if you carry on, eventually it will fail.

 

Mac, most drivers have a peak rating of 4 times the RMS rating, they tend to have a program rating of 2 times the rms. The program rating on the speaker says exactly this. This is what I'll handle for short peaks and it's ideal for headroom.

 

DSP can be helpful here, you can have a slow attack compressor / limiter for when the output goes above -3dB (or equivalent within your system gain structure that makes the amp travel above -3dB) it can allow short transients but throttle anything longer and larger.

 

 

As for more speakers not bigger amps. Dan hits the nail on the head. Subs are slightly more forgiving in their positioning in some ways, infact often with correctly spaced subs you can improve the way the LF is reproduced and get a more even coverage, but this is not through having not enough amp power. As for mid/top. You put in the number of boxes that you need to cover a given area, if that area stays the same but you need more power, you need more power from the amps IF the speakers allow it, or you need speakers that are more efficient. More speakers will result in massive coverage issues with comb filtering and lots of phase issues. That's before we go into the sight lines from the people who have paid to see the show, and as Dan mentions, space in the vehicle, weight, setup time.

 

 

 

To the original question now, Steve. Be careful. You'll be fine I'm sure. It'll sound fine and it'll probably be loud enough. But just as with a large amp, you need to be careful how hard you drive it as it can clip earlier than you think. You see a big high power speaker and naturally expect a big sound, if you don't get it you turn it up and then the underpowered amp struggles. I can't see you having problems with it but :

 

A) You won't get the maximum volume out of the speaker that you could with a suitably rated amp

B) A suitably rated amp for this box would cost a fortune

C) Control in the LF might not be as good at any volume though this is not an issue with purely vocal monitoring.

D) Have fun with it.

[Brian]

Avoiding Integrals so early in the day let's do it empirically...

 

 

Let's say we have +/- 40V DC rails in our ideal amplifier.

 

We can swing the 8 Ohm load +40V and -40V to make a sine wave.

 

The sine wave will have an RMS value of 40/√2 = 28.3V. In a 8 Ohm load this means that 3.5 A RMS will flow (28.3/8).

 

So our average power will be 28.3V * 3.5A = 100 Watts

 

We can also swing +40V and -40V to make a square wave.

 

With the load at 40V we have a current of 5A (40/8) giving us a power of 40V * 5A = 200 Watts.

[mackerr]

Mac, most drivers have a peak rating of 4 times the RMS rating, they tend to have a program rating of 2 times the rms. The program rating on the speaker says exactly this. This is what I'll handle for short peaks and it's ideal for headroom.

 

Unfortunately, none of these specs say "exactly" anything. There are no standards set by AES, or any other governing body, about how any of these measurements are made, nor what kind of failure determines the upper limits of performance. They are all simply guidelines determined in an unspecified way by each manufacturer. The manufacturer gets to pick the waveform, and the time frame for any of these specs, none of which is reflected in the data sheets that I have seen.

 

Mac