Loudspeaker Power Handling Vs. Efficiency
-
Strapping Young Stu
- Posts: 1016
- Joined: Sat Dec 30, 2006 7:01 pm
- Location: Dorset, UK
A little quote I just picked up on another forum which might help to put a bit of extra real world perspective on the views of doubling cabinets and stacking rules (some spelling and punctuation corrected)
Rog Mogale
"a wall like a floor will reinforce the output as you now have the speaker and another virtual speaker or mirror of the speakers output adding to it. This is providing the speaker is close enough to couple in wavelength terms to the adjacent floor or wall. So a horn that requires a mouth of 5000 sq cm to have a certain low cutoff can have this figure halved if its in contact with the floor or quartered if its on the floor and against one wall. Hence half space, quarter space and so on.
So at what point does a stack of speakers all combining in phase change from one operating condition to another? Well, if the stack is on the floor then its in half space, but I'm not convinced the more cabs you add the more it operates like it was in quarter space. It will happen to a certain extent but I think you still have a new larger speaker with a longer horn in half space. I never automatically change from half space to quarter space or eighth space if I model more than one speaker. For me its just a larger speaker with a longer horn and bigger mouth if its a horn. Its still in half space. Likewise I always model mid tops in free space even if there are 8 of them. Its still a speaker, all be it a large one thats in free space. Its a difficult question to answer and I have a couple of times before, but a stack of speakers operating conditions don't change as much as you think unless you go very big. The bigger the speaker and the larger its radiating surface is the less it starts to obey the inverse square law. Why? well you are more likely to be in the near field of a sound source if it is marginally larger than you are. Its like you are physically nearer and will also visually look like this as the stack grows.
Its a debate that still continues. I think you still have a larger speaker with increased coupling to the floor, a longer horn path and bigger mouth if using horns and the increase in efficiency that brings, but still working in the same space. You can also add any effects brought about via any decrease on the inverse square law if the stack is large enough. But one thing thats doesn't happen is that you automatically add 6dB everytime you double the amount of speakers. You will get 3dB from double the amp power as a result of the lower impedance and only if the amp delivers twice the output into the new lower load, but you can never get 3dB from a doubling of sensitivity. So be careful when knowing the output of one cab and then addding 6dB everytime you double the amout to work out the new combined output. It will never be as much as you expect due to cabinet losses which destructively combine to reduce the sensitivity and the fact that most amps don't double their output into half the impedance."
I personally found this insight very interesting and it confirms to me that theory and practice are very different.
Rog Mogale
"a wall like a floor will reinforce the output as you now have the speaker and another virtual speaker or mirror of the speakers output adding to it. This is providing the speaker is close enough to couple in wavelength terms to the adjacent floor or wall. So a horn that requires a mouth of 5000 sq cm to have a certain low cutoff can have this figure halved if its in contact with the floor or quartered if its on the floor and against one wall. Hence half space, quarter space and so on.
So at what point does a stack of speakers all combining in phase change from one operating condition to another? Well, if the stack is on the floor then its in half space, but I'm not convinced the more cabs you add the more it operates like it was in quarter space. It will happen to a certain extent but I think you still have a new larger speaker with a longer horn in half space. I never automatically change from half space to quarter space or eighth space if I model more than one speaker. For me its just a larger speaker with a longer horn and bigger mouth if its a horn. Its still in half space. Likewise I always model mid tops in free space even if there are 8 of them. Its still a speaker, all be it a large one thats in free space. Its a difficult question to answer and I have a couple of times before, but a stack of speakers operating conditions don't change as much as you think unless you go very big. The bigger the speaker and the larger its radiating surface is the less it starts to obey the inverse square law. Why? well you are more likely to be in the near field of a sound source if it is marginally larger than you are. Its like you are physically nearer and will also visually look like this as the stack grows.
Its a debate that still continues. I think you still have a larger speaker with increased coupling to the floor, a longer horn path and bigger mouth if using horns and the increase in efficiency that brings, but still working in the same space. You can also add any effects brought about via any decrease on the inverse square law if the stack is large enough. But one thing thats doesn't happen is that you automatically add 6dB everytime you double the amount of speakers. You will get 3dB from double the amp power as a result of the lower impedance and only if the amp delivers twice the output into the new lower load, but you can never get 3dB from a doubling of sensitivity. So be careful when knowing the output of one cab and then addding 6dB everytime you double the amout to work out the new combined output. It will never be as much as you expect due to cabinet losses which destructively combine to reduce the sensitivity and the fact that most amps don't double their output into half the impedance."
I personally found this insight very interesting and it confirms to me that theory and practice are very different.
Last edited by Strapping Young Stu on Tue Feb 26, 2008 6:15 pm, edited 1 time in total.
-
Dave Non-Zero
- Posts: 1939
- Joined: Thu Jul 13, 2006 4:59 am
- Location: Dundee, Scotland
- Contact:
you missed a few.Strapping Young Stu wrote: (some spelling and punctuation corrected)
Rog Mogale
"...space, quater space and so on. "
"operating condidtion to another? "
"the lower impeadance "
I personally found this insight very interesting and it confirms to me that theory and practice are very different.
I understand the fact that most amps dont deliver exactly double the power into half the impedance, but what are 'cabinet losses'?
i also think that you cant think of your speakers as having 104db efficiency and doubling gives you exactly 111db efficiency with a peak of 13xdb. Its been shown enough times on this forum that so many factors come into effect in practise like displacement limits, power compression etc, that assuming 6db per doubling is at best a rule of thumb, not a mathematical certainty. In other words, i agree.
-1 for thought terminating cliches.
Built and/or own:
8 x T48 24" 3015LF
6 x DR280
2 x DR250 old style beta10
2 x T36s 20" delta15L
1 x TAT
1 x dual Lab12 30" T60
In Progress:
2 x DR280
Built and/or own:
8 x T48 24" 3015LF
6 x DR280
2 x DR250 old style beta10
2 x T36s 20" delta15L
1 x TAT
1 x dual Lab12 30" T60
In Progress:
2 x DR280
-
Strapping Young Stu
- Posts: 1016
- Joined: Sat Dec 30, 2006 7:01 pm
- Location: Dorset, UK
-
Bill Fitzmaurice
- Site Admin
- Posts: 28646
- Joined: Tue May 02, 2006 5:59 pm
Many times I've measured one and two identical cabs with the same amp driving them and the increase at low frequencies is always 6dB. If the voltage remains constant and impedance is halved current, and therefore power, will double and the small signal LF output will go up by 6dB every time.
Where confusion can arise is that when the driver/cab count is doubled the high frequency roll-off frequency of the system will go down, by a factor of 0.7, so the 6dB figure does not apply broadband.
As far as amp ratings are concerned, they are quoted at full power. You won't get a doubling of the full power rating with halved impedance load. But you will at small signal levels.
Where confusion can arise is that when the driver/cab count is doubled the high frequency roll-off frequency of the system will go down, by a factor of 0.7, so the 6dB figure does not apply broadband.
As far as amp ratings are concerned, they are quoted at full power. You won't get a doubling of the full power rating with halved impedance load. But you will at small signal levels.
-
Strapping Young Stu
- Posts: 1016
- Joined: Sat Dec 30, 2006 7:01 pm
- Location: Dorset, UK
When you have a mouth as big as this the effective horn length tends to appear a little longer. It a smaller effect of the v-coupler. The quote above was in response to someone asking about how modelling in half space and quarter space actually holds up in real life and whether when you model with multiple drivers it should be in half space or quarter space. Rog appears to have found more correlation between real world measurements and modelling in half space, rather than quarter space.Ron K wrote:How does the horn length change with stacking more cabs???new larger speaker with a longer horn in half space
I can see the mouth size changing but isn't the horns length of each cab still the same?
And yes those are giant scoop bins - 5.4m horn which loads to 25Hz - Even in Rogs opinion they dont have many practical applications.
Stu
-
Bill Fitzmaurice
- Site Admin
- Posts: 28646
- Joined: Tue May 02, 2006 5:59 pm
That will be the case with almost any grouping of horns, but especially with bass horns. It has nothing to do with the path length, and everything to do with total mouth area. The SPL charts for groups of Tubas and Titans show in detail what happens with stacks of horns. As for boundary reinforcement, coupling with a boundary occurs when the source is less than a quarter-wavelength distant, pure and simple.Strapping Young Stu wrote:
When you have a mouth as big as this the effective horn length tends to appear a little longer