Power requirement?

Greetings all,

I have hesitated to which sub-forum to post the topic, but as it is relevant to a low-frequency part of a multi-way speaker, I believe that this sub-forum is the best.

Background

For one reason or another, I had to give away my fully horn-loaded system and live with headphones. I am now in a position to acquire/build a pair of speakers, though not by replicating my former system. I have been contemplating a three-way, preliminarily a MEH above a 15 inch low-frequency driver. I seem to have a preference for a sealed box; which may well be because something always bothered me in a certain frequency portion of ported box. So, I will limit the following drivel to that topology.

Problem definition/proposed standard

From my understanding, one of the criticism of sealed box, and I am paraphrasing, is a “lack of low frequency extension and low efficiency”. Admittedly, this is a rather ill-defined statement, so let me attempt to qualify it vis-à-vis my preferred music genre.

As I tend to listen mostly to classical music, I have found two topics that appear to be relevant: “What high efficiency speaker for Classical Music?”, https://www.diyaudio.com/forums/multi-way/292232-efficiency-speaker-classical-music-26.html, and “A Test. How much Voltage (power) do your speakers need?”, https://www.diyaudio.com/forums/multi-way/204857-test-voltage-power-speakers.html.

I understand that different people may have different opinions regarding the required SPL, still, given the title of at least the first topic, one would expect at least some commonality, but the amplifiers range from a 2A3 to high-power solid state, and associated differences in the amplifier’s power.

In an attempt to understand the power requirement, I propose the following. According to my understanding majority of classical pieces do not have much content below about 40 Hz. Of course, there are exceptions, e.g., P.I. Tchaikovsky - 1812 Overture; R. Strauss - Also sprach Zarathustra; some G. Mahler pieces – 8th symphony, but they are that – exception. Since I live in a quiet environment and my listening preferences are on average 80 dB or less, assuming that the dynamic range of classical music is 20-30 dB it follows that 110 dB at 40 Hz peaks should be the standard.

Method

1. Driver selection

My first naïve idea was to look for high-efficiency 15 inch drivers, but, as I quickly found, they have about the same efficiency, cf., Table 1. So this does not appear to be a silver bullet.

2. Required power vs Vb

Recalling J. A. Hofman’s “iron law”, I decided to investigate the relationship between an enclosure’s Vb and a power that would satisfy the above-proposed standard. I started with Vb=100l, not due to any sophisticated reasoning, but simply because I had built such an enclosure; the idea having been that I had to start somewhere, and by comparing a model for the specific Vb with a measurement, it may be reasonable to correlate to different volumes.

Using Unibox v. 4.08, I had adjusted the power to the required SPL at 40 Hz. I repeated the exercise for different volumes, cf., Table. 2.

Notes: (i) Table 2 summarizes the model without any consideration for a room gain;
(ii) I was not concerned about exceeding the Xmax for the frequencies below the 40 Hz as this is an implementation and not conceptual issue.

Conclusion/Discussion

1. The parameters Fb and even more F3 are not strongly dependent on the change in Vb;
2. There is an inverse relationship between the Vb and the Qts;
3. There is an inverse relationship between the Vb and the power, which agrees with the “iron Law”; and
4. There is no escaping the conclusion that, assuming that the above-defined standard is to be achieved, a power higher than presented in the links is needed.

Provided that the conclusions are correct, and please do not hesitate to point any flaws in my diatribe, the above now presents at least two questions:

1. What would be the best architecture for an amplifier that will be most of the time supplying power 30 dB below peak, considering preliminary frequency of operation 40 Hz – 350 Hz?; and
2. Disregarding WAF, is there any guidance on selecting the Vb, considering the Vb to Qts relationship?

Kindest regards,

M

Table. 1

Driver Fs [Hz] SPL /1W/1M [dB]
TAD1601A 28 97
TD-15M 34.7 97.8
JBL2226 40 97

Table. 2

Vb [l] Fb {Hz] F3 [Hz] Qts Max SPL [dB] Power [W]
101.7 56.1 60.81 0.655 116.30 155
111.9 54.15 60.81 0.632 116.30 145
122.1 52.47 60.26 0.617 116.10 140
132.3 51.01 61.38 0.601 116.00 135
142.4 49.72 60.81 0.585 116.00 130
152.6 48.57 61.94 0.576 115.60 125
 
I would think something like Troels Gravesen's SEAS 3-way classic would hit the spot as a classical music speaker, and reasonable appearance:

SEAS-3-Way-Classic

The sort of bass you can get really depends on your room. I suppose Troels' speaker is for a reasonably large room. And a good acoustic is always helped by lots of soft furnishings.
 
If your target for f-3dB (?) is 40Hz, which driver would you need, presuming you are not in for active correction? I missed your criteria I guess, but the drivers you mention won’t work as they are deigned for vented (BR) use.

Regarding amps: any decent AB-design or D-design will do. Even a mediocre amp will, in fact, because you’re contemplating the frequency range where the room response reigns.
 
I think your question shows two important aspects:

1) seems to me you are overthinking this 🙂

2) you already did excellent research and calculations, so much so that you are in much better condition to answer this than any of us; I would grab a nice cup of tea, play some soft Classical Music in the background, and carefully read, reread and weigh your own data to reach a conclusion.
 
Conclusion/Discussion

1. The parameters Fb and even more F3 are not strongly dependent on the change in Vb;
2. There is an inverse relationship between the Vb and the Qts;
3. There is an inverse relationship between the Vb and the power, which agrees with the “iron Law”; and
4. There is no escaping the conclusion that, assuming that the above-defined standard is to be achieved, a power higher than presented in the links is needed.

Provided that the conclusions are correct, and please do not hesitate to point any flaws in my diatribe, the above now presents at least two questions:

1. What would be the best architecture for an amplifier that will be most of the time supplying power 30 dB below peak, considering preliminary frequency of operation 40 Hz – 350 Hz?; and
2. Disregarding WAF, is there any guidance on selecting the Vb, considering the Vb to Qts relationship?

Kindest regards,

M

M,

I'd suggest your conclusions are correct in spotting correlations, but they're incomplete overall: there are many Thiele-Small parameters which will make a difference to the resonances, roll-offs, and box sizes. Narrowing it down to Qts alone is an over-simplification.


To answer your questions:
1 - I'd pick something meaty and solid-state.
2 - Lots of guidance, but it's in the simulators. As noted above, they utilise all the T/S parameters, and will therefore provide more useful results than mere Vb:Qts ratios.


Some further notes:
- When it comes to sizing a sealed box, my recommendation would be to size the box so that Xmax is reached when the amplifier reaches maximum output. As you make a sealed box smaller, increasingly large power inputs will be required to utilise the maximum (mechanically-limited) output from the driver. If you make the box (IMO) too-small, then it becomes impossible to use all of the output, because the amplifier can't drive the speaker hard enough.
There are counter-arguments, though: a smaller box (as you've noted) is less efficient, so the voice coil & magnet are dealing with larger power inputs (with flux modulation etc, that'll lead to increased harmonic distortion) than a larger box achieving the same SPL.

- 40Hz is, in my opinion, the minimum target for LF extension: I even make sure my PA speakers (where maximum output is a very high priority) are flat to 40Hz. For home HiFi, 30Hz is a better target IMO.

- The intended bandwidth of these woofers is the range where room-related issues are at their most noticeable. I would strongly recommend implementing some DSP, even if it is just for these drivers. My current system has a pair of 8" woofers per side in sealed boxes, and I'm flat to 10Hz, thanks to a very helpful room. Without EQ, the bass is actually over-blown: there's a +15dB mountain at 40Hz, and still enough "help" at 10Hz to get 10Hz level with >100Hz. A flat in-room LF response is a wonderful thing.

Chris
 
Greetings all,

first, thank you all for the replies. I do appreciate them, although I cannot quite correlate some of them with the concept I was trying to understand. To wit:

Summary; power is cheap. Spend your time on things that matter in the design.

Well, yes, power can be cheap, if one did not try to get 150W in class A. 😀. However, before I started the investigation, I did not even know what power I needed.

Hi System 7,

I would think something like Troels Gravesen's SEAS 3-way classic would hit the spot as a classical music speaker, and reasonable appearance:

thank you for your reply, though I am not sure, how is it relevant. Apart from not having asked for a specific speaker recommendation, with all due respect to Troels Gravesen, apparently, many people like his designs, he would not be on my list.

Hi Markbakk,

thank you for the reply.

If your target for f-3dB (?) is 40Hz, which driver would you need, presuming you are not in for active correction? I missed your criteria I guess, but the drivers you mention won’t work as they are deigned for vented (BR) use.

No, you did not miss anything. What I was trying to understand was how much power I need for a target of 110 dB/40Hz. I specifically stated that I am, at this point not concerned with an implementation.

I used the TAD1601A as an example, since I do have it. I calculated its Efficiency Bandwidth Product (EBP) of 77, which according to the "guidelines" indicate that albeit perhaps not ideal, the TAD does satisfy the target. Additionally, Jeff Bagby noted:

Where you are headed is correct. If the modeling indicates that it will work well in a given application, either ported or sealed, then that is what you need to go by. EBP is only a general guidline, mostly from a time when modeling wasn't as easily accomplished.
, cf., EBP--Ported or Sealed -

Techtalk Speaker Building, Audio, Video Discussion Forum
. Similarly, if I interpret Dr. Geddess correctly, as long as the driver can be equalized for the target, that is all that matter.

And my simulation data indicates that, albeit perhaps not the ideal, the TAD will work.

Regarding amps: any decent AB-design or D-design will do. Even a mediocre amp will, in fact, because you’re contemplating the frequency range where the room response reigns.

Thank you.

Hi JMFahey,

thank you for the reply.

1) seems to me you are overthinking this 🙂

Could you please be more specific as to what am I overthinking?

2) you already did excellent research and calculations, so much so that you are in much better condition to answer this than any of us; I would grab a nice cup of tea, play some soft Classical Music in the background, and carefully read, reread and weigh your own data to reach a conclusion.

Thank you for the vote of confidence, but, as you can gather from Chris666's answer below, there is a lot I do not know.

I want to emphasize that I am mostly concerned with the LF portion, I had successfully build the MEH that will go above before.

Hi Chris661,

thank you for your comprehensive reply.

I'd suggest your conclusions are correct in spotting correlations, but they're incomplete overall: there are many Thiele-Small parameters which will make a difference to the resonances, roll-offs, and box sizes. Narrowing it down to Qts alone is an over-simplification.

Can you please elaborate? My understanding has been, that one wants:
(i) a low Free air resonance (fs), to push the roll-off as low as possible, (ii)an Xmax value that together with the driver's Sound Pressure Level allows for the target SPL to be reached, and, (iii) if one still insist on the EPB guidance, proper Qes parameter.

Once these parameters are set, there is a direct relationship between the remaining parameters Qtc and Vb, in that one can select Qtc and calculate Vb or vice versa.


To answer your questions:
1 - I'd pick something meaty and solid-state.

Indeed, I was just wondering, given the dynamic range, whether an AB or class D would be better for handling this.

2 - Lots of guidance, but it's in the simulators. As noted above, they utilise all the T/S parameters, and will therefore provide more useful results than mere Vb:Qts ratios.

Please see my request how to I approach this above.

Some further notes:
- When it comes to sizing a sealed box, my recommendation would be to size the box so that Xmax is reached when the amplifier reaches maximum output. As you make a sealed box smaller, increasingly large power inputs will be required to utilise the maximum (mechanically-limited) output from the driver. If you make the box (IMO) too-small, then it becomes impossible to use all of the output, because the amplifier can't drive the speaker hard enough.
There are counter-arguments, though: a smaller box (as you've noted) is less efficient, so the voice coil & magnet are dealing with larger power inputs (with flux modulation etc, that'll lead to increased harmonic distortion) than a larger box achieving the same SPL.

As I understand the simulations, this is a a chicken-and-an-egg problem. The combination of Xmax and the driver's Sound Pressure Level, indicates, that for 110 dB at 40 Hz, one needs power allowing about 7.8-7.9 mm Xmax. This can be achieved by an infinite combinations of Vb and power.

And I cannot quite qualify it in my mind. Initially, I though about selecting Vb to minimize power, for the reasons that you had stated. That would imply larger Vb.

But, then I read complains about (i) larger Vb enclosures having "boom bass", whatever the term means, and (ii) I started to wonder, whether it does matter practically. Remember, my average listening level is 80dB or less, for which at, e.g., Vb=120l the model implies 0.05W. Will the occasional peaks do cause thermal/flux modulation problems?

I am afraid that this is an unanswerable question, because one would have to analyze the time interval of peaks-to-troughs.

Hence my quandary regarding the Vb.

- 40Hz is, in my opinion, the minimum target for LF extension: I even make sure my PA speakers (where maximum output is a very high priority) are flat to 40Hz. For home HiFi, 30Hz is a better target IMO.

- The intended bandwidth of these woofers is the range where room-related issues are at their most noticeable. I would strongly recommend implementing some DSP, even if it is just for these drivers. My current system has a pair of 8" woofers per side in sealed boxes, and I'm flat to 10Hz, thanks to a very helpful room. Without EQ, the bass is actually over-blown: there's a +15dB mountain at 40Hz, and still enough "help" at 10Hz to get 10Hz level with >100Hz. A flat in-room LF response is a wonderful thing.

Maybe you are correct with the 40 Hz, I based the figure on the fact that the horns in my room were flat to this figure, and I never felt lack of bass.

Regarding the EQ, I agree, but this is an implementation issue, that will depend on measured room gain of the prototype. I need to get a handle on building the LF portion first.

Kindest regards,

M
 
Last edited:
Well, yes, power can be cheap,
No. In 2021, power is cheap.

if one did not try to get 150W in class A.
The only thing that makes class A more difficult than AB is cooling and the large heat sinks if you insist on passive cooling. Forced air or water, both of which I have used can be done very cheaply and silently if you put your mind to it.

However, before I started the investigation, I did not even know what power I needed.
As power is cheap an excess is not an issue, especially at LF, so why obsess about calculating it out to a tiny margin of error? In the real world, it's not going to matter all that much.
 
M,

The infinite combinations of Vb and power required to hit 7.9mm of peak excursion at 40Hz are simply points on a curve. Pick a point on the curve to be your Vb, and off you go.


Flux modulation is pretty well instantaneous (at least, compared to the time intervals involved in low-frequency reproduction), so you'll find the resulting harmonic distortion will occur at large signal levels.
In my opinion, a high-quality driver being used within its limits (NB - 7.9mm of excursion will be getting towards the limits of a lot of 15" midbass units, so dedicated subwoofers drivers (Faital 15HP1060, Beyma 15SW1300Nd, etc etc) or multiple drivers should be considered) will manage this just fine, and it's a non-issue unless you're going to be seriously exceeding the driver's thermal power handling.

ie, using a much-too-small box and dropping lots and lots of power in there, but still staying within Xmax, will result in more harmonic distortion than a larger box using less power to hit the same SPL.
In short, use a good driver and a sensibly-sized box. You'll be fine.


No preference on amplification in this frequency range - so long as there's plenty of clean power, class AB or D will do the job.


I also think you might be over-thinking this, but I enjoy working through the process.


Chris
 
Hi Chris661,

thank you once again for your time.

The infinite combinations of Vb and power required to hit 7.9mm of peak excursion at 40Hz are simply points on a curve. Pick a point on the curve to be your Vb, and off you go.

I understand that, I was just responding to your proposition "to size the box so that Xmax is reached when the amplifier reaches maximum output", which appeared to suggest a single value.

Flux modulation is pretty well instantaneous (at least, compared to the time intervals involved in low-frequency reproduction), so you'll find the resulting harmonic distortion will occur at large signal levels.

I did not know that (among other things I do not know); I thought that it was a thermal related issue.

In my opinion, a high-quality driver being used within its limits (NB - 7.9mm of excursion will be getting towards the limits of a lot of 15" midbass units, so dedicated subwoofers drivers (Faital 15HP1060, Beyma 15SW1300Nd, etc etc) or multiple drivers should be considered) will manage this just fine, and it's a non-issue unless you're going to be seriously exceeding the driver's thermal power handling.

ie, using a much-too-small box and dropping lots and lots of power in there, but still staying within Xmax, will result in more harmonic distortion than a larger box using less power to hit the same SPL.
In short, use a good driver and a sensibly-sized box. You'll be fine.

The two paragraphs are little convoluted for me not being a native speaker. Are you suggesting that the TAD with its Maximum rated power of 300W should be O.K., provided that I select Vb to keep the max power under, let us say 50%, of the maximum rated power; furthermore, considering that the peak power will be infrequent?

I also think you might be over-thinking this, but I enjoy working through the process.

You might, of course, be correct, but at the same time a little forethought might prevent costly mistakes.

Kindest regards,

M
 
If your target for f-3dB (?) is 40Hz, which driver would you need, presuming you are not in for active correction? I missed your criteria I guess, but the drivers you mention won’t work as they are deigned for vented (BR) use.

No, you did not miss anything. What I was trying to understand was how much power I need for a target of 110 dB/40Hz. I specifically stated that I am, at this point not concerned with an implementation.

IMHO if your target is F3 at 40Hz then all the alignment in table 2 won't satisfy your requirement. All you have done was to prove that a chosen driver can have a SPL of 110dB at 40Hz with enough voltage applied, irrespective of the max SPL the driver will have at a different frequency. It turns out that the max SPL is at around 116dB, so you have found more or less the F6 point.

I don't know how you calculated the values in table 2, but just as an example I put the T/S parameters of the TAD1601A in Unibox, and found that with a 120L box tuned to 35Hz, the F3 point is around 39.5Hz, and you hit the 110dB at 40Hz with 64W of power.

Ralf
 
Hi Chris661,

thank you once again for your time.

I understand that, I was just responding to your proposition "to size the box so that Xmax is reached when the amplifier reaches maximum output", which appeared to suggest a single value.

You're welcome.

Yes, if you have a particular amplifier you wish to use (or an upper limit to how much power you wish to send to your drivers), then you can often size the cabinet so that Xmax is reached at your chosen power level.

NB: This only works so far. A 1w amplifier will never drive a 1KW-rated driver to Xmax.


I did not know that (among other things I do not know); I thought that it was a thermal related issue.

Power compression (thermal build-up which reduces the efficiency of the motor) also happens, but that's a separate issue.

Flux modulation occurs when currents flow through the voice coil: the permanent magnetic field becomes deformed, which results in some harmonic distortion. More current = more distortion. Of course, there are lots of factors that come into play.

Purifi have put a lot of work into reducing harmonic distortion with their drivers, and their papers are quite interesting, if you wish to read further.


The two paragraphs are little convoluted for me not being a native speaker. Are you suggesting that the TAD with its Maximum rated power of 300W should be O.K., provided that I select Vb to keep the max power under, let us say 50%, of the maximum rated power; furthermore, considering that the peak power will be infrequent?



You might, of course, be correct, but at the same time a little forethought might prevent costly mistakes.

Kindest regards,

M

Apologies. Your English is excellent, but I can see how that paragraph was particularly difficult. You appear to have managed to follow it, though.

Yes, I would say that hitting the TAD with occasional 150w peaks will be fine, and should still be very low in distortion compared to most "HiFi" speakers.

All the best,
Chris
 
Hi Ralf,

thank you for your reply.

IMHO if your target is F3 at 40Hz then all the alignment in table 2 won't satisfy your requirement. All you have done was to prove that a chosen driver can have a SPL of 110dB at 40Hz with enough voltage applied, irrespective of the max SPL the driver will have at a different frequency. It turns out that the max SPL is at around 116dB, so you have found more or less the F6 point.

Could you please re-phase your statement?

My initial post expressed the desire to reach 110 dB peaks at 40 Hz, which according to my interpretation of the data in Table 2 shows as possible.

I don't know how you calculated the values in table 2, . . .

Once I set the power for the specific Vb to achieve the 110 dB at 40 Hz, I copied the values from the fields "Design by Vb and Q" in the "Closed Box" section.

Please, do not hesitate to point any flaw in my thinking, I want to make it right.

I don't know how you calculated the values in table 2, but just as an example I put the T/S parameters of the TAD1601A in Unibox, and found that with a 120L box tuned to 35Hz, the F3 point is around 39.5Hz, and you hit the 110dB at 40Hz with 64W of power.

Thank you for the effort. Per my initial post, I am rather afraid of a vented box because it seems that all of them that I listened to had some frequency that bothered me; in addition, it is my understanding that it is rather difficult to build and tune a vented box.

Saying that, the power savings and associated flux modulation as discussed by Chris661 is attractive. I do not want to push my homework on you, but if it would not be much of a bother to do the calculation for e.g., 140l - 150l, and post the values entered so that I can learn how to do so, I would appreciate it. I could then build the enclosure first as sealed an then perhaps attempt to convert it to vented.

Hi Chris661,

Yes, if you have a particular amplifier you wish to use (or an upper limit to how much power you wish to send to your drivers), then you can often size the cabinet so that Xmax is reached at your chosen power level.

Ah, that is how you proposed to approach it. I do have a 55W amplifier, but I would not take it as a limitation. I can build another one.

Apologies. Your English is excellent, but I can see how that paragraph was particularly difficult. You appear to have managed to follow it, though.

Yes, I would say that hitting the TAD with occasional 150w peaks will be fine, and should still be very low in distortion compared to most "HiFi" speakers.

Please, no apologies; you had been extremely helpful, and it is not your fault that I cannot speak English.

Kindest regards,

M
 
With the TAD and a closed box you will need some active compensation to reach 40Hz flat. Irrespective of the max SPL. I told you your driver won’t do it uncorrected and Ralph put it in another way by suggesting a vented system.

That being said, a Linkwitz Transform or a high Q highpass a la Keele isn’t complicated. Even some cheap DSP will do. But I don’t read you are contemplating such solutions.
 
Hi markbakk,

thank you for the reply.

With the TAD and a closed box you will need some active compensation to reach 40Hz flat. Irrespective of the max SPL. I told you your driver won’t do it uncorrected and Ralph put it in another way by suggesting a vented system.

I believe that we are mis-communicating, but I might be wrong, so let me re-state my though process.

1. My inquiry started by defining the SPL levels I consider necessary for realistic reproduction.
2. I was then trying to establish power needed to reach the SPL level, assuming a sealed box.
3. In response, Ralph appears to suggest, that my conclusions regarding the SPL are incorrect. Since I did not understand his reasoning, I requested an explanation.
4. You seem to agree with Ralph, but, since you did not provide any explanation, I am no wiser.

In my understanding, 1-4 above are conceptual, in that I am trying to understand the need for power, so that the concept can be implemented.

That being said, a Linkwitz Transform or a high Q highpass a la Keele isn’t complicated. Even some cheap DSP will do. But I don’t read you are contemplating such solutions.

Your quote is moving the initial conceptual inquiry to an implementation. I do understand that due to the shape of the SPL as a function of frequency, to obtain a flat SPL response, an equalization is required, be it Linkwitz transform, with which I am familiar, Q high-pass, with which I am not, or any other shaping technique.

Clearly, the Linkwitz transform works by maintaining/extending a flat response by the amplifier providing the speaker with more power at the lower frequencies. ESP - The Linkwitz Transform Circuit However, if one does not have enough power for the equalization to reach the target as per 1, how can the equalization be achieved?

Thus, my understanding is that the power per Table 2 is needed to achieved the equalization.

Does it make sense or did I make it more of a mess?

Kindest regards,

M
 
No, now it makes sense. Be it that the TAD will deliver your 110dB at 40Hz with much less electric power in a vented enclosure than in a closed one. Without Linkwitz transform or Keele-like B4 tuning. Plus it easily will do more than 110dB. Put otherwise, chances are it will do these levels with less distortion in the vented enclosure. If that would matter to you, closed systems require drivers sporting very linear motor and suspension plus massive amounts of Xmax. Haven’t seen Klippel tests of the TAD, but it might be suboptimal for that.
 
Hi markbakk,

No, now it makes sense.

Thank you again for your reply, confirming that my thought process is not completely flawed.

Be it that the TAD will deliver your 110dB at 40Hz with much less electric power in a vented enclosure than in a closed one. Without Linkwitz transform or Keele-like B4 tuning. Plus it easily will do more than 110dB. Put otherwise, chances are it will do these levels with less distortion in the vented enclosure. If that would matter to you, closed systems require drivers sporting very linear motor and suspension plus massive amounts of Xmax. Haven’t seen Klippel tests of the TAD, but it might be suboptimal for that.

Yes, that was what Ralph suggested. Perhaps I need to overcome my paranoia about the vented alignment, find a tutorial for Unibox, and model such an alignment.

Kindest regards,

M
 
I listen to classical. I've played classical, in a 120 piece band for 6 years , a 70 piece orchestra (all district) a couple of hours, (The HS orchestra was out of tune and excruciating to a pitch trained bassoonist). I was in the TXmusedassn All State Band 1968 for about 10 hours. While the band could hit 110 db, & All State band hit 120+ with antiphonal UTex brass on the balcony, I don't find it necessary in the home to enjoy the experience. About 90 is fine. I'm not deaf, I have hearing to 14 khz.
I furthermore find 54 hz low enough for most reproduction, excepting the last note of Also Sprach Zarathrustra, which I heard live once @ Ky Center for the Arts. I play a lot of organ albums, and the 16' pipe rank is enjoyable enough even without the 32' rank. Most LP's after 1958 had the lowest octave suppressed by the recording engineer, to avoid the market killing skips of the tonearm on the extremely popular "fine wood consoles" in most professional's homes of the sixties. (The Colombia LP's from 1954-59 had the lowest octave). If I wanted to listen to 32 hz, an 18" subwoofer as built by Allen organ driven by a 3rd channel would be the ticket. Unfortunately PA 18" woofers are built for the rock & roll market, and are detuned below 40 hz. Electronic crossovers to split off the extreme bass and mix it to mono are cheap: I bought a used Rane CK-225t for $50 last spring.
If you have a passion to build, okay, but the troels suggested above will set you back $$$$. By contrast, we in the USA have another alternative. Used bar band gear. Go to the Peavey dealer with a piano CD, say Appassionnatta Sonata, and audit some SP2's. $600 new. I managed to score a pair of 2004 SP2 for $400 this spring, combined with the Rane. Could have got a class D stereo amp for $150 but personally, I don't want class D. I have a CS800s class AB amp I picked up for $160 with freight. .02% HD @ 240 w/ch 8 ohms. Worked but I put new PS e-caps in it for reliability since it is 20 years old. I looked for the SP2(2004) or SP2-XT speakers under 250 miles for 9 months on craigslist+ebay, after my SP2-XT were stolen 9/20.
As far as the 2004 spec SP2, HD 2nd harmonic rated 20 db down from 5 w 1 m from 100 hz to 10 khz, 3rd harmonic 30 db down 80 hz to 8 khz. That is on the datasheet. What other speaker specs HD? BTW, I run my rig 1/8 w base level in the music room, 1 v pp into 8 ohms, leaving room for 72 db peaks on 1812 Overture from a 70 watt amp. I thought a 70 W amp was fine with 102 db speakers on classical in my 14'x11'x30' music room, but mine was home built, popped a wire off, and started humming. On to CS800s which is repaired or M-2600 which is also working (but a little over on HD due to TL072 op amp).
SP2(2004) speakers are vented, 15" 1505-8kadt woofer, with a volume of 123 L if they are made of 3/4" plywood. 1505-8kadt is spl 99, Qts .35 VAS 143 L Xmax .99 mm Overall speaker response is 98 db 1w1m, the SP2-XT were 102 db 1w1m.
If you want to build, build the 32 hz subwoofer. I don't see many of those on the used market. I've got a line on an Allen 18" abandoned in a church I support with pipe organ installation help, but that is a one time deal. As I say, used PA woofers like Peavey SP118 don't go low enough.
I initially started a project to build replacement copies of the stolen SP2-XT which crossed to the 2" CD horn @ 1200 hz. Woofers were going to be Eminence Delta-pro-15 which are spl 101.8 db Qts .40 Vas 243 L Xmax 4.3 mm, which I bought already. That project stopped for summer mowing-tree-cutting-tractor-maintenance season, after I found the SP2-2004 in the spring. I have the woofers & the used Peavey RX22 drivers with used plastic horns. Next issue is to lift the SP2 onto the poles with tweeters 8' above the floor where they belong. They are too heavy for me, 84 lb.
There are other shopping goals, like the JBL4722 which does go to 30 hz. Never came up on cl/ebay within 1000 miles of here.
If you go shopping on the used market, take a CD player an amp & some cables. Many PA speakers are damaged and get repaired with inferior drivers. Mine have been low use, and dead accurate reproducers of piano even in a 5 sided trailer LR that the first pair were located in. The 2nd owner was paranoid about allowing my CD player to touch his amp, so I used my own MMA-875t (mono). Amps get blown often in bar bands by people plugging guitar amp 75 watt output into the amp inputs.
I'd say a little listening instead of cutting wood is in order. Not everything on the market is smoke & mirrors. If you're under 500 miles, I may get my system operational next month or two. It is fall, time for electronics, piano & organ practice, pipe organ installation.
 
Last edited:
Best advice we could give here: build a proven design or buy a new or second hand main studio monitor I guess. But getting there is more than half the fun, so let's not discourage the TS here.

Next best advice: try before you buy. Or rather, build a reasonably priced design like the Asathor, documented here and thread here. And check if your goal really is what you want. Keep it for reference, trying to get your design better than that. And be surprised for that will be a challenge.