Best waveguide under 17.75x6.75" mouth for The Selenium D220Ti

...

I've been trying to buy some EOS-6 and EOS-8 from Erich. He says he still has some cases of them *somewhere* but can't find them just now...:confused:

I didn't want to leave this comment dangling on a negative note. Erich located his stock of small EOS waveguides that were listed on the old diysoundgroup.com site, and went beyond the call of duty to make sure I got what I was looking for, for a nice price. A pleasure to deal with, if you understand his busy-ness and have a bit of patience. :)
 
The venerable QSC waveguide (no longer available) is awesome. This one: B-52 PHRN-1014 1" Horn 10" x 14" Bolt-On appears to be identical. It should unload at around 1050Hz. The PH612 should unload at around 1400Hz, so if you plan on mating this with a 15" woofer you might prefer the larger waveguide.

Nice catch.

I've built speakers with the QSC waveguide, the PRV clone of it, the 18 Sound XT1086 and the Pyle PH612.

Some random observations:

1) this appears to be the PRV clone. It has a different mounting plate than the QSC

2) the BMS and JBL ring radiators work exceptionally well on the PH612. (It was designed for them.)

3) the XT1086 is basically my "go-to" horn. The other two are nice, but the 18 Sound is hard to beat at it's size.

4) if you can live with a big cabinet, go for the QSC clone

All of these waveguides have measurements in "the great waveguide list"
 
Patrick Bateman - Have you tried the Peerless DFM2544 with the QSC/PRV/B52 waveguide? I was hoping that with the relatively large Xmax (0.3mm) I could get the combo to go down low enough to meet up with a JBL 2226. I'm not too concerned with the CD's non-linearity as I can EQ with DSP (it will be an active system). I am planning to build a sealed 2.5 foot^3 cabinet placed right up against the wall which should net me an f-6db of 35Hz. Add a little DSP low frequency boost and it should be OK without subs for music.


My plan B is to go to a 12" mid-woof. Perhaps a Dayton PA310, or if win the lottery a TD12H, both of which could reach a little higher (~1kHz)
 
I bought the Dayton and B52 horn and sent them to Scott Hinson (SpeakerScott) to measure. He measured them and the LTH102 with a B&C DE250. Results: Less of a horn than waveguide, they're way too shallow and expand too rapidly (which would make sense to anyone who understands horns), so they don't have good LF response. I uploaded his REW measurements. I think what I'm going to go with is this because it is much more hornlike, with a less rapid expansion and deep, meaning it has better directivity than just about any other horn for a 1" CD in the same size and it also has a coverage angle I like. I might use this DIYSG CD with it, I don't know yet, but it's a great candidate: Denovo Audio BA-750 Compression Driver FaitalPRO | HF Horns | LTH102

Here are his REW measurements: He also did them hot at 170F, and you can see the affects there, very good data: Tweeter Measurement Saturday.mdat - Google Drive
 
Last edited:
Depends on how you define, or understand "directivity" and what sort of pattern control you are looking for, I guess. I just loaded that mdat file into REW and had a look at the curves.

The LTH102 is a tractrix horn, which by definition will tend to beam more narrowly as frequency increases. This is the antithesis of "constant directivity".

In my world, constant directivity means the frequency response curve maintains (within reason) the same basic curve/slope as you move off-axis. The overall amplitude changes evenly: lower freqs should drop nearly as much as higher freqs as you move off-axis.

The high temperature curves are academic and just throw noise into the mix.
Let's cut through the weeds: I narrowed my focus to 4 curves.

2: B52 2V
17: LTH102 0H

I assume these are on-axis curves measured under identical or similar conditions.

7: B52 40H
21: LTH102 40H

I assume these are 40 degree off-axis curves measured under identical or similar conditions.

Given these assumptions, the B52 is about 3dB hotter than the Faital in the 1kHz to just below 3kHz range, on-axis. Above 3kHz, they are identical, for all practical purposes.

Now, let's look at the 40 degree off-axis graphs:

The FaitalPro is now 5dB hotter than the B52 below about 1.5kHz.
Once you get a bit above 3kHz, the FaitalPro averages 5dB down from the B52, up to 7dB difference at about 12kHz.

To me, this indicates that the B52 has far better (10-12dB) off-axis response consistency compared to on-axis, than the FaitalPro DUT. This defines the goal of constant directivity.

If you want laser-beam highs, and more low-range off axis, go with the LTH102. If you want a consistent balance of lows-highs on and off-axis (within the range of the horn), go with the B52. This is the more constant-directivity option.

I ignored the Dayton H812 for the purposes of this comparison, because its curves have certain elements of both of the two devices discussed, and would only cloud the comparison- astute observers will draw their own conclusions.

Thank you for posting these measurements, they provide fodder for an interesting brain exercise. With a 1" driver, you're not going to use it below 1kHz anyway, are you? Even with the Denovo BA-750, which is, by all accounts I have read, an excellent driver. Any variations in the "low-end 'loading' range" of the horn can be dealt with in crossover network simulation and measurement.

Peter

Edit: regarding the "they're way too shallow and expand too rapidly" observation: Any constant-directivity horn/waveguide will have a flare rate which averages very close to conical. Which means that, for example, a 90 degree horn which is 12 inches wide shall be 6 inches deep. Give or take variances for an OS throat entry, mouth roundover, etc. A longer horn may have other desirable characteristics for certain applications, but constant directivity is not one of them.
 

Attachments

  • rew1.png
    rew1.png
    114.1 KB · Views: 429
Last edited:
Steve- I hope my explanation made some kind of sense. It was typed in the wee hours with the aid of a bottle or so of wine, but a sober re-read tells me I didn't stray far.

I've only been a student of the science and sound of horns for 50 years or so, I hope some of the right stuff has stuck in my brain.:)

Lots of folks really like those Dayton H812 horns. All horns are a compromise of one thing against another, with a usable 3 octave/1 decade range at best.
The Daytons are a clone of a JBL PT waveguide which was made legendary by our dearly departed friend and mentor, Zilch, in his Econowave designs. You may be aware of this, but it bears repeating, and provides fresh search-engine fodder for the uninitiated ;)

I bought 4 of those B52 (QSC clone) horns. Two to play with in a stupid 5-way Hartley/JBL/Eminence system I'm working on: Hartley 18 IB subs, JBL 2204H mid-bass, JBL 2123H mid, JBL 2426J hi-mid, Eminence N151M-8 tweeter. That should be fun when I put it all together.

but I'd love to do a hack-it Unity horn design

If you haven't, you should read this thread: Small Syns
Bill Waslo did a small unity/synergy-horn design built on a Denovo SEOS-15 waveguide. I've been thinking the other pair of B52 horns might lend themselves well to that exact approach.

Have fun with this project. I hope my input is helpful in any small way.

Peter
 
Thank you Peter. It has made a lot of sense. I have read that thread, but I don't think I want to use the SEOS because it's more of a waveguide. Doesn't really provide the acoustic impedance benefit of a horn. I downloaded his calculator. What exactly is the other horn you want to use? Just curious.
 
Last edited:
For which application?

I'm thinking of trying the B52 for a Waslo-style synergy horn.

I'm planning to use the other pair of B52 with JBL 2426 drivers (that may change- I might swap them for Radian or Celestion or BMS) in the 5-way. But, I own the 2426s, so they'll get a fair trial first.

I bought some Denovo EOS-6 that I was planning to try with the N151s in the main system above 7-10k-ish, and I'll still check them out, but I took a plaster cast of one so I could visualize the flare, and they may be E, but they sure aren't OS :eek:

They look more like an Elliptical Exponential horn as I look at the cast mould. So they wont be CD... but does that really matter when I'm only asking them to cover just over an octave? Maybe not. But I'm going to buy a pair of B&C ME10 horns to audition and measure alongside them, and see which I prefer for the vhf tweeter.
 
because it's more of a waveguide. Doesn't really provide the acoustic impedance benefit of a horn.

This is an old saw which has been hacked to death by the likes of Earl Geddes, Wayne Parham, and David Smith.

Granted, the SEOS waveguides *are* too short, which causes measurable response ripple in the passband. Most people who use them find that they are very nice, and the ripple can be effectively dealt with in the crossover design, and/or with careful EQ. Mr. Parham may differ on that opinion.

But simply put, a constant directivity horn has very simple mathematical constraints. Conical. With allowances for throat and mouth transitions. Earl Geddes will say: Oblate Spheroidal. But OS is as near as Damn! is to swearing, to conical overall. The refinements are in the curve which matches the driver exit angle to the horn angle, for the most part. Pardon me, Earl, for over-simplifying.

So, some basic Trig will define the horn proportions. A 90 degree horn will be twice as wide as it is deep. +/- minor fudge factors.

The "acoustic impedance benefit of a horn"-- sorry, this has been debunked many times over. The lower usable frequency limit of a horn is determined solely by the mouth dimension. A 12" horn is good down to 1kHz. A 24" horn is good down to 500Hz, as a rule of thumb. Any nonsense about flare rates and acoustic impedance benefits is just that, nonsense. The only thing that is affected is directivity.

To clarify, I am talking about horns which operate in the range above the Schroeder frequency of the room. Bass horns are a different animal, directivity is irrelevant, and the concept you refer to does, in fact, apply.
 
Last edited:
Depends on how you define, or understand "directivity" and what sort of pattern control you are looking for, I guess. I just loaded that mdat file into REW and had a look at the curves.

In my world, constant directivity means the frequency response curve maintains (within reason) the same basic curve/slope as you move off-axis. The overall amplitude changes evenly: lower freqs should drop nearly as much as higher freqs as you move off-axis.

Given these assumptions, the B52 is about 3dB hotter than the Faital in the 1kHz to just below 3kHz range, on-axis. Above 3kHz, they are identical, for all practical purposes.

Now, let's look at the 40 degree off-axis graphs:

To me, this indicates that the B52 has far better (10-12dB) off-axis response consistency compared to on-axis, than the FaitalPro DUT. This defines the goal of constant directivity.

If you want laser-beam highs, and more low-range off axis, go with the LTH102. If you want a consistent balance of lows-highs on and off-axis (within the range of the horn), go with the B52. This is the more constant-directivity option.

This explanation was great and the measurments are also a great find. Can you help me understand or tell me at which frequency the horn loading ends on the B52...? The Dayton specs say 600hz while others have said that a horn that size can only play down to about 1000hz but the measurements should show us right? If you can point me in the right direction...
 
The "acoustic impedance benefit of a horn"-- sorry, this has been debunked many times over. The lower usable frequency limit of a horn is determined solely by the mouth dimension. A 12" horn is good down to 1kHz. A 24" horn is good down to 500Hz, as a rule of thumb. Any nonsense about flare rates and acoustic impedance benefits is just that, nonsense. The only thing that is affected is directivity.
Debunked? By whom?
It seems you're ignoring the basics of horn physics.

A pure conical horn with the mouth size of a garage door won't load the driver as well as a trumpet-shaped long exponential or hyperbolic horn with a mouth circumference only 1/10th the size of the door.
Of course, the latter's directivity is anything but constant and the throat impedance won't be (nearly) as smooth.

6-Figure2-1.png

Of course