DIYRM / Scott Hinson Multiple Entry Horn build

[HiMu]

I built Scott Hinson's MEH for my wife's 50th birthday party soundsystem. I wanted to try it as a woodworking challenge as I don't have a proper table saw or a cnc. Woofer flares and the throat adapters were done by a cnc shop, but while waiting for the parts I made the woofer flares with a plunge router as a practice. All of the final dimensioning of the parts was done with handtools.

Yes, looks like a propa challenge for a novice, quasi-luddite, mostly handtool woodworker. Let's go!

I roughly cut the vertical secondary flares with this craptastic table saw that doesn't cut straight, no matter how well I set it, the right side of the table is bent. These were the only cuts that I did with this... abomination.

Rough routing of the woofer flares while waiting for the parts from the cnc shop.

And after fine tuning. These are within 1-1.5mm identical with the cnc made parts. Achievement unlocked!

Hand planing the compound mitre flares. The long bevels were cut with a track saw, which required some masking tape and CA glue to prevent the thin parts from moving during the cut. These were impossible to do with my table saw, I tried... and ruined some furniture grade baltic birch plywood.

Parts ready for assembly, at this point I was still waiting for the speaker elements to arrive. In this photo are the throat adapters and woofer flares from the cnc shop. I ordered 4 adapters but got 5 because there would've been a blank piece roughly the size of the adapter. Nice! Now I have parts for 4 speakers and 1 extra adapter for whoopsies.

Test fitting the B&C DCX464 coaxial compression driver to the adapter

Test fitting the B&C 10NW76 woofers and the coaxial compression driver to the primary horn flare before glueing. The complete speaker had to be done in 5 different glue ups.

A lot of thought has gone into this design, not much wasted space inside the enclosure.

The finished speakers. The horn still needs 1-2 rounds of paint and lacquer + final touches for the enclosure.

Finished these a week ago, overall build time was 4.5months but active build time was roughly 3 months, rest was waiting for the parts. My original plan was to take proper measurements outside but the speaker element delivery was delayed for more than a month, and unfortunately now it's already too windy and rainy.

These sound very good, not fatiquing and easy to listen to, very good imaging and surprising amount of bass. I'd describe these as a combination of a pa, studio monitor and HiFi speaker, all expectations were exceeded regarding sound quality.
I'm very satisfied with the build, especially considering my skill level as a woodworker, this was my 2nd pair of speakers that I've built from raw, not pre-cut stock, 4th speaker build overall.

Scott Hinson's 44 page paper, sketchup files and an .stl file for the throat adapter can be found here (doesn't require logging in): https://www.facebook.com/DIYRM/post...-supporting-files-the-bc-3-w/636200011629136/

 

[Audio Science Pal]

Looks good!

 

[kemmler3D]

Impressive! I've been curious about MEH designs because in theory they have a lot of advantages but I haven't had a chance to hear any in person, save a couple minutes in a concrete room once... (Sounded bad.)

 

[IngMishael]

I built Scott Hinson's MEH for my wife's 50th birthday party soundsystem. I wanted to try it as a woodworking challenge as I don't have a proper table saw or a cnc. Woofer flares and the throat adapters were done by a cnc shop, but while waiting for the parts I made the woofer flares with a plunge router as a practice. All of the final dimensioning of the parts was done with handtools.

Yes, looks like a propa challenge for a novice, quasi-luddite, mostly handtool woodworker. Let's go!
View attachment 481888
I roughly cut the vertical secondary flares with this craptastic table saw that doesn't cut straight, no matter how well I set it, the right side of the table is bent. These were the only cuts that I did with this... abomination.
View attachment 481889
Rough routing of the woofer flares while waiting for the parts from the cnc shop.
View attachment 481890
And after fine tuning. These are within 1-1.5mm identical with the cnc made parts. Achievement unlocked!
View attachment 481897
Hand planing the compound mitre flares. The long bevels were cut with a track saw, which required some masking tape and CA glue to prevent the thin parts from moving during the cut. These were impossible to do with my table saw, I tried... and ruined some furniture grade baltic birch plywood.
View attachment 481892
Parts ready for assembly, at this point I was still waiting for the speaker elements to arrive. In this photo are the throat adapters and woofer flares from the cnc shop. I ordered 4 adapters but got 5 because there would've been a blank piece roughly the size of the adapter. Nice! Now I have parts for 4 speakers and 1 extra adapter for whoopsies.
View attachment 481891
Test fitting the B&C DCX464 coaxial compression driver to the adapter
View attachment 481893
Test fitting the B&C 10NW76 woofers and the coaxial compression driver to the primary horn flare before glueing. The complete speaker had to be done in 5 different glue ups.
View attachment 481894
A lot of thought has gone into this design, not much wasted space inside the enclosure.
View attachment 481913
The finished speakers. The horn still needs 1-2 rounds of paint and lacquer + final touches for the enclosure.
View attachment 481886View attachment 481887

Finished these a week ago, overall build time was 4.5months but active build time was roughly 3 months, rest was waiting for the parts. My original plan was to take proper measurements outside but the speaker element delivery was delayed for more than a month, and unfortunately now it's already too windy and rainy.

These sound very good, not fatiquing and easy to listen to, very good imaging and surprising amount of bass. I'd describe these as a combination of a pa, studio monitor and HiFi speaker, all expectations were exceeded regarding sound quality.
I'm very satisfied with the build, especially considering my skill level as a woodworker, this was my 2nd pair of speakers that I've built from raw, not pre-cut stock, 4th speaker build overall.

Scott Hinson's 44 page paper, sketchup files and an .stl file for the throat adapter can be found here (doesn't require logging in): https://www.facebook.com/DIYRM/post...-supporting-files-the-bc-3-w/636200011629136/

Incredible mastery on your end

 

[HiMu]

Impressive! I've been curious about MEH designs because in theory they have a lot of advantages but I haven't had a chance to hear any in person, save a couple minutes in a concrete room once... (Sounded bad.)

Thanks!
These synergy/unity/meh's do make a lot of sense acoustically, that was one of the reasons I got interested on building these without having ever heard this type of speakers. Hinson has done some clever and unconventional choices in this design, which are explained in his paper.

One of the first things I noticed when hauling these to my room, they are positioned quite a bit higher than what would be optimal for traditional "flat" front baffle speakers - these don't have nearly as much negative effects from tilting. Somewhat similar to how Genelec Ones and other coaxial "point source" speakers behave in Atmos setup, there's no noticeable phasing when moving until you get to the extreme vertical/horizontal limits. I can hear that these aren't "measure perfect" 60x90 constant directivity, but are at least in the ballpark.

Highly recommend to go and listen to similar speakers, especially in a non-pa/party/club environment if opportunity arises.

When thinking of how quickly 3d printing has developed since the core patents expired, I predict that we're going to see some nice designs in the next few years.

 

[kemmler3D]

When thinking of how quickly 3d printing has developed since the core patents expired, I predict that we're going to see some nice designs in the next few years.

Definitely where my head is at on these. I'm not sure when I'll have space or time to build new speakers, but 3D printing a fancy horn seems like a good use of it. If some hero with the skills and software wants to simulate a really optimized, curvy sort of MEH design at some point, plebes like me can produce a perfect copy with .1mm precision at home, with no more strain on the brain than cooking a frozen pizza.

This is the sort of STL I would pay decent money for...

 

[HiMu]

If some hero with the skills and software wants to simulate a really optimized, curvy sort of MEH design at some point, plebes like me can produce a perfect copy with .1mm precision at home, with no more strain on the brain than cooking a frozen pizza.

A hero exists! John White has the Solana, which is pretty well documented, I think Erin's Audio corner measured a version of it on Klippel.

Solana — JW Sound

Highly capable point source loudspeaker optimized for flexible deployment in a range of environments.

www.jwsound.live

Klippel measurements

Solana Measurements — JW Sound

Measurements were conducted using a Klippel nearfield scanner. Compression measurement was conducted using a ground-plane method.

www.jwsound.live

 

[mrick39]

Excellent work!

My three speaker (L,R,C) MEH build of Scott's design for my wife's 53.765 year birthday will commence shortly, along with the completion of two SBH18 18" Acoustic Elegance sealed subs to act as stands for the Left and Right MEH's.

(I'm going to need a bigger house...)

 

[kemmler3D]

A hero exists! John White has the Solana, which is pretty well documented, I think Erin's Audio corner measured a version of it on Klippel.

Solana — JW Sound

Highly capable point source loudspeaker optimized for flexible deployment in a range of environments.

www.jwsound.live

Klippel measurements

Solana Measurements — JW Sound

Measurements were conducted using a Klippel nearfield scanner. Compression measurement was conducted using a ground-plane method.

www.jwsound.live

That's pretty much what I was thinking, although the dispersion angle is smaller than I would personally want and there seems to be a little directivity error around 1khz, which you'd think they'd be able to iron out with such an advanced design technique, I wonder what the story is there. On the other hand, the distortion and compression graphs look pretty phenomenal, so I'm not sure they need to make excuses for this design either. Maybe they moved the crossover to prioritize that over directivity, dunno. Definitely a good proof of concept of the kind of thing I'm imagining.

I think we're entering a new era of DIY where the horns and waveguides we can buy (print) are on a new level and put home builds back into contention with SOTA commercial designs, because in a sense they ARE SOTA commercial designs.

Being able to 3D print (in pieces) and assemble very large parts can be very cost effective compared to manufacturing and shipping something comparable. I have a N53 QRD that I printed and built myself for around $500 all-in. It weighs over 70lb, is quite fragile, and is huge (53" on a side) so would probably cost a good fraction of that to ship...

 

[Curvature]

Multiple Entry Horns? Meh.

 

[fluid]

there seems to be a little directivity error around 1khz, which you'd think they'd be able to iron out with such an advanced design technique, I wonder what the story is there.

This is better described as midrange narrowing or waistbanding and not directivity error. The predominant reason for this to occur is due to the baffle size and edge rounding. When a baffle gets big enough without sufficient chamfering or rounding over, this occurs. See here
https://audiohorn.net/mid-range-beaming-and-narrowing/

This is a similar effect described by Keele and dealt with by the secondary flare on Conical horns

https://www.dbkeele.com/PDF/Keele (1975-05 AES Preprint) - Whats So Sacred Exp Horns.pdf

The designer is aware of it and shows how to combat it

https://www.diyaudio.com/community/threads/portable-battery-powered-meh-build.415747/post-8107695

So it becomes a question of practicality vs polar smoothness

 

[HiMu]

Excellent work!

Thanks!

two SBH18 18" Acoustic Elegance sealed subs

I suddenly know what I want for christmas...

That's pretty much what I was thinking, although the dispersion angle is smaller than I would personally want and there seems to be a little directivity error around 1khz, which you'd think they'd be able to iron out with such an advanced design technique, I wonder what the story is there. On the other hand, the distortion and compression graphs look pretty phenomenal, so I'm not sure they need to make excuses for this design either. Maybe they moved the crossover to prioritize that over directivity, dunno. Definitely a good proof of concept of the kind of thing I'm imagining.

There is a quite long development thread at diyaudio, this post should answer some of your questions https://www.diyaudio.com/community/...-powered-meh-build.415747/page-5#post-8015721
And the follow up that @fluid mentioned.

Being able to 3D print (in pieces) and assemble very large parts can be very cost effective compared to manufacturing and shipping something comparable.

The development can be done by just changing some modular parts of the prototype and measuring the changes. Truly fascinating stuff how this can be done!

 

[Curvature]

This is better described as midrange narrowing or waistbanding and not directivity error.

It is directivity error. The rise of the DI curve reverses, and this can also be seen in the early reflections plot, where curves compress in that region relative to on-axis. It's not significant, but there.

 

[kemmler3D]

This is better described as midrange narrowing or waistbanding and not directivity error. The predominant reason for this to occur is due to the baffle size and edge rounding. When a baffle gets big enough without sufficient chamfering or rounding over, this occurs. See here
https://audiohorn.net/mid-range-beaming-and-narrowing/

This is a similar effect described by Keele and dealt with by the secondary flare on Conical horns

https://www.dbkeele.com/PDF/Keele (1975-05 AES Preprint) - Whats So Sacred Exp Horns.pdf

The designer is aware of it and shows how to combat it

https://www.diyaudio.com/community/threads/portable-battery-powered-meh-build.415747/post-8107695

So it becomes a question of practicality vs polar smoothness

Even with the 45mm round over it looks like there's a little bump in the directivity... A little one to be sure, but I couldn't say why it's there.

 

[a4eaudio]

It is directivity error...

Can you provide a good official reference to the definition of "directivity error"? Or is it just a poorly worded term for a general directivity mismatch?

 

[Curvature]

Can you provide a good official reference to the definition of "directivity error"? Or is it just a poorly worded term for a general directivity mismatch?

I don't know that there is an official definition of that term specifically. The CTA2034 document and Toole's book detail the relevant background and interpretations. I searched the standard but the term isn't featured.

I don't see any problem with it being called an "error" in any case. Just like one man's midrange narrowing is another's mismatch.

Funnily enough, @fluid gave this explanation here: https://www.audiosciencereview.com/...error-in-the-measurements.31106/#post-1097155

 

[kemmler3D]

I don't know that there is an official definition of that term specifically. The CTA2034 document and Toole's book detail the relevant background and interpretations. I searched the standard but the term isn't featured.

I don't see any problem with it being called an "error" in any case. Just like one man's midrange narrowing is another's mismatch.

Funnily enough, @fluid gave this explanation here: https://www.audiosciencereview.com/...error-in-the-measurements.31106/#post-1097155

Had the same result when I looked for an official definition.

Actually "error" might be a better term in this context because it's not a mismatch in the standard sense of crossing one direct-radiating driver to another too high. On the other hand "error" might sound too much like a value judgment - I just wanted to note that the change in DI was not monotonic, which was a little surprising because the design seems to have been done in a really advanced way. I understand this stuff is not easy (I only have a vague sense of how you'd go about designing a MEH) but based on the website it sounded like they used some kind of optimizer.

 

[fluid]

Can you provide a good official reference to the definition of "directivity error"? Or is it just a poorly worded term for a general directivity mismatch?

There is no universally accepted definition to my knowledge, this is a term that has come from Amir's reviews, usually with an arrow where he identifies the error. Like any element of language it can mean different things to different people when they read it and see what it refers to in a measurement. In most speakers reviewed the reason for the mismatch in directivity comes from two drivers having very different radiation patterns in the crossover region. A tweeter that is wide radiating at it's lower pass band being mated to a woofer that is rising sharply in directivity. This generally creates a peak or dip in the DI around the crossover.

It is directivity error. The rise of the DI curve reverses, and this can also be seen in the early reflections plot, where curves compress in that region relative to on-axis. It's not significant, but there.

The reason I don't think it is useful to call this directivity error is because the cause for it does not stem from the two drivers having different radiation patterns in the crossover region. The cabinet edges or a horns edges cause this to happen. The universally accepted term for this is waistbanding, it may well have been known in earlier studies but the Keele paper I linked is the earliest description and investigation of it, I have seen. When there is a specific cause and a well accepted term to describe it I think it should be used as that actually aids in understanding of the problem.

In a MEH all the drivers radiate from the horn and unless the ports are located at much beyond a 1/4 wavelength distance apart from each other or another source, they radiate as one or at least close enough for jazz. But they still suffer from waistbanding depending on the width of the horn and the edge termination. Conical designs often use the secondary flare like Keele. Other designs aiming for specific directivities need to more carefully consider the edge.

I don't really like the term error being used in this context as for me that signifies a mistake. It's possible a speaker designer is unaware of the cause for any given directivity change and they made an error in the design. But unless they have been living under a rock for at least the last ten years it is more likely a deliberate choice either to reduce cost or complexity and they accepted the trade off.

But that is perhaps my own semantic issue. I really just wanted to point out the cause of what is seen in the measurements as there is more than one reason why the directivity can have variations.

 

[a4eaudio]

...I don't really like the term error being used in this context as for me that signifies a mistake. It's possible a speaker designer is unaware of the cause for any given directivity change and they made an error in the design. But unless they have been living under a rock for at least the last ten years it is more likely a deliberate choice either to reduce cost or complexity and they accepted the trade off...

This is exactly correct and the point I was going to make but I choose not to post my response earlier in they day. The dictionary definition of "error" is a "mistake". Unless we are talking about Danny Ritchie, or Zu Audio, or someone else who doesn't take measurements, then it is NOT an error, it is a tradeoff chosen by the speaker designer. It would be much more appropriate to simply refer to it as a directivity "mismatch" which is just a factual statement without any judgmental connotation.

It is often stated that it is "just semantics" to dismiss the argument, but note that semantics is the branch of linguistics and logic concerned with meaning. Meanings matter. The majority of people reading speaker reviews on ASR are not registered forum members but people doing research about speakers either to decide what to buy or to learn. It is a disservice to suggest to them that basic design decisions are "mistakes".

 

[fluid]

Just like one man's midrange narrowing is another's mismatch.

This is often the case that different words are used to mean the same thing. The perspective of someone who is designing speakers and waveguides might be quite different from others, and for me the why is important. I can't avoid it if I don't understand it unless I get lucky. For me those two things are not the same although they might look the same in a measurement, I can understand why it might seem like a distinction without a difference.

I just wanted to note that the change in DI was not monotonic, which was a little surprising because the design seems to have been done in a really advanced way. I understand this stuff is not easy (I only have a vague sense of how you'd go about designing a MEH) but based on the website it sounded like they used some kind of optimizer.

There was optimization of paramaters to achieve certain things and there was more information about that process in one of the threads. This is quite different to optimizing the whole radiation pattern to a target, which is considerably more complicated, but that would not be obvious to anyone who has not used these tools.