RT-2 Horn sub, the result

[Elbert]

Regarding driver Fs:


Bass drivers for direct radiator applications typically have a low fs (lower than fb), medium Qts and fairly high BL. Suitable bass horn drivers, by contrast, fs is higher than fc, Qts lower than 0.35 and higher BL.

ref. the WIKI section.

According to the above, the Eminence Delta-pro 12A driver I chose should be appropriate, or at least so i thought...

[bjorno]

Hi Elbert,
In the late seventies, I made a couple of the RT-2 horns using the recommended and other different drivers.

They all had typical horn mid-bass FR characteristics but not enough BW to call the horn a sub that IMO should at least cover 1.5 octave below 80 Hz.

RT-2 is more useful if complemented with satellites FR starting from about 160 Hz-.

Hint:Try to model your RT-2 horn using the given Hornresp template that can be found at this link:

http://www.kensonpro.com/linnaraudio...simulering.htm

b

[technofreak]

Originally posted by Elbert
Regarding driver Fs:


Bass drivers for direct radiator applications typically have a low fs (lower than fb), medium Qts and fairly high BL. Suitable bass horn drivers, by contrast, fs is higher than fc, Qts lower than 0.35 and higher BL.

ref. the WIKI section.

According to the above, the Eminence Delta-pro 12A driver I chose should be appropriate, or at least so i thought...

Delta Pro 12 is a midrange/midbass driver, very good from around 80 to 2000 Hz in a BR. It is regarded as one of the best midrange drivers from Eminence, along with Beta 8 and Alpha 6.

As for a driver's suitability for horn loading, Wikipedia is right - this is a valid approach for midrange horns and many PA bass cabs, which does not need to go below 50 Hz.

However, very good bass horn can be designed around more "typical" driver, too, for example, LABhorn, that goes well into 30-s. The driver used in it, Eminence LAB 12, has T/S parameters more close to home or car sub than a typical PA sub - indeed, I've found a fairly cheap 12" car sub that has parameters close to LAB12 and simulations in Hornresp show, that it, in fact, can be used in LABhorn.

The driver is Signat Fury 12 and costs less than 100€ here, in Estonia (LAB12 is over 250€ here). Only worrying thing about this driver is it's claimed 1" voice coil - seems too small for such driver.

Bottom line is, I would certainly try some other driver with lower FS and maybe higher Xmax.

[Elbert]

Fantastic bjorno!

I did have a go at some hornresp simulations when trying to select a driver, but it is obvious that the model in the link you refered to is way better than what I managed!

Now I can experiment with some different driver parameters to better get an understanding of whether or not my dissapointment is caused by sub-optimum driver selection or inherent limitations in the horn it self!

I could for instance do a simulation using the parameter of the SEAS 13" I have, and then try it out in practice to see if any simulated changes will materialise in real life.

Can't wait to get home and try this out!

[AndrewT]

Hi,
it's obvious looking at the box pic it can never operate as a sub-bass horn. The view of the inside of the box shows it will not even operate as a bass horn.

Three factors determine the LF response of a horn.
Mouth area.
Flare rate.
Path length.

All three factors must be satisfied or come very close to achieve the target response. One can truncate the horn before reaching the desired Mouth area for the loading situation. That is the first compromise.
One can cut off the horn short. That's the second compromise.
One can flare the horn too quickly. That's the third compromise.

There are many other factors to take account of, but these just guild the lily once the basic size is determined.

It looks like all three LF response factors have not just been curtailed, they have been slashed.
Compare the mouth area of 6LabHorns to this box opening.
Compare the path length in a LabHorn to this box.
Compare the Flare rate to that in the LabHorn.
This box was doomed to failure before the designer even sat down to sketch it out.
It is simply too small.

It should make an upper Bass horn or possibly a Mid horn, but too many bends/folds.

Addendum.
one needs 12Labhorns in free space to achieve the 27Hz cutoff.
6LabHorns are needed for floor loading to give the desired Mouth area. That is ~4sqm (~40sqft) for 27Hz in 2Pi loading.

[Elbert]

Andrew,

You are in deed correct in your observations!

These issues were allso adressed in the original article series (in swedish), and the approaches explained and justified.

What my pictures doesn't really show is that the space between the floor of the horn and the flor on which it will stand, is in fact the final section of the horn together with the triangular foot. This makes the horn somewhat longer than the internal horn foldings.

The horn mouth is formed by the slot between the horn flor and the floor. Admittedly, this is does not amount to a huge area, but the theory behind this was that with the elongated mouth this resulted in, the effective mouth area would be doubled due to reflection from the floor or something.

But yes, you are right, even with all the compromises and trics in the world, this can never be a true deep bass sub, you can only bend the rules of physics so much!

Nevertheless, Since I've gone through the trouble of building it, I can just as well play around with it.

If I can conjure out as much as 1/2 octave further down in response compared with the initial results, it might yet prove sufficient for engaging listeing experiences!

[Elbert]

Wow jbell!

If you can get the sort of performance you described out of those modest (size-wise) horns, There should be hope for my build as well!

Sounds like excellent value for money too, must admitt I've spent somewhat mopre on my build so far.. which is probably part of the reason I'm a bit hesitant to burn my project! (not to mention the risk of asphyxiating from the fumes given off by all that burning filler and glue!)

Can you describe the other systems you use or point me to some reference? JUst so I can better understand what you are comparing with.

[jbell]

billfitzmaurice.com

The T18 is a 7' (or there abouts) length horn. Not sure what your hornresp params are of your horn, but if you are going to get 40hz, you need at least that length. I found that adding an 'extra wrap' to the T18 gave me great response down to 35hz, made the cabinet more efficient overall (I could get to 126 with less power) but once I hit 126, that was it, more power didn't do much... If you want real low, the fitzmaurice table tuba (13' length 22hz horn) is better suited. It's also an open rear chamber front loaded horn, so in some respects may be thought of like a tapped horn, scoop, etc..

I have issues with the response characteristics of undersized 1/4 wave length horns, like what BF has on his site. However I've found that for the table tuba and the T18, that a series inductor can flatten them out enough to be usable, especially in dual driver cabinets.

A mcm 55-2421 is a 4ohm driver, and the cabinet adds 2 ohm in horn inductance, so each driver is considered 6 ohm. A pair of drivers in parallel, is 3 ohm, so every 4ohm stable amp I've tried has driven it with no issues, partially because you never run more than a couple watts of power. The 3 ohm load also makes it easy to get an inductor that will work well. (12mh is usually enough)

one last, 120hz eq at -6, will not help as much as you think. The inductor is -0 at 40hz, -6@80hz, -12@160... (6db/oct) in addition to the -12db/oct crossover, which makes 160 @ -24.

However, if you check out any of the tapped horn threads, or play around with hornresp, you'll see that added inductance 'changes' the response of the horn in more ways that just 6db/oct crossover. This is true for front loaded horns, as well as tapped horns.

hope this helps

[Elbert]

Thanks for all the details jbell!

Now, finally the weekend is here and I'm back in the den!

Just hooked up the old nikko parametric EQ and set the sliders thus:

126 Hz -12 dB
63 Hz - 3 dB
31 Hz +2 dB

That certainly took care of some low mid-bass whoomp, it sounds much drier and neutral now..

But those lowest notes are still missing.

It's particularily noticeable when I play a certain track on a jamariquai (did I spel that right?) record I have where there's a synth sweep that runs straight down in to the basement.

Where as my old BR sub at least follows the sweep down, admittedly dropping in level, the horn still just runs out of breath.

Its the same when play a lot of other albums, just seems like the foundation just isn't quite there..

I haven't looked in to what an inductor reactance will do with the electroacoustics as you pointed out, but I'll see if I can play around
with hornresp a bit later on..

So, the next step will be to borrow a SEAS W33 woofer from one of my main speakers and se how that fares in the horn. (if I can make it fit inside)

To be continued..

[AndrewT]

Originally posted by Elbert
................
126 Hz -12 dB
63 Hz - 3 dB
31 Hz +2 dB............................
.............
Where as my old BR sub at least follows the sweep down, admittedly dropping in level, the horn still just runs out of breath.

have you looked at the driver cone excursion when playing this torture track?
When a horn unloads (stops working as a horn) the driver behaves as if it had no horn at all. If a back box is fitted (your's has) then that becomes the only load the driver sees.
Is it a sealed box?
You can model it and see what you expect in low end response. Expect interesting results.

If the back box leaks badly, expect your driver to give up the ghost after it has run out of breath.