Hey folks! 
Long story short: I currently have access to a pair of Coral Beta 8's through my father. I hope to use these in a DIY enclosure which is based off of the original BLH plans, but taller and narrower.
My question is; Would the acoustic properties of a modified enclosure be the same as the original so long as the internal volume stays consistent? From what I've understood the throat size is a critical factor, and in my dumb head i figure that so long as the decrease in total width is compensated for with a proportional increase in depth or height it should turn out all right, no?
Springing for anything else is out of the question since I'm a student and means are tight enough as is -- So I'd really like to make the most of what i have access to. Ultimately the Beta 8's are far and away my best bet unless I want to try and steal my fathers' Quad ESL-2905's without him noticing, which is highly unlikely.
Bonus question:
I also have a pair of Gamma VLD-12's that i could theoretically use as tweeters/supers. Could i enlarge the throat size (or general horn size) of the Beta 8 enclosure to extend the bass response at the cost of HF response so i could two-way it up with the Gammas? Or am i just asking for trouble?
Thankful for any and all advice, so long as it isn't "Give up".
Cheers!
Long story short: I currently have access to a pair of Coral Beta 8's through my father. I hope to use these in a DIY enclosure which is based off of the original BLH plans, but taller and narrower.
My question is; Would the acoustic properties of a modified enclosure be the same as the original so long as the internal volume stays consistent? From what I've understood the throat size is a critical factor, and in my dumb head i figure that so long as the decrease in total width is compensated for with a proportional increase in depth or height it should turn out all right, no?
Springing for anything else is out of the question since I'm a student and means are tight enough as is -- So I'd really like to make the most of what i have access to. Ultimately the Beta 8's are far and away my best bet unless I want to try and steal my fathers' Quad ESL-2905's without him noticing, which is highly unlikely.
Bonus question:
I also have a pair of Gamma VLD-12's that i could theoretically use as tweeters/supers. Could i enlarge the throat size (or general horn size) of the Beta 8 enclosure to extend the bass response at the cost of HF response so i could two-way it up with the Gammas? Or am i just asking for trouble?
Thankful for any and all advice, so long as it isn't "Give up".
Cheers!
Right, IF the various volumes, path-lengths, expansions are the same you can in theory reshape it however you want.
You can truncate the horn down to the desired larger throat area to reduce its useful (mid) bass BW, i.e. driver goes lower before horn takes over, but if it doesn't line up with the driver's specs you might wind up with something you can't EQ/tweak to a seamless 'hand-off' between horn, driver outputs.
If wanting the horn to go lower, then a new larger design is required.
You can truncate the horn down to the desired larger throat area to reduce its useful (mid) bass BW, i.e. driver goes lower before horn takes over, but if it doesn't line up with the driver's specs you might wind up with something you can't EQ/tweak to a seamless 'hand-off' between horn, driver outputs.
If wanting the horn to go lower, then a new larger design is required.
Ah, makes sense. So theoretically if i were to "unfold" the middle piece to make it a taller, Hedlund horn-style enclosure it'd be fine so long as it's all according to spec.
I went ahead and modeled the original enclosure in fusion 360 to be able to calculate the general parameters of the horn, and so far I can't really make sense of them. It has a throat diameter of ~12cm, a path length of ~1.728m and a final diameter of ~37cm. The coupling chamber is also quite small. Intuitively it seems fairly underdimensioned.
Is it common to have a fairly high cutoff point? I realize that dimensioning the horn for a cutoff of 37Hz isn't realistic, but with these dimensions i figure the cutoff ends up in the ~200Hz range. Or does the "extending" effect of the room itself really do that much for the low end response?
I'm currently reading through as much literature pertaining to the subject as possible, but any help is much appreciated.
I went ahead and modeled the original enclosure in fusion 360 to be able to calculate the general parameters of the horn, and so far I can't really make sense of them. It has a throat diameter of ~12cm, a path length of ~1.728m and a final diameter of ~37cm. The coupling chamber is also quite small. Intuitively it seems fairly underdimensioned.
Is it common to have a fairly high cutoff point? I realize that dimensioning the horn for a cutoff of 37Hz isn't realistic, but with these dimensions i figure the cutoff ends up in the ~200Hz range. Or does the "extending" effect of the room itself really do that much for the low end response?
I'm currently reading through as much literature pertaining to the subject as possible, but any help is much appreciated.
If your 200 Hz estimate comes from the the speed of sound divided by horn length (343/1.73), then you would need to divide the frequency by four because the pipe is a quarter wave resonator. (343/1.73/4) ~= 50 Hz. This is a rough estimation, the taper may affect the frequency also. But it is still a reasonable estimation. The woofer chamber can be smaller than you expect because the pressure wave in the chamber is partially summed with the pressure in the horn. This makes the chamber effectively larger. Tapered quarter wave pipes are easy to approximate but trying to calculate an exact value is much more difficult. (and unnecessary, in my opinion)
Yes and no. It's a horn -but it's a heavily compromised horn (like most of this type are) in terms of its expansion (aka bulk). In other words, there is a mismatch between the impedance-matched frequency determined by the expansion & terminus area, and the QW cutoff determined by the length & taper. Basically, it's a standing-wave generator from its QW cutoff frequency up to the point that it's actually impedance matched; most of the time with designs of this kind, the latter is actually a fairly narrow BW before you reach the upper corner frequency and the acoustic low-pass, although you have to factor the floor reflection boundary condition into this: at the very least, in practical conditions this design is in 1/2 space rather than free-space so the virtual terminus area is doubled. In theory. Whether it quite functions like that depends on the solidity & any covering that might be present on that floor though. If it's rammed into corners, you may pick up some more also, but how much is pushing the boundaries of what could be called a SWAG into a WAG. 😉Is it common to have a fairly high cutoff point? I realize that dimensioning the horn for a cutoff of 37Hz isn't realistic, but with these dimensions i figure the cutoff ends up in the ~200Hz range. Or does the "extending" effect of the room itself really do that much for the low end response?
I'm currently reading through as much literature pertaining to the subject as possible, but any help is much appreciated.
Correct.
Yes since the box only loads the driver to its upper mass corner (Fhm) where T/S theory peters out, so for a given Fs: the lower its Qt', the higher its Fhm = the smaller its throat/chamber and vice versa:
Fhm = 2*Fs/Qts'
Fs: Fhm*Qts'/2
Qts': 2*Fs/Fhm
(Qts'): (Qts) + any added series resistance (Rs)
Have yet to read it, but presume MJK's BLH doc is the most technically correct one to date, though haven't compared it to my adaptation of Keele's '77 FLH doc to see if there's any potentially audible difference.
All that said, there's the ongoing fallacy that a horn must have a very powerful motor (low Qt) when nothing could be further from the truth as stated since its desired Fhm, (Rs) governs Qts' and its inductance for its transient response.
This is easily shown by simming a driver in an infinite baffle (IB) and as one's speaker alignment 'slides' backwards down the slope to Fs/Fb its bandwidth (BW) increasingly narrows up with decreasing frequency, ergo its Fhm = increasing Qts'!
......or in my case, mostly wishful thinking since my listening room is a stick built 'floating' floor foundation construction to withstand/absorb the high winds/actual mini twisters that can/has slammed into it over the decades.
Just messed around with this calculator to get in the ballpark. But that makes sense! So in practice a horn tuned for ~40Hz wouldn't be that much larger, then? I can do a fairly tall speaker, roundabouts 150cm.
I am determined to understand whatever you just said... eventually. I'm familiar with standing waves from transmission line theory in electrical engineering, but I'm not entirely wise as to how they work with regard to horn loaded speakers. The taper is also a mysterious thing.
Yes, I've been perusing the MJK doc. Also been looking through his other website, and this forum, for more info. Ultimately my main barrier to understanding is the jargon. Lot of F's, Q's and V's to remember!
Really appreciate the response so far. Has helped a great deal in clearing things up. At this point i'm wondering if I should just build the original enclosure and be happy with it, but I'm very much the overdoing it-type. If there's performance gains to be made with a few adjustments I'm happy digging just that bit deeper.
You're welcome!
Don't recall reverse engineering this particular BLH design, but historically, any pre/early T/S design includes some amp added output impedance (series resistance) that increases Qts (Qts'), which in turn makes the entire horn design under-damped, though of course one can use a cheap pot* to dial it in in-room and replace with an adjustable $$ power resistor or a grid of small resistors.
The other major issue that Scott noted, room boundary loading, often just accounted for any horn, floor axial/path-length truncating, so backed up near/against a wall or in a corner ideally needs to be accounted for a seamless 'hand off'.
* FWIW, I did this with all speaker systems since the room dominates down low and most times each speaker needs different EQ, tunings.
Don't recall reverse engineering this particular BLH design, but historically, any pre/early T/S design includes some amp added output impedance (series resistance) that increases Qts (Qts'), which in turn makes the entire horn design under-damped, though of course one can use a cheap pot* to dial it in in-room and replace with an adjustable $$ power resistor or a grid of small resistors.
The other major issue that Scott noted, room boundary loading, often just accounted for any horn, floor axial/path-length truncating, so backed up near/against a wall or in a corner ideally needs to be accounted for a seamless 'hand off'.
* FWIW, I did this with all speaker systems since the room dominates down low and most times each speaker needs different EQ, tunings.
Right -- It's going to be stood fairly close to a wall, so that's good to know.
According to this webpage the Beta 8's have a fairly low Qts (0.285). Does this then mean that the enclosure might be under-damped to compensate for the relatively low Qts of the driver? From what I've gathered an ultimate Q of ~0.7 is ideal.
To make sure I've understood you guys correctly:
And when you say this would "reduce it's useful (mid) bass BW" you're merely stating that the resonant peak would be shifted more towards sub bass territory?
According to this webpage the Beta 8's have a fairly low Qts (0.285). Does this then mean that the enclosure might be under-damped to compensate for the relatively low Qts of the driver? From what I've gathered an ultimate Q of ~0.7 is ideal.
To make sure I've understood you guys correctly:
By this you mean that I could shorten the path length (thus proportionally increasing the throat size) to lower the effective Fs?
And when you say this would "reduce it's useful (mid) bass BW" you're merely stating that the resonant peak would be shifted more towards sub bass territory?
And by this you mean that in reducing the Fs i might unintentionally introduce a "trough" in the FR of the speaker where neither horn nor driver can effectively produce a certain frequency?
Is there some calculator i could use to mess around with these parameters? I know of AKABAK and Hornresp, but both of them have rather severe learning curves by the looks of it. I would like to be able to enter the driver parameters, throw various path lengths, exponential curves, coupling chamber sizes and throat diameters at it, and recieve and approximated frequency response. No room correction or anything, just something that allows me to visualize these parameters interacting with eachother so i can get a better feel for it.
Additionally -- I realize that I'm now asking things that are past the scope of my initial question, so if there's some other thread or easily digestible reference document I'd be better off reading, let me know.
Sorry. 😉 To try to simplify: take a mathematically 'ideal' horn. The type doesn't matter: just imagine that you've got your desired horn expansion formula & used it to design a horn with no changes, that's tuned to, say, 40Hz just to pull a number out of the air. We'll also assume free-space conditions, i.e. the horn is floating a couple of thousand feet up in the air, so there are no boundaries nearby to consider. That horn will be, since you used the mathematically ideal expansion calculations, impedance matched down to its tuning frequency (caveat -there may be some end-correction to deal with, but we'll leave that aside).
This is all fine until the elephant in the room comes along: that ideal horn is by most current domestic standards, huge. You can reduce the size of it in practice by using room boundaries (floor, walls, ceiling) to virtually increase the size of the terminus -but you'll ideally need to adjust the entire horn expansion profile to account for that, & usually it's not going to hit a given expansion profile precisely anyway as it's very hard to control the position of the enclosure relative to the available boundaries (assuming suitable ones are even available).
What often happens, especially with designs that have a forward-facing terminus is that people either
a/ Take the full-sized horn they calculated and truncate it, i.e. simply chop the end off before it gets too large & live with whatever is left,
b/ Take the horn they calculated, leave the filter chamber & throat alone, and compromise the expansion rate so it fits the desired space once folded up, or
c/ Some combination of those two.
When you do any of these, because the horn terminus (mouth) isn't full-sized for designed cutoff frequency (using the above example, 40Hz) any more, it's no longer impedance matched down to that frequency. But its physical length may not necessarily have changed in proportion to the reduced expansion rate. So what you've got in effect is a pipe with a physical length that's tuned to 40Hz but which no longer has sufficient expansion for horn loading in the usual sense down to that frequency. Net result is that you may (will) have output at 40Hz, but it's from standing waves / pipe resonance / TL action, and output will continue to be from pipe resonances up to the frequency where the effective terminus size is large enough for impedance matching to occur, and you transition over to horn loading in the usual sense of the term. This is where a lot of the large peaks & nulls you see in the low frequency response of many back-loaded bass horns comes from -pipe resonances from an undersized, aka compromised, horn expansion.
Thanks for taking the time Scott, that clears some of it up a good bit.
But -- What does modifiying the expansion rate realistically achieve? I figure that a shorter path length in combination with a greater expansion rate might help some aspect of the horn, but it'd still affect the standing wave side of things, no?
Also, from what I've understood the filter chamber acts as a low pass of sorts. Is that its' primary function, or is there more to it?
But -- What does modifiying the expansion rate realistically achieve? I figure that a shorter path length in combination with a greater expansion rate might help some aspect of the horn, but it'd still affect the standing wave side of things, no?
Also, from what I've understood the filter chamber acts as a low pass of sorts. Is that its' primary function, or is there more to it?
Usually a smaller box is the goal, or for a 40Hz horn in free space you'd have a nearly 45ft^2 terminus area per channel. Being generous, since most of us have floors 😉 we'll call it half that. Still pretty 'substantial' for most domestic situations.
That's its primary use. There are exceptions, but to a point we'd be getting into the field of reactance-annulled back-horns there, & you can probably count the number of those on one hand, with fingers to spare.
It is unfortunate that proper T/S are thin on the ground.
Really makes designing a proper horn much easier.
Here is an example of a horn that uses the corner of a room to acheive an ≈ 8x larger effective mouth.
dave
Really makes designing a proper horn much easier.
Here is an example of a horn that uses the corner of a room to acheive an ≈ 8x larger effective mouth.
dave
Yes in that it might/likely be designed for a higher Qts': (Qts'): (Qts'): (Qts) + any added series resistance (Rs)
No. you're lowering the horn's HF (mid bass in this case) gain BW as defined by its effective Fhm, which in turn requires a higher Qts' for a given driver Fs to maintain its mean/'critical' damping (neither under nor over damped).
Maybe more easily understood from a horn loading POV along with all the extra math (pg. 7):
http://www.xlrtechs.com/dbkeele.com...Preprint) - LF Horn Design Using TS Paras.pdf
There was a very good one early on, but disappeared along with a lot of other interactive technical sites in a vicious virus attack that also basically wiped clean my first then brand computer/BU and all my Lotus 123, Excel, Word worksheets (I forget the name of the IBM ones I used on the plant's System 38? whatever was current in '94) even though I had an up-to-date premium antivirus program.
Maybe these will suffice for 'grokking' the basics?: https://audiojudgement.com/folded-horn-speaker-design/
http://www.mh-audio.nl/Calculators/GExpohorn.asp
Yeah, until the proper math was derived, horns were designed full size and a section of it with the desired gain BW was sectioned out, so to speak (great way for math challenged DIYers) and maybe how the Coral was done from just looking at its cutaway.
Yeah, the midrange proof of concept I did was small bass horn size to get enough acoustic damping, so really only practical for sealed compression horn apps and even then requires such small chambers that IME can only be done over a narrow range of BW designs with very limited driver choices to handle the compression except at very low power.
Aah. So the expansion rate is generally already "compromised" in most conventional folded horns, then?
Ding
-- I've generally conflated your use of the term "Horn" with the speaker unit as a whole. Makes much more sense now. 
Those things don't seem half bad either -- Avesburys, was it? So many options....
-------------------------------------------------------------------------------------------
I've started messing about in Hornresp now, using T/S parameters i found here. Came from a German guy, so they must be correct. Used the "input wizard" to get me into the ballpark, but it still doesn't look good at all -- to my eyes, at least. Here's what it's looking like so far:
Currently feeling like it's a hopeless endeavor and that I should just build something "good enough". Plenty of useful information for me to absorb over the next few days, though. If anyone has and helpful insights or tips I'll gladly receive them, but I'll stop pestering you for now. It's clear that I have some reading to do!