No announcement yet.

HT - How Little power is Surprisingly Needed?

  • Filter
  • Time
  • Show
Clear All
new posts

  • HT - How Little power is Surprisingly Needed?

    I had a conversation with a salesman the other day that got me thinking about how much power is really needed for a home theater vs what a salesman might want to sell me.. Mine HT is 500 square feet and the math I did surprised me, so I thought I better run this past the people that really know this kind of thing.

    A little googling found that peak commercial theater sound levels are around 95 decibel, but that is really loud. Noise of 90 decibels feels loud to the ears. Lawmakers in California are trying to pass legislation to restrict commercial theaters to a maximum of 85 db. So 95 db gives plenty of head room.

    I'm going to focus on the subwoofer since I know I'm going to use two 12" ultimax subs. Maybe four later, but starting with two. These subs are rated at 86.7 dB for 2.83V/1m. I am going to make an assumption that equates to 86.7 db / watt since that is a standard rating, but I'm lost on how someone gets from 2.83 volts with a two ohm speaker to one watt. Sure P = EI, but I'm just going to push the "I believe" button until corrected.

    The primary seating position is ~4 meters from the subs (and the L-C-R speakers). So to reach 95 db at 4 meters I would need to double that output twice at the source. I know there is some benefit from reflection, but I'll negate for calculating max power. So 95 db + 6db + 6db = 107 db. Doubling power increases out put 3 db, so from 1 watt I would need to double power 4 times to 8 watts.

    Well that can't be right so lets through in 3 db of head room. Now we are at 16 watts. Ok another 3 db of head room = 32 watts.

    Clearly I must be doing something wrong or there is a big misconception out there about how much power is "really" needed. A 1000 watt amp would generate 118 db. Or more reasonably a 250 watt amp could generate 111 db. Still way too much.

    Did I go horribly wrong somewhere or miss some serious consideration?

  • #2
    P = V² / R

    The "2.83V/1m" rating is practically equal to "1W/1m" when looking at 8Ω nominal loads. When 2.83V is going into a 4Ω load, that equals 2W input power. So to get the 1W/1m equivalent, subtract 3dB from the 2.83V/1m rating of said 4Ω speaker.

    The sensitivity rating of a speaker is typically in the mid-band (EBP, aka Fs/Qes, up to 2*EBP). Below that it is largely governed by the low-end alignment. You also have to keep in mind the sensitivity of the human ear to different frequencies. Midrange at 95 dB is a lot different than 30Hz at 95 dB, as we perceive it. Make no mistake about it, to reproduce very low bass frequencies you need a lot of clean power AND lots of headroom for dynamics. Midrange and treble - not so much.


    • #3
      Correct on all points, but the fletcher Munson curve is esentilly flat to 1,000 Hz at 90 db and totally flat at 100 db. So no extra power needed there. It's been a long time, but I seem to recall that efficiency was generally tested at 1,000 Hz. How is that suppose to work for a subwofer? Makes no sense. At least yet.

      After thinking about this for a while I have concluded teh industries subwoofer efficiency ratings are if not meaningless, surely subject to enough variance that they don't mean much. At what frequency is the rating determined - not specified., could be outside it's intended operating range. Under what conditions were they deteremined, anechoic chamber or closed room - not specified. DVC subs - with coils in series or parallel - not specified. Recommendation is to buy a big amp. OK. I guess I'll give in. Sure woudl have been nice to have used math though. Oh well.


      • #4
        Duplicate post deleted.
        Last edited by richlandrick; 05-26-2019, 04:20 PM. Reason: Duplicate Post


        • #5
          If you use WinISD and input the T/S parameters for a driver you can see the SPL chart for different power inputs (watts or voltage) and different distances (in meters). Would that help check/confirm your estimates?


          • #6
            Model the sub in its enclosure in WinISD. You can see SPL over the frequency range graph (Bode plot) given a signal input (e.g., input power - watts). Here's an example showing the various WinISD parameters for a vented 6.5" TangBand sub at 50 W.

            Click image for larger version

Name:	Sub.png
Views:	1
Size:	202.9 KB
ID:	1413955


            • #7
              Way cool. Thanks.


              • #8
                Originally posted by richlandrick View Post
                Sure woudl have been nice to have used math though. Oh well.
                You still can, you just need to know the math. You do have a copy of Vance Dickason's "The Loudspeaker Design Cookbook", right?


                • #9
                  This is for the um15, not the 12. 10 watts and 95db seems absurdly wrong ...

                  Click image for larger version  Name:	Capture.JPG Views:	1 Size:	175.5 KB ID:	1413968


                  • #10
                    Bass is separate from the main speakers. Full range doesn’t use nearly the power that people think. Back in the day I had Magnepans on a 10W Class A amp. Low efficiency speakers in a low power amp. In a medium size room it was always enough power for listening. And having an external powered sub takes care of the power needs.


                    • #11
                      Leave the Fletcher-Munson curve out of it completely. It doesn't apply here and will only confuse things. 87dB + 20dB = 107dB, so you'd need 100W. Add in 6dB of headroom and you're at 400W. If the Ultimax sub is 4 ohm and sensitivity rated at 2.83V, then it is only 84dB @1W and these numbers all double.

                      Tall Boys
                      NRNP Computer Sub
                      The Boxers
                      The Hurricanes
                      The Baronettes


                      • #12
                        Originally posted by Kornbread View Post
                        This is for the um15, not the 12. 10 watts and 95db seems absurdly wrong ...

                        Click image for larger version Name:	Capture.JPG Views:	1 Size:	175.5 KB ID:	1413968
                        Maybe a question of how the T/S parameters are stated vis-a-vis the dual coils ....


                        • #13
                          Originally posted by Kornbread View Post
                          This is for the um15, not the 12. 10 watts and 95db seems absurdly wrong ...
                          I'm hoping someone else chimes in on the math, because I can see in WinISD this isn't exactly right (likely due to the dual voice coils)...But I don't see why this is "absurdly" wrong. (I'm assuming you were using 1 m listening distance.) According to the spec sheet the Ultimax 15 is 86.5 db sensitivity at 2.83v and is dual-voice-coil 2+2=4 ohms impedance. Thus, 86.5 db at 2 watts at 1 meter.
                          Increasing watts to 4, then 8, then 16 produces 89.5, then 92.5 then 95.5 db SPL. Which is not far from what you have. This does not account for the headroom peaks people are talking about.

                          That is at 1 meter. Doubling the distance reduces SPL by 6 db (I think??). So if your listening distance is 4 meters (using 4m rather than 3m so I can deal with doubling), you have lost 12 db and need to double the 16 watts 4 more times. So you need 256 watts to produce 95.5 db at 4 meters. If you want 20 db headroom for peaks, that is a lot of power. But your modelling seemed reasonable for the assumptions that were being used.


                          • #14
                            I've been modelling drivers for the last how ever many months, designing for a dual 15" + horn. With the dual 15"' Lavoce 154.02, 116db at 26 watts, comes to mind. Dual Acoustic Elegance AETD15M-8, 116db at 34 watts. Dual FailtalPro 15H520, 116db, 62 watts. All three of these setups are geared towards signal fidelity using the JBL 2216ND (M2 Woofer) as a reference point of maximum allowed excursion before IMD is a factor in a 2 way configuration, and in a dual woofer situation have half as much as the hypothetical status quo based off of the single JBL. Is that Hifi enough?

                            I forgot to say that for modelling, this is at 1 meter.
                            Last edited by camplo; 05-27-2019, 01:25 PM.