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

[Flaesh]

KI

künstliche Intelligenz? In my opinion, the basic idea of a speaker made of a round piece of cardboard and a expanding pipe has been around for quite a while and is almost obvious without KI.

ps: Indeed, working with the GPT is mentioned in the opening post. I probably automatically skip or immediately forget such things.

 

[Wolf11Man]

I use KI every day, for things I know but don't remember in detail. Just like using a pocket calculator. The creative decissions are made by me, from experience.
Like not to pair a 6.5" HIFI driver with quite weak dynamics and a plastic cone from woven polypropylene to a PA compression driver.
88 dB/W/m woofer with a 108.5 dB/W/m will make anyone who hardly knows what a decibel is shake his head. KI helps you to spend money on this combination.
If you have no idea about speaker building, you can not even tell which part of the KI's answers are blatant nonsense.

I asked KI: "how many speaker with 88 db/watt do I need to match the SPL of one 108.5 dB/watt speaker?"
KI answer:
You would need approximately 112 speakers with an 88 dB/W sensitivity rating to match the Sound Pressure Level (SPL) of one 108.5 dB/W speaker, assuming both are fed the same power, and the 88 dB speakers are placed together.
Here is the breakdown of the calculation:

The Rule: Every doubling of the number of speakers (when coupled) increases the total SPL by 3 dB.

 

[voodooless]

You would need approximately 112 speakers with an 88 dB/W sensitivity rating to match the Sound Pressure Level (SPL) of one 108.5 dB/W speaker, assuming both are fed the same power, and the 88 dB speakers are placed together.

So what? You can just pad down the tweeter. Sure it’s a waste, and for sure you should frown on the idea, but it doesn’t create an inherent problem.

Or you also have 100+ dB/watt horn loaded monster subwoofers?

Efficiency just tells you how much of the input power comes out as sound, nothing more…

There are plenty of better reasons for this combo not to work.

 

[sheeple]

I would keep the 6.5 inch woofer if you listen from around 3 meters. All these big boxes are mostly unnecessary in my eyes. Instead, add a cheap 8-inch subwoofer from Dayton to the back panel, if you are handy, you could even add resistive slots for cardioid, and you arrive at something similar as the new small Dutch&Dutch 6c. You must not.

You asked for remarks about possible issues with the design: it is the horn. The issue with all these waveguides is they present a rather small image. The presentation is stunningly clear and accurate, it is a true marvel, but the music sounds as if you were looking at a miniature. At least if the listening distance is short. If you have a big room and listen from at least 4 meters, it gets better. Which is a setting that these PA devices were designed for originally.

There is sadly only few wide horns, JBL has some 120 degrees, but I think they are for 1.4 inch drivers and also they are relatively wide, somewhat contradicting the small woofer or would have to be mounted off-baffle.

 

[Ktacos]

I agree that if you use a CD and horn you tend to have to be pretty far away for them to not sound like crap. I built some VBS 10.2 last year and sold them fast because you had to be like 9ft away before they sounded right. I also tend to find that CD's just don't sound nearly as good as nice dome+waveguide. They also tend to be quite narrow so there's not much room interaction going on which can really put a damper on the speakers spatial properties. I have that DE250 and have tried it in various horns and don't really feel it can keep up with my SB26ADC+waveguide in terms of fidelity.

I also find that that with most woofers and CD's+horns, the drivers impulses are quite misaligned, far more so than a typical two way bookshelf for example. Active filtering is needed to align them and can make a pretty big difference.

 

[MEGB1262]

Hi there,

there is a chance to improve the efficiency of the midrange speaker with a front horn, something like this

but it makes the development probably more challenging, you need a decent woofer too and it will change to three way or at least you need to go satellite / subwoofer

hope it helps, Stefano

P.S.:
a naive approach to using AI in a project like this can lead to strange results, but the thread opener is starting to learn and asking for help from the experts here or not?
he should be careful to start the thread this way in DIYAudio

In the other forum, most people including the moderation team completely lose it if you even mention the word “AI”

 

[Wolf11Man]

A polypropylene cone in a woofer is about the worst you can put inside a horn. A horn works by putting high pressure on the the cone.
You need a very stiff cone and a strong driving force to make that work, not a soft, flexible one.

 

[Super Panda]

Sure it’s a waste, and for sure you should frown on the idea, but it doesn’t create an inherent problem.

Not a waste at all. Yes, some energy is lost as heat but provided you use decent resistors with low inductance this is the best way to "match gains" since you total headroom will be limited by the woofer anyway. DSP-padding comes at a cost to dynamic range - inaudible but still, who dismisses an opportunity to optimise?
I embrace L-pads regardless of what anyone thinks of passive components.
And just to be clear, I'm agreeing with you if that wasn't evident

I'm actually in the process of matching the XT1086 to an SB Audience NERO-12MWN700D for a 2-way system and I estimate they'll match at around 1200 Hz @ 90° but this remains to be seen. The horn will be placed freestanding which will probably make it bloom a bit more at lower frequencies without baffle support, hence the wider than stated 80° dispersion at estimated crossover. The CD (ND1090) will get a 10 db pad which should leave a 3-4 db sensitivity delta in favour of the CD.
So to the OP, good luck on your project! I would take heed of the larger woofer advice others have voiced here but I think it'll be just fine regardless

 

[voodooless]

I'm actually in the process of matching the XT1086 to an SB Audience NERO-12MWN700D for a 2-way system and I estimate they'll match at around 1200 Hz @ 90° but this remains to be seen.

That’s already a totally different setup vs a 6.5” driver. You already gain almost 10dB in efficiency. You’ll also have much better directivity matching.

DSP-padding comes at a cost to dynamic range - inaudible but still, who dismisses an opportunity to optimise

Just use a low gain amp…

 

[MEGB1262]

A polypropylene cone in a woofer is about the worst you can put inside a horn. A horn works by putting high pressure on the the cone.
You need a very stiff cone and a strong driving force to make that work, not a soft, flexible one.

Hi there,
theory and practice
i think your theoretical approach may apply to long bass horns, not to short front horns that do not load below 350 - 400 Hz
i tried with the Sonido SFR200A custom drive unit with no whizzer cone
the front horn shouldn't have worked here either, but it dampens a nasty surround resonance and boosts efficiency in the midrange by 6 to 8 dB
look at the sonido.hu gallery there are a lot of folks using the drive unit with the very light paper cone with a front horn mounted
- Stefano

 

[Flaesh]

Horn Speakers - Is it me or.......

The sad aspect: whenever I did such experiments, horn-loaded monitors were despised and ranked the worst by almost everyone, even by horn lovers among the engineers in blind tests. The only noteworthy exception was the Strauss main monitor which I have linked above. (sigh) I understand, but...

www.audiosciencereview.com

 

[Audionaut]

high performance horn speaker, and the joy of making something myself that works.

This is your first DIY project.
I don't want to discourage you, but the way you're planning to do it doesn't make much sense.
You can use the SEAS woofer, but then a dome tweeter in a waveguide is enough to achieve consistent dispersion and level.

You can use the D250 compression tweeter and a horn, but then a midrange or mid-bass driver from the PA segment makes much more sense. Complemented by a powerful 15-inch woofer, for example.

Your SEAS woofer is about 70 dB at 40 Hz and 75 dB at 50 Hz at 1 W/1 m.

Even a dome tweeter with moderate efficiency will suffice, as that’s what you’ll need to base your design on for a passive speaker.

You want to configure and build everything yourself from scratch, but that’s not trivial and isn’t really realistic for a first-time project that’s supposed to work in the end.
Sound will come out of the speaker and you might be happy about it, but it’s more beneficial for the learning process and not very promising for the final result.

There are many tried-and-true design proposals on the market, including for horn projects. I would use those if I were you.
If you want to experiment and learn, buy a speaker management system like the DCX2496 or, as the cheapest but very effective solution, an ADAU1701 board.
Experimenting and learning with passive components has become expensive and, if the right values aren’t available, frustrating, because you can’t just make changes on the spot.

In any case, whether active or passive, a calibrated measurement microphone and the appropriate Software is needed, which you need to know how to use and be able to interpret the results from.
That’s the most important part, and it’s not something you can wrap up in a few hours. If you don’t measure properly, you’re just guessing.
So you also need to know where potential sources of error might arise during the measurement itself—and there are quite a few of them.