My first DIY speaker design - 2-way with horn waveguide

[unmutual]

Would like some feedback on my plans, I have ordered the parts, but not started building yet.

The reason for building this is simply that I want to experience a high performance horn speaker, and the joy of making something myself that works.

The stars of the show:

Compression driver: B&C DE250

Waveguide: Eighteen Sound XT1086

Woofer: SEAS H1571-08 U18RNX/P

Cabinet: 18 mm MDF, appr. 22L, ported (bass reflex)

Target tuning frequency: ≈40–43 Hz
Estimated system F3: ≈42–45 Hz

Acoustic crossover at around 1.2–1.5 kHz
Woofer crossover network: 2nd order
Tweeter crossover network: 3rd order + L-pad
Acoustic slopes likely: LR4 / asymmetric LR4

Crossover prototyping will be done with the help of Xsim and REW.


For full disclosure, this design has been worked out using ChatGPT, through a long process with many iterations. I understand that this raises many issues, and without proper review, the results are not likely to be impressive. The original cabinet size it suggested was not big enough to fit the drivers and waveguide, quite amusing that for all the capabilities of AI it still can make such elementary mistakes. So I am certainly not intending to follow all its suggestions blindly to the letter. I have found it very useful as a starting point, instead of spending hours trawling Youtube.

Anyway, I am very interested in some human feedback on the overall concept, and if there are any pitfalls I have already managed to fall into.

 

[alex-z]

I would pick a bigger woofer. You can make a 6.5" woofer play down to 40Hz but it won't have the dynamic range to justify a horn + compression driver.

Something like a Dayton SIG225 makes more sense.

Dayton Audio SIG225-4 8" Signature Series Woofer 100W Driver 4 Ohm

Dayton Audio SIG225-4 8" Signature Series Woofer 100W Driver 4 OhmDayton Audio SIG225-4 8” Signature Series Extended Range 100W Driver 4 OhmThe Dayton Audio Signature Series was designed to give speaker designers access to the pinnacle of speaker technology at the value expected from Dayton...

www.parts-express.com

XSim is a poor tool for crossover design. Use VituixCAD instead, it has native off-axis response modelling, an enclosure modeling tool, even an auto-optimizer for crossover design.

VituixCAD Loudspeaker simulator

Use the enclosure tool to size your cabinet. Brace the inside thoroughly, chamfer the rear of the woofer hole, lightly stuff with porous absorption to absorb reflections.

Once that is done measure the response of each driver in the cabinet. This will give you the data for crossover design.

How to make quasi-anechoic speaker measurements/spinoramas with REW and VituixCAD

NOTE from ASR staff: The following content has useful tips and details, but as REW and VituixCAD continue to be actively upgraded, some parts of this thread may be obsolete. Be sure to consult the respective software sites for the latest updates. Quasi-anechoic measurements are a way to...

www.audiosciencereview.com

Make sure to use loopback timing reference if you have an XLR mic, or acoustic timing reference if you have a USB mic. Otherwise the phase/time alignment will not be accurate.

 

[Smithlog]

Looks like a solid starting point! Your driver choices and crossover plan make sense, just double-check cabinet bracing and port tuning, and you’ll likely learn a lot during prototyping.

 

[Flaesh]

on the overall concept

1.4 (1.5), less often 2", driver, usually a 15" woofer, proportional waveguide, crossover at around 800 (or lower if possible) Hz - our (meaning all of us who preferred such systems) usual concept. (This does noT mean that small speakers are bad))

I doubt that an electrical second order will create an acoustic fourth order for this woofer.
The horn-driver combo will require CD-compensation.
A digital crossover greatly simplifies things)).

 

[Flaesh]

Do you want to go with a carefully finished speakers right away? I've had various horns mounted on top of the woofer boxes for 15 years).
I would use a 10" speaker with 1086.
Will you be using subwoofers? Personally, I'm a proponent of using subwoofers in home systems, especially this type, with a relatively high-range midbass. For several reasons. If the future system is not near field.

Ha! I'm talking more about my concept than yours, I apologize!))

 

[unmutual]

Do you want to go with a carefully finished speakers right away? I've had various horns mounted on top of the woofer boxes for 15 years).
I would use a 10" speaker with 1086.
Will you be using subwoofers? Personally, I'm a proponent of using subwoofers in home systems, especially this type, with a relatively high-range midbass. For several reasons. If the future system is not near field.

Ha! I'm talking more about my concept than yours, I apologize!))

I did consider to have the horn mounted externally, but part of the rationale for the project is to practice my routing skills. Also, on the off chance that the end result is better than just acceptable, I might want to actually put the speakers in a room. The WAF would be low with an external horn.

 

[Flaesh]

Okay, so be it.
The 6.5" speaker will be a limiting factor. But if you don't demand significant)) volume from it, you can\may live with that.
By the way, what speakers do you have? And room and speakers\listeners placement?

 

[unmutual]

I am prepared for the possibility of discovering that it was a design flaw to go with a 6.5" woofer. This is for a secondary setup, main reason was that I wanted to experiment with a horn loaded design. I plan to use these speakers in the home cinema room (which is about 15 m2) and use a subwoofer as well. They will replace the Jamo E370 floorstanding speakers that I discussed in a previous thread.

 

[unmutual]

Should add that I first considered a Purifi woofer but found them too expensive for a first attempt at DIY, hence I chose the SEAS driver instead.

 

[Flaesh]

E370 floorstanding speakers that I discussed in a previous thread.

I saw that thread, it's closed.
You're swapping two 8" woofers for one 6.5".
The E370 is slightly narrower than the 1086, so it won't fit.

 

[unmutual]

Yes I am now building a new enclosure, I have abandoned the idea of using an existing one.

 

[Flaesh]

I'm late!)

Regarding the DE250, the frequency response in 1086 will\may differ from the DE250 datasheet. I'll have to measure it. CD-compensation may be necessary.

 

[voodooless]

I would also opt for a larger driver, 8 or even 10”. It would probably match the directivity better, besides more output.

 

[Flaesh]

 

[hex168]

I also think that you are going to see a directivity discontinuity with a 6 1/2" woofer.

Consider, for example:

LaVoce WAF123.01 12" Ferrite Woofer 8 Ohm

LaVoce WAF123.01 12" Ferrite Woofer 8 OhmThe LaVoce WAF123.01 is an impressive 12" woofer that features an impressive aluminum basket which provides exceptional cooling. The excellent 99 dB sensitivity is perfect for mating to other efficient drivers. The gasket on this driver allows it to be...

www.parts-express.com

That will work in a 22l box and achieve your low frequency target with a Linkwitz transform or similar EQ. Max output at 41Hz (Low E on a bass guitar) is 105 dB. Much, much more at higher frequencies.

FYI, good measurements on the Eighteen Sound XT1086 here (at least for the clone tested):

MTG Designs - Eighteen Sound XT1086 (clone)

Manufacturers Info Brand: ??? Model: XT1086 Style: Elliptical Coverage (H x V): 80 x 60 Minimum Frequency: 800 Hz Throat: 1" Mounting Type: 2-Bolt Dimensions (W x H x D): 260 x 215 x 126mm

www.mtg-designs.com

Note that the XT1086 uses a diffraction slot; current practice in horn design appears to frown on those.

Compare to these:

MTG Designs - Denovo SEOS-12

Manufacturers Info Brand: Denovo Model: SEOS-12 Style: Super Elliptical Oblate Spheroid Coverage (H x V): 90 x 60 Minimum Frequency: --- Throat: 1" Mounting Type: 2/3-Bolt Dimensions (W x H x D): 334mm x 162mm x 92mm

www.mtg-designs.com

MTG Designs - Denovo 2825

Manufacturers Info Brand: Denovo Audio Model: OS-11B Style: Oblate Spheroid Coverage (H x V): ---- Minimum Frequency: 1200Hz Throat: 1" Mounting Type: 2/4-Bolt Dimensions (W x H x D): 285mm x 250mm x 110mm

www.mtg-designs.com

Available here:

Waveguides

Waveguides

www.cinergyaudio.com

I'd pick one of those two waveguides based on your preference for vertical control (better lower with the OS-11) vs. lobing around crossover (probably worse with the OS-11 but needs to be simulated in Vituixcad using various center-to-center spacing).

View attachment 519051

Nothing wrong with a DE250, but this newer driver looks interesting too:

https://en.toutlehautparleur.com/compression-driver-lavoce-df10-144lk-8-ohm-1-0-inch.html

As far as I know, it has to be ordered from Europe, but that's not that bad given its base price.

As long as you are doing a wcveguide two-way, how about some reticulated foam a la Geddes?

Poret® Foam Shop - SWISSTROPICALS

Poret® Foam by SwissTropicals US and Canadian customers, see terms for shipping rates. For all other countries, please, Email me first to receive a shipping quote for your order. Poret® Foam is a brand product made by our partner EMW® Filtertechnik GmbH in Diez, Germany. SwissTropicals, LLC is...

www.swisstropicals.com

Question for Dr. Geddes about his WG foam plug

I've read on your website a brief description and purpose of the open cell foam that you use in your wave guides. I'm looking for a little more information. Are the foam cells random in size/distribution or are they ordered in such a way as to scatter (or perhaps absorb?) sound at certain...

www.diyaudio.com

 

[hex168]

View attachment 519051

Cool nomograph! Where's it from?

 

[Flaesh]

Cool nomograph! Where's it from?

I don't remember, it's from some pretty old book. It's very simplified, but it helps to understand the principle.

 

[MEGB1262]

Hi there,

the 18s xt1086 elliptical horn for the 1 inch driver need already for a decent "look" at least a 10 inch woofer and a minimal cutoff of > 1200 Hz

the 18s xt1464 need at least a 12 inch or 15 inch woofer crossed above 800 Hz

all this horns have their advantages and disadvantages

as far as i know here the real challenges are

A )
serious time alignment, see this German video with English text translation

https://www.strauss-elektroakustik.ch DE / EN available

B )
decent passive crossover with symmetrical acoustic mid range - high range filter slopes
one of the last developments available from Mr. Bernd Timmermanns of Hobby HiFi magazine adopted a 5th order low pass filter and a 3rd order high pass filter to achieve this in a 12 inch woofer 1 inch horn driver + horn combination

C )
so all in all it will be difficult to go through this very easy, but there are a number of well documented DIY projects for this type of speakers

D )

however, the baffle can be built without complex CNC machining, which is a major advantage
here just an example of a quite clean build i had realized some years ago

it was two way with 2 subwoofer as stand mounts - looks like 3 way but it is two way + subwoofer below 100 Hz
8 inch Sonido SFR200A custom built with no whizzer cone and 4.5 inch LMH rTrak wood horn with 18S NSD1095N compression driver
one of my last developments without good measurement equipment, was only possible with tons of try and error with extensive listening tests and active crossovers with 2 miniDSP 2X4 HD in double mono 1X4 HD input / output routing configuration

hope it helps, Stefano

 

[kemmler3D]

For what seems to be a very successful recent DIY 2-way horn project on this site, check out this thread: https://www.audiosciencereview.com/...d-15-and-audiohorn-x-shape-2-way-build.66717/

As others have mentioned one of the main benefits of a horn speaker is to get very high dynamics, so using a small woofer somewhat defeats the purpose.

 

[Wolf11Man]

The thread opener first asked KI and then proudly presented his result to humans. Would be funny, if it would not lead to such results, probably triggering wrong purchase decissions, unrealistic expectations and loss of money. KI wants to please you and keep the "communication" going. If it doesn't know anything about a theme like speaker building, so as part of it's programming, it simply invents advice. If it would be a human we would call him a cronic liar or troll. The alogarithm behind it takes any stupid comment on the web and mixes it up to an even more stupid advice. KI can not even construct a usefull speaker, what else does it mess up? Does your medical doctor, while you consult him, type into a computer? Does he, maybe rely on a KI based diagnostic system? Does this idea make you feel a little sick?
I fear we will see many more still born projects from KI assisted wanabe speaker designers soon, to keep in thread. We are in a phase where KI is silently pushed into our lives, becomming the number one "I know anything better" source for information. It will be hard for any well informed, human specialists to push against an KI invented answer in the near future. The majority of people will take the KI answer as the final truth. Sooner than anyone will expect, the commercialized KI will push you to buy certain products, even if asked for neutral advice, as a certain brand has paid a covert advertising fee to the KI steering company.
Advertising counts on KI opperated, affiliate bias/ undisclosed paid endorsement, the administration simply has to change the laws a little.
This is one of the driving forces behind the KI growth, which sucks up all capital that is availble, leaving nothing for usefull investments.
Beside from commercial interest, anyone can imagine what bad actors will do when KI is released to pretend it is a human voice on the internet. Any rationale, real opinion will be drowned out by countless comments from bots.