When we think about watts in a multi way speaker, are there any convenient, fact based rules for how power is distributed in music an therefore drivers?
According to Eminence:
"As a general rule of thumb, we estimate that the high frequency driver will see about 20% of the system power with a 1.5 kHz-2.5 kHz crossover, 15% with a 3.5 kHz-5 kHz crossover, and 10% with a 5 kHz and up crossover (all based on a minimum 12dB/octave slope)."
https://eminence.com/blogs/blog/high-frequency-driver-power-handling
"As a general rule of thumb, we estimate that the high frequency driver will see about 20% of the system power with a 1.5 kHz-2.5 kHz crossover, 15% with a 3.5 kHz-5 kHz crossover, and 10% with a 5 kHz and up crossover (all based on a minimum 12dB/octave slope)."
https://eminence.com/blogs/blog/high-frequency-driver-power-handling
With a 2-way crossover, the 50-50 power split is generally something in between 300-500 Hz in most types of music.
From the Eminence Loudspeaker Book:
| Crossover frequency (Hz) | Percentage power above crossover point |
| 150 | 80 |
| 200 | 72 |
| 300 | 55 |
| 500 | 40 |
| 1000 | 25 |
| 3000 | 20 |
| 5000 | 10 |
Rule of Thumb. On second read I found the meaning changed quite for the better 😀
@Galu
Thanks for that. I'm honestly a little surprised the typical subwoofer crossover area (~80Hz) doesn't reduce power nearly as much as I though. I thought that would be the 40-50% point. That doesn't happen until ~ 300 Hz according to that.
I'm thinking a lot about 2-way vs. 3-way speakers and 2-way vs. 2-way plus subs as I approach my next project. I'm converting a 2-way with 6.5" mid-woofer to a 3-way and based on this chart it seems the reduction in energy in the (currently) midwoofer is substantial.
Thanks for that. I'm honestly a little surprised the typical subwoofer crossover area (~80Hz) doesn't reduce power nearly as much as I though. I thought that would be the 40-50% point. That doesn't happen until ~ 300 Hz according to that.
I'm thinking a lot about 2-way vs. 3-way speakers and 2-way vs. 2-way plus subs as I approach my next project. I'm converting a 2-way with 6.5" mid-woofer to a 3-way and based on this chart it seems the reduction in energy in the (currently) midwoofer is substantial.
Last edited:
I guess they were assumimg equal efficiency over all the ways. When a sub is capable of really deeeeeeeep bass then it usually has amn efficiency that is much lower than all the other drivers in the speaker. In this case the "150Hz/80% above" rukle will not be valid anymore.
Regards
Charles
Regards
Charles
Right, but I wasn't really thinking about efficiency so much as how crossover points affect the power level of a specific driver.
We often say that a high pass filter on your satellites will reduce the power the main amps need to produce and therefore what the satellites get but that is apparently is a lot less than I imagined.
I have had four and five way big systems for a couple of decades now and when working out how much power was required for each band I used a 70 volt / 600 Watt into 8 Ohm amp. Via oscilloscope I measured peak voltage and current with a current loop.
The fact the subs took the lions share of the current was no surprise. What I did not expect was the peak voltage requirement was in the mid bass-lower midrange.
Barry.
The fact the subs took the lions share of the current was no surprise. What I did not expect was the peak voltage requirement was in the mid bass-lower midrange.
Barry.
@1audiohack Good insights. I think it's still true in multi-way systems as we tend to prefer 4 Ohm drivers for the mid-bass or in 2-way systems we often find the minimum Z at around 150 Hz or so.
The crossover points will affect the average power level distribution between drivers.
Assuming flat response and equal sensitivity and impedance for each driver in a multi-way system, the peak voltage required for each driver will be identical for peaks of the same sound pressure level.
I made a spreadsheet a while back that would calculate approximate power to drivers across a 4-way crossover. You should be able to do the same pretty simply, just a matter of plotting a matching curve to that scatter chart (Eminence data) and then doing the formulas from there.
Mine wasn't based on the Eminence data (I think I got the info from a reference on Elliott's pages) but it's likely to be very similar.
Mine wasn't based on the Eminence data (I think I got the info from a reference on Elliott's pages) but it's likely to be very similar.
Thanks for all the ideas everyone. my needs are a lot less elaborate than many of you imagine. The eminence table will really help me going forward.