So for reasons that defy logical explanation, the wife and I would up owning a 130 year old, gothic revival church. This would normally have little to do with a 20Hz horn loaded sub, except that we are toying with the idea of living there.
The church itself is about 60 feet long and 20 feet wide, with transepts off to the sides of the altar that are about 6 feet deep and 11 feet wide.
One of the big challenges is that the celing height is over 30 feet at the peak. This is one huge volume of a room. And also an impressivbe one at that!
The transept that the "workbench" AKA dual horn loaded sub goes into is off to the left in this pic.
The sheer volume of the place sets a challenge in providing solid bass. At the same time this provides a few opportunities to build a speaker like I will never have the chance to do again.
The bulk of the hall is a pretty massive living room, and when at the church, my hobby desk is in one of the transepts. The Wife has given permission for me to adapt this as a "workbench". Yes, she knows there is a cunning plan to make this a giant sub.
I toyed with tapped horns, vented, sealed and infinte baffle. In the end a combination of the practicalities of generating adequate acoustic power for the space and what drivers I had laying around led to a horn loaded sub. The overwhelming factor here being the efficiency provided by a horn loaded sub.
Tapped horns do OK, but dont really help that much with cone excursion. Sealed, bented and infinite baffle work, but would reqire me to buy new drivers - and in such a room would drive me to big expensive drivers and big power mps if I were to achieve bass antwhere toward 20Hz.
So a horn it is. But I need to get a horn mouth that is ludicrously big to ger anything like decent bass toward 20Hz. So how does this work?
The transept is shown below...
Yes, that is a bed on the old altar
🙂
The horn loaded sub will actually be two in mirrir image, operating either side of a workbench. The air of these will effectively turn the 1/2Pi space against the wall into a 1/4 PI space. This reduces the total mouth area required by the horn to "work" as a 20Hz horn.
This helps, but we still need a monsterous horn mouth. So what gives?
There will be a 900mm (three foot) wide section as a desk, with 1200mm (four foot) wide subs either side. The two of these will each see what is effectively 1/8 space. They will also face into the transept. This is pretty important in this design.
View attachment Horns_In_Hall.pdf
At the back of the desk there will be a panel that closes off the desk - making the area behind the desk and into the transept an integral part of the horn. After exit into the transept, the area of the horn section is effectively the area behind the horn to the rear wall of the transept. This is 330cm wide by 70Cm deep, or 23,100cm square for the pair of horns, or 11,500cm2 each. This part of the expansion is controlled by the placement of the desk, and is not "made up" by me, and is not an "optimistic guess".
This will then expand out to the width of the transept and exit into the hall at the face of the wall. This part is a bit more of an optimistic guess... but I think will realistically expand to twice the 23,000cm2 at the exge of the trsnsept.
Each of the wooden (folded) horns is far from modest. They are 1036mm deep, 1200mm wide and 800mm tall. The desk will be between these and slightly below the height of the top of the subs ) as the sub exterior height is 836mm, somewhat taller than is comfortable for a desk.
The horns themselves are characterised by:
- Drivers: Two 25 inch Richrd Allan HD15 drivers per horn
- S1 (at the horn mouth) 700cm2
- L12 280cm
- S2 (at the exit of the folded horn into the transept) 6000cm2
- Vrc (behind the drivers) 180 litres
- This is a 20Hz flare rate from the drivers to the horn mouth. The upper frequency is set by the entry from the drivers into the horn, a touch over 200Hz.
In the transept we have
- L23 - about 100cm (known and set by placement of bench)
- S3 - 11,500cm2 (area of transept behind the bench)
Then from the exit of the overall horn coupled into the transept
- L34 - the length of the transept to the hall
- S34 - 36,000 this is a bit of a guess. Admitted.
At the back of the desk, there will be a further extension of the horn into the depth of the transept, which extends the horn loaded section to the rear of the transept.
From the expansion to the back of the transept, there will be a further expansion back into the hall proper.
Lets look at the main "horns"...
View attachment Top View.pdf
The main horns are constructed from 18mm "structafloor" - basically flooring board. This is both tough, rigid and cost effective. These subs use a LOT of timber. And bracing.
The sub internal height is 800mm. All panels except the top and bottom are 800mm tall. This allowed me to get the local hardware store to rip the top and bottom off 3600mm "yellow tongue" flooring sheets to make the panels. Mind you there are 5 3600mm sheets and two 2400x1200mm sheets in these boxes. Just the timber will be way over 1/4 ton before I start!
With the drivers there is no way these things are moving around.
View attachment Top Panel Drill.pdf
I have to cut and drill all the panels before I take them into the hall. Not because I cant get them in the door (there are double doors in the belltower) but because of the weight of the things.
CAD is super helpful here.
What are drivers? The model is pretty insensitive to the drivers. I gave four old school Richard Allan HD15 drivers. Two per sub (there are reasons why I did this....) works a treat. Dropping in for example an JBL2226 gives better results, if the difference actually matters.
Looing at the response using the Richard Allan drivers... with no effort to make the plot look pretty...
This includes the path length from the mouth of the horn through to the transept and up into the hall.
And this is for 2.83V into 8 Ohms. Which is 2W per box, or four watts total input. Just goes to show the power of horn loading
🙂
So what is the cone excursion like at 115dB?
This says that at 2.83V input, and close on 115dB output right down close to 20Hz, the cone excursion peaks at 1.1mm. Even witht he old school Richard Allan drivers, which I would rate to +/- 4mm or so in a "near engh to linear" region, there is miles of excursion to go. 50 watts per driver and things will be getting rather loud in the hall...
To be honest, I don't see these ever getting turned up too far, as I am worried that the old stained glass windows will vibrate too much and break. I would hate myself, and the wife would probably cut my head off and hide it inside one of the speakers. No really, she would be that angry... I would go as far as saying "without a doubt, I will go missing if I break the windows"
The timber is bought, and cut. I am currently finishing off the drilling and prep of the panels to take up to the church and screw together. In the next week or so they will be up and running.
They will be run via a DSP crossover which will allow me to program the delay required in the main speakers. The amp driving them will likely only be 50 Watts into 8 ohms - if I leave it like that I will not be tempted to crank them up
🙂
The main speakers are quite high efficiency. To match the old school Richard Allan bass drivers, to match the subs, I dug out some 10 inch Richard Allan CG10's. These are about 97dB/w/m. The tweeters, high efficiency domes are running with nil attenuation to match the efficiency of the 10 inch bass mids.
This will end up being a seriously high efficiency system, and I have litle doubt, cable of some awe inspring bass in such a huge volume of a room.