Collaborative Tapped horn project

[MaVo]

I wonder how i can determine, if i should simulate horns in 2*PI, 1*PI or 0,5*Pi. I have attached a sketch of my audio room, with intended placement of the subs. This scheme seems to be the best to supress room modes, which is why i use it. Right now, there are small closed box woofers, which i want to replace with tapped horns.

Should i simulate this setup as one horn with 4 woofers in 0,5*PI or is it better to assume the worst case which would be the response plot of a single horn in 2*PI?

The room has only one door and a very little window. Since its in a basement, two walls are of "infinite" thickness and the others are concrete.

Can someone help me with this?

[GM]

Don't know how you accurately sim it due to the complex summing involved, but I would just sim a single corner and since each doubling adds 3 dB, just add 6 dB to whatever you come up with.

GM

[hm]

Hello Mavo,

thats a joke or overkill,
for such a small room,
i would use bearly one TH,
4 ? where you will sit and what distance
to the TH, 0,5 m ?,

du dokumentierst damit ziemlich viel Unkenntnis, schade.

[Patrick Bateman]

Originally posted by MaVo
I wonder how i can determine, if i should simulate horns in 2*PI, 1*PI or 0,5*Pi. I have attached a sketch of my audio room, with intended placement of the subs. This scheme seems to be the best to supress room modes, which is why i use it. Right now, there are small closed box woofers, which i want to replace with tapped horns.

Should i simulate this setup as one horn with 4 woofers in 0,5*PI or is it better to assume the worst case which would be the response plot of a single horn in 2*PI?

The room has only one door and a very little window. Since its in a basement, two walls are of "infinite" thickness and the others are concrete.

Can someone help me with this?

It might be worthwhile to stack two of the tapped horns on top of each other, with the top one inverted. Then measure the response in-room.

Distributing the subs flattens the response as you know, but there's nothing to stop you from distributing them *vertically* too. Stacking them with the mouth near the ceiling accomplishes that.

[MaVo]

Hi GM, you saw the problem Thanks for the hint.

Hi Patrick Bateman, your hint is also well reveived. Indeed this would even more smooth the frequency response, but only affect the vertical modes, between ceiling and floor. The lowest of those is in the 70s, so this isnt my first concern.

For those of interest, i can only recommend the harman papers on room response, especially the multiple sub paper.

[Chris8sirhC]

What on earth are you using for your mains that could keep up with that setup, and which design did you end up going with?

[David McBean]

Originally posted by GM

Greets!

OK, true, but it's normally derived also since it's very rarely published, so still not sure why you prefer it over the ~universally used Mms.

Understood, thanks, but I was hoping you'd publish the formula. Oh well, doing a bit of Googling yielded this from Bill Geiger:

[Mmd] = [Mms]-2*([Sd]^2)*[Ma1]

[Ma1] = 8*[p0]/(3*[pi]^2*[a]) - Acoustic Mass of Air Load on one side of equivalent piston (kg/m^4)

[p0] = 1.18 kg/m^3 - Air Density

[pi] = 3.14159

[a] - Driver Effective Radiating Piston Radius (m)

[Sd] = [pi]*[a]^2 - Driver Effective Radiating Area (m^2)

GM

Hi GM,

Bill’s expression for the mechanical mass of the air load on both sides of the diaphragm of a driver mounted in an infinite baffle 2*([Sd]^2)*[Ma1] is taken from Leo Beranek’s book “Acoustics”. It is an empirical formula based on the observation of results, not one derived from fundamental theory. What the formula actually says is that the loading mass on each side of the diaphragm can be considered to be a layer of air equal in area to that of the diaphragm equivalent piston, and equal in thickness to 8/(3*pi) or ~0.85 times the piston radius. This is obviously a rather simplistic approach to take. You will notice that the approximation formula is independent of frequency. This means that the value becomes less accurate as the resonance frequency of the driver rises.

The value for Mms in Bill’s formula [Mmd] = [Mms]-2*([Sd]^2)*[Ma1] assumes that the driver is mounted in an infinite baffle. In many cases the manufacturer’s specification for Mms is actually based on an unmounted driver radiating into free space. This is often not made clear in the information provided.

Hornresp uses a more accurate and theoretically correct method by calculating the actual acoustical reactance of the air load on each side of the diaphragm, with frequency being taken into account. Without wishing to get too technical, the calculations require the use of the Struve function, the handling of which is not a trivial exercise. Unfortunately the precise result is not given by way of a single simple expression, which is why I did not include a formula in my previous post.

The reason I favour using Mmd rather than Mms is because Mmd is an absolute physical value that can be directly and very accurately measured. Mms on the other hand can only be determined indirectly. Mms values provided by driver manufacturers can sometimes be rather inaccurate, or it may be unclear whether the figure applies to a driver in an infinite baffle or to an unbaffled driver radiating into free space. Hence my preference to use Mmd where specified, or to calculate Mmd from Sd, Cms and fs (the driver free-air resonance frequency) when Mmd is not available.

Hope this helps to explain why Hornresp is the way that it is :-). In principle, I try to make the results calculated by the program as accurate as possible.

Kind regards,

David

[MaVo]

Originally posted by Chris8sirhC
What on earth are you using for your mains that could keep up with that setup, and which design did you end up going with?

If this question was directed at me, the answer is a pair of horns, which i designed according to tom danleys unity/synergy horns. But i dont play very loud, i just think that having lots of headroom and efficiency is a nice thing for sound quality and nerd-like satisfaction.

[Chris8sirhC]

yeah, it was directed at you. Which tapped horn did you end up building? With the drivers you have, there was a design that was good to 18hz or so, and then one that was good down to 14-15hz.

[MaVo]

i see, i want to use the bms 18n600 as soon as its released, since i figured i should use the best driver i can find or else i will upgrade too soon - kinda a psychological thing i suppose. i will use four subs for room mode cancelling. i am not interested in high volumes, but in clean and effortless sound. ill build some little monitor boxes to give away from my 12 inch drivers i have right now, since they are more of a midrange than a woofer - trying to clone a summa should be fun

parameters are: