Spreadsheet for Calculating Power Amplifier Requirements

[terryforsythe]

The question "how much power do I need?" sometimes is asked. But, the correct answer is dependent on numerous variables. To help, I have created a spreadsheet for calculating amplifier power requirements and made it publicly available at the following link:

Amplifier Power Requirement

docs.google.com

The yellow boxes are for user data entry. I left the permissions on those cells unblocked, so it should work. If not, the equations can be copied into a spreadsheet on one's own computer.

For "Desired Maximum SPL", ensure the value is high enough to cover transients. Adding an additional 15-20dB may be sufficient in most cases, but perhaps a bit higher in some cases (e.g., uncompressed music).

Some amplifiers can output transient power higher than their RMS rating, but that probably varies from amplifier to amplifier and I don't often see that specified anymore. If it is accurately specified, though, the transient power rating would be what to use for comparison to the "Amplifier Required to Get Desired Maximum SPL" assuming that the "Desired Maximum SPL" is set to include transients.

NOTE: This spreadsheet does not consider the room response and is intended to get someone into the ball park of where they need to be with regard to amplifier power. For a more detailed analysis, also analyze the room (e.g., using REW), and take those results into consideration when setting the "Desired Maximum SPL".

EDIT: I have updated the spreadsheet to add estimates for room gain and gain from speaker placement.

 

[Blumlein 88]

Well that really would only work in an open field. I understand why you did it, but it will actually over-estimate power required in rooms indoors. Many listening positions in domestic rooms are around the critical distance where sound from reflections more or less equals direct sound from the speaker. It probably will require 3 db or so less than the spreadsheet shows which is about half the power in an average room.

 

[terryforsythe]

Well that really would only work in an open field. I understand why you did it, but it will actually over-estimate power required in rooms indoors. Many listening positions in domestic rooms are around the critical distance where sound from reflections more or less equals direct sound from the speaker. It probably will require 3 db or so less than the spreadsheet shows which is about half the power in an average room.

NOTE: This spreadsheet does not consider the room response and is intended to get someone into the ball park of where they need to be with regard to amplifier power. For a more detailed analysis, also analyze the room (e.g., using REW), and take those results into consideration when setting the "Desired Maximum SPL".

 

[terryforsythe]

Well that really would only work in an open field. I understand why you did it, but it will actually over-estimate power required in rooms indoors. Many listening positions in domestic rooms are around the critical distance where sound from reflections more or less equals direct sound from the speaker. It probably will require 3 db or so less than the spreadsheet shows which is about half the power in an average room.

That is a very good point, though. Recently I was involved in some of the discussions in the thread for the Topping B100 power amplifier review. I think it has adequate power for a lot of situations. But, there are those who think different, and are adamant about it. I see how it can be very confusing to those just trying to determine which amplifier to buy. I hope the spreadsheet can alleviate at least some of the confusion.

 

[DonH56]

You can look at this online calculator for equations to modify your spreadsheet if you wish:

Peak SPL Calculator

 

[terryforsythe]

I see the spreadsheet is getting used.

I added a note to the spreadsheet as a reminder to add additional dBs to the "Desired Maximum SPL" to account for transients.

 

[terryforsythe]

Well that really would only work in an open field. I understand why you did it, but it will actually over-estimate power required in rooms indoors. Many listening positions in domestic rooms are around the critical distance where sound from reflections more or less equals direct sound from the speaker. It probably will require 3 db or so less than the spreadsheet shows which is about half the power in an average room.

You can look at this online calculator for equations to modify your spreadsheet if you wish:

Peak SPL Calculator

Thank you for the input!

I have updated the spreadsheet to account for estimations of gains from the room and speaker placement.

 

[MRC01]

What's neat about using the linked spreadsheet is that you can see in real time anyone else who's also using it. Seems like mostly a bug since they can stomp on any changes you're making, but partly a feature since it's fun to watch
Of course we can make our own copies.
Another fun thing for Magnepan owners is to set efficiency at 86 and impedance at 4 ohm and see how much power you need

 

[pogo]

I use this one, which also takes the music/sound material into account: Link
What many underestimate is the effect of tone controls/room corrections.

 

[terryforsythe]

What's neat about using the linked spreadsheet is that you can see in real time anyone else who's also using it.

I was just in there adding the cells for room gain and gain from speaker placement.

I'm not sure if the spreadsheet can be downloaded. One way to do so is to copy the spreadsheet and paste it into a locally open spreadsheet program. The equations may not copy directly, though, so you would need to copy them one by one into the cells.

EDIT: The spreadsheet is downloadable. I just downloaded it and the equations came through fine.

 

[Blumlein 88]

What's neat about using the linked spreadsheet is that you can see in real time anyone else who's also using it. Seems like mostly a bug since they can stomp on any changes you're making, but partly a feature since it's fun to watch
Of course we can make our own copies.
Another fun thing for Magnepan owners is to set efficiency at 86 and impedance at 4 ohm and see how much power you need

Even more so if you look at measurements at Stereophile where different models were found to be 80-82 db instead of the spec 86 db. You do get a bit of a boost from it being a panel which isn't radiating over as wide an angle.

 

[paulcc]

There is also a handy app, PAcalculate, to do this calculation and many more. It's on the Apple app store and probably Google as well. Matt Poes suggested it.

 

[kyuu]

What many underestimate is the effect of tone controls/room corrections.

Any idea of a handy rule-of-thumb to account for this? Subtract, say, 3dB to cover it maybe?

 

[pogo]

Depends on the conditions. More dB may also be required for FR compensation.

 

[terryforsythe]

Any idea of a handy rule-of-thumb to account for this? Subtract, say, 3dB to cover it maybe?

It depends.

EQ added at the lower frequencies (e.g., room correction) is more significant with respect to power requirements. Typically, the power used for musical content decreases with increased frequency. So, adding 6dB at 80Hz probably is more of an issue than adding 6dB at 3kHz. See the below RTA spectrum analysis (first graph) I measured with REW (today). I played samples from four songs, louder than I normally listen, and measured the peak values. Values above 600 Hz were down over 25dB from the peak at 90 Hz. (Note that the RTA does not capture the peak transients, though.)

On the other hand, music transients contain more higher frequencies than lower frequencies. So, you probably do not need as much transient headroom in the low bass region in comparison to the higher frequencies. The second graph below is a time domain measurement playing four songs in full (last week), the last two of which I was not able to detect any audible compression. The tan line (LAFMax) indicates the maximum SPL over time and the red line (LZpeak) indicates the peak due to transients. (Not important to this discussion, but the teal line (LAE) is the sound exposure level, i.e., the impact on your ears over time from listening to music at the volume.)

Given the above, adding 15-20dB above your normal listing volume to "Desired Maximum SPL" probably is sufficient to account for a reasonable amount of EQ and transients. As noted, uncompressed music may contain higher transients, and thus increasing the value a bit is not a bad idea if you listen to a lot of uncompressed music. Given my measurements, for my system, which I listen to at 2m, a "Desired Maximum SPL" of 100dB is sufficient. In my family room I sit 3m from the speakers, so I would need a little over double the power if I used the same speakers there.

(Note that I generated the two graphs on different days and did not make sure the volume levels were the same. In each case I just turned the volume up louder than I normally listen. I only included them to help the understanding of the concepts, not as scientific data.)

 

[MRC01]

Even more so if you look at measurements at Stereophile where different models were found to be 80-82 db instead of the spec 86 db. You do get a bit of a boost from it being a panel which isn't radiating over as wide an angle.

Yes and FWIW, Stereophile measured the Magnepan 3.6 at 84 dB / 2.83 V, 2 dB lower than spec. With all that, the above site recommends 400 wpc for me. When people say they suck down power and you need a big amp, they are not joking. Especially for high dynamic range music - and I mean that literally, in other words music that is perceptually mostly "quiet" not "loud". Too many people (present company excluded) think "high dynamic range" = "loud" when nearly the opposite is true.

 

[Blumlein 88]

Yes and FWIW, Stereophile measured the Magnepan 3.6 at 84 dB / 2.83 V, 2 dB lower than spec. With all that, the above site recommends 400 wpc for me. When people say they suck down power and you need a big amp, they are not joking. Especially for high dynamic range music - and I mean that literally, in other words music that is perceptually mostly "quiet" not "loud". Too many people (present company excluded) think "high dynamic range" = "loud" when nearly the opposite is true.

I owned some Maggie 3.3R speakers for a time. The spec on those were 84 db, but I don't know if they were that or a bit lower.

 

[Sokel]

View attachment 397417

(Note that I generated the two graphs on different days and did not make sure the volume levels were the same. In each case I just turned the volume up louder than I normally listen. I only included them to help the understanding of the concepts, not as scientific data.)

We also have to point out that this chart is (A) weighted for the average and max (not for peaks,that's (Z) weighted) .
It would be a lot different (higher) with (C) weighting,specially if the captured music contained a bit of lows.