A Bookshelf Multi-Way Point-Source Horn

Founder of XSA-Labs
Joined 2012
Paid Member
This originally started in this thread which asks the question "why aren't we building a bookshelf synergy monitor" and answers it with "we are" or at least "let's try and see what we get."

I got inspired back in the Trynergy thread to build a semi bookshelf point source horn

The results with the SB Acoustics SB65WBAC25-4 were very encouraging that Bushmeister went and ordered the drivers and some off the shelf waveguides to start experimenting.

So here is sort of my vision of the end goal. It's all very fluid and we try to make things smaller where possible. The reality is it might be closer to a small stand mount. Decisions of which WG to bass the system on sets overall size.

524229d1452489305-bookshelf-multi-way-point-source-horn-xbush-01-sb23nrxs45-8-design-sketch.png


Thanks to Bushmeister for the willing spirit to tackle this project. We are going to work together to come up with a new speaker with development and plans happening in real time ad hoc. It starts with Bushmeister picking the off the shelf 18 Sound waveguide.

Here is a summary of the completed speaker built by Bushmeister.

I guess we are calling the "XBush". Post 1248

Some highlights of the XBush:

Predicted sensitivity and response before EQ using a Harsch XO at 500Hz and SB23NRXS45-8 woofers:
536558d1457612468-bookshelf-multi-way-point-source-horn-xbush-01-sb23nrxs45-8-spl-sensitivity-2.83v-50mm-12mm-thick-65hz-hpf.png


The business side of the horn normally not visible when boxed up:
536573d1457612760-bookshelf-multi-way-point-source-horn-img_20160120_214528.jpg


Mounted inside cabinet:
536580d1457612930-bookshelf-multi-way-point-source-horn-img_20160203_072759.jpg


Finished speaker:
536582d1457612930-bookshelf-multi-way-point-source-horn-img_20160211_183015.jpg


Acoustic XO slopes for Harsch:
536584d1457613073-bookshelf-multi-way-point-source-horn-slopes.jpg


Harmonic distortion at 100dB and 1m (nice!). Note that this is just a 2-wa speaker with two sealed 8in drivers - that manage to produce nice deep bass extension with low distortion:
536585d1457613073-bookshelf-multi-way-point-source-horn-dist-2-.jpg


Measured acoustic phase:
536586d1457613073-bookshelf-multi-way-point-source-horn-phase-1-.jpg


Measured Horizontal Polar Directivity:
536587d1457613073-bookshelf-multi-way-point-source-horn-polar.jpg


Measured Vertical Polar Directivity:
536588d1457613073-bookshelf-multi-way-point-source-horn-vert-polar.jpg


Step Response:
536589d1457613073-bookshelf-multi-way-point-source-horn-step.jpg


Measured Group Delay:
536590d1457613073-bookshelf-multi-way-point-source-horn-gd.jpg



######### XRK971's XBUSH ############

Not fully up to snuff yet when compared to Bushmeister's amazing effort, but here is Xrk971's attempt which uses a Faital Pro LTH142 waveguide, dual Dayton RS180P-8 woofers, and same SB65WBAC25-4 mid/tweeter:

Internals with all 3d printed driver to horn adapters (here shown with early version using DC200-8 wofoer):
528945d1454484423-bookshelf-multi-way-point-source-horn-lth142-swcond-woofer-01.jpg


Later version with RS180P-8 woofers:
530860d1455371559-bookshelf-multi-way-point-source-horn-image.jpeg


Detail of SB65 throat adapter:
526756d1453501391-bookshelf-multi-way-point-source-horn-sb65-long-adapter-photo.png


CAD drawing of adapter:
526771d1453502400-bookshelf-multi-way-point-source-horn-xbush-throat-adapter-v3-wireframe.png


All 3D files (STL) for xrk971’s Xbush are here:
A Bookshelf Multi-Way Point-Source Horn

Fitting into box for first time:
529771d1454885057-bookshelf-multi-way-point-source-horn-baffle-black-1.png


Measured Polar directivity of SB65 in LTH142 waveguide:
526844d1453538138-bookshelf-multi-way-point-source-horn-sb65-lth142-eq-6th-fdw-polar.png


3d printed woofer to waveguide adapter:
527484d1453821187-bookshelf-multi-way-point-source-horn-horn-woofer-adapter-01.jpg


3d printed woofer chamber volume reducer plug:
527784d1453987194-bookshelf-multi-way-point-source-horn-woofer-mount-05-volume-filler1.png


Polar directivity with woofer holes:
528426d1454257077-bookshelf-multi-way-point-source-horn-sb65-lth142-eq-6thfdw-polar-vertical-big-holes-repaired-10db-scale.png


529809d1454908866-bookshelf-multi-way-point-source-horn-baffle-black-2s.png


Used in stereo with uTrynergy on left and Xrk's speaker on right:
531021d1455442962-bookshelf-multi-way-point-source-horn-utryn-xbush-stereo-setup.jpg


Acoustic XO slopes of Xrk's speaker:
534203d1456601648-bookshelf-multi-way-point-source-horn-xbush-harsch-xo.png


Phase of Xrk's speaker (still working on getting it flatter and proper 55deg phase bump for Harsch):
534204d1456601648-bookshelf-multi-way-point-source-horn-xbush-harsch-phase.png


Harmonic Distortion of Xrk's speaker (not very good - many leagues below what Bushmeister is able to do):
534206d1456601648-bookshelf-multi-way-point-source-horn-xbush-harsch-hd.png


I suspect the higher HD has to do with a number of things including: driver choice, cabinet volume and wall material, cabinet resonance damping methods, waveguide inertness via 4kg of epoxy Bondo, and larger holes (60mm vs equiv 29mm).
All pieces and adapters on Xrk's speaker were 3d printed PLA vs hand cut and epoxy for Bushmeister's excellent effort. Sometimes low tech gives better results.

So, my speaker is still a project waiting for more work to improve the phase, the step response, and improve the harmonic distortion.
 
Last edited:
  • Like
Reactions: 3 users
Founder of XSA-Labs
Joined 2012
Paid Member
It's fun to go wild when someone else has to make it.

It's really 3 way sealed speaker with tweeter and mids going out of horn and woofer is big sealed box firing down. Box can be made simple rectangular prismatic but where's the fun in that. :). Also really I am trying to keep volume small and size visually slimmer.
 
Sealed three way monitor bushmeister linked the other thread is serius speaker build, looks a good reference to compare Bookshelf Multi-Way Point-Source Horn.

Discovered my WinSpeakerz program can model band pass box so for fun did model SB12NRX25-8 in Q 0,7 sealed verse same having a band pass front box and port. xrk971 if you find time could you run Akabak and see if it agree frq response for band pass setup, data used is 32 watt input, 1,999 liter back box volume, 0,05 liter front box volume, port diameter 34mm, port length 30mm, front box resonance 1000Hz, BW2 low end roll off for sealed example is 108,9Hz for band pass looks more like 88Hz with a couple dB lost sensitivity verse ordinary sealed.
 

Attachments

  • 1.PNG
    1.PNG
    86.6 KB · Views: 674
Founder of XSA-Labs
Joined 2012
Paid Member
Thanks for running that Byrtt. Once it couples into the horn - the behavior may be different in that horn adds a few more dB mostly due to directionally because of horn. There will be overshoot peak from port but that can be EQ'd out. Does look like 500Hz is a good XO point. I need to estimate to expansion profile of the 18Sound WG and put that geometry into the model for Akabak to predict this well. I guess it's time to put some simulation to this to predict bandpass hole size and location for optimum results.

Yes, seeing Bushmeister's work on that very formidable 3 way sealed box gives me high confidence we are going to get this horn to work well.
 

ra7

Member
Joined 2009
Paid Member
I admire the spirit of X and also am a little jealous of the speed with which he builds speakers. Wish I could turn all my wild ideas into actual builds this fast.

The point of the synergy horn is controlled directivity in the low frequencies. This, by design, means the mouth of the horn must be large and the horn itself must be relatively large. The lower in frequency you want to control where the sound goes, the bigger the horn must be. As diyers we can build whatever size speakers we want. But can we really call these "bookshelf?"
 
Founder of XSA-Labs
Joined 2012
Paid Member
For me, this is a FAST horn and if I can get controlled directivity down to the bottom of the full range at 500Hz I am good. That cleans up the ambient spray a lot.

But like I said, it depends on how big your books are:)

There's no reason why we can't all build crazy new speakers fast and all the time. For me it was lack of workshop so I use foam and glue.

I had a crash course in designing and building over 3 dozen speakers in past 3 years. I average 1 a month. Not that fast - some people whip out wood speakers in a few days. But trying new things is fun. The Synergy was a challenge but doable in foam.
 
Last edited:
Wow Guys! I go to bed and a whole design starts to come together whilst I am sleep!!

Thank you so much for the input and effort you guys are putting in!

If we can make this work it would be fantastic! I am very excited about this project. So far, I have built multiple full range speakers in various boxes - ikea salad bowls, fish bowls, boom boxes. etc.

I have built a number of straight forward two way designs - mostly using SB acoustics drivers - which I think are fantastic value. One set I built in left over engineered oak floor boards - which created a surprisingly dead cabinet when combined with adequate damping.

I have built those high end three way monitors- which I am still overjoyed with, and as a result of making, I sold my studio monitors which were my previous reference.

I have tried my hand at CLD baffles, decoupling of drivers with sorbothane etc, various bracing techniques - including CLD bracing - which works amazingly well.

But all these techniques have been with direct radiator builds, and whilst I have always attempted to keep a smooth, even power response and carefully manage lobing behavior, I have not done a CD design like this before.

Xrk - love the sketch! However, I was thinking more about trying to keep this as bookshelf as possible (of course all ideas are fluid at present!), so I was thinking about getting the horn itself to play as close to 80hz as possible, then use subs below.

I use four distributed subs with my three-way scan-speak design, and it works incredibly well.

However unfortunately, I am still awaiting these SB65s, which is very frustrating!
 
Another thought I have had:

Tapping the horns - they are extremely rigid, light, with a very high pitched 'ring' on the 'tap test'. In my experience, this tells me they will likely have very high stiffness with high resonances. These can often be tamed very successfully with existential damping - of the form used in cars (think dynamat) to dampen resonances in the metal structure. (This form of damping is much less useful for lower resonances you get in heavy thick MDF/Ply constructed cabinets - for which the best technique is CLD.)

So I am thinking, that covering the outer surface of the horn profile with 4mm thick bitumen damping product could help reduce unwanted resonances and lower distortion further.

I hope this, combined with the sorbothane gaskets - which should minimize direct energy transfer and mechanical excitation of the horns - should give us a very clean output.
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
You are in luck because according to Byrtt's sim of the woofers in a 2liter(ea) rear chamber it goes down to about 80Hz. Now it is much smaller and closer to a "bookshelf" vs a design with built in woofer. XO around 80Hz with your distributed subs is perfect then. I still need to confirm Byrtt's simulation but we are probably at least not completely off. There is always LT to push it up a bit at expense of excursion so max SPL may be affected.

Do you know if the profile of the WG is available anywhere? I would like to model this horn with bass injection holes and we can get close to optimal before cutting injection ports.

It would be useful if you can take the factory PDF drawing and use a ruler to get me at least 5 dimensions of horn height and width at 5 different axial distances. Every 20% of length basically. This would be enough to form a rough model in Akabak.

Also, can you can please use either geometry to estimate the volume of the front cone of the woofer formed if you place it against a flat wall. If you have driver in hand easy to use dry rice or beans and measure volume.

These hard numbers come a long way to getting to a rough model.

Thanks
 
Last edited:
Founder of XSA-Labs
Joined 2012
Paid Member
Once this is done it would be great to measure it Geddes style and send the files to him. If you ask him nicely he will likely map it for you.
GedLee LLC

Good idea - I am sure Bushmeister has capability to measure polar response.

2d color contour plots of SPL vs frequency aren't exactly hard if you have scientific plotting SW that does interpolation from an array. Like Mathcad can do this quite readily. I have not plotted DI maps before but this type of plot is standard fare in Mcad.
 
Founder of XSA-Labs
Joined 2012
Paid Member
Another thought I have had:

Tapping the horns - they are extremely rigid, light, with a very high pitched 'ring' on the 'tap test'. In my experience, this tells me they will likely have very high stiffness with high resonances. These can often be tamed very successfully with existential damping - of the form used in cars (think dynamat) to dampen resonances in the metal structure. (This form of damping is much less useful for lower resonances you get in heavy thick MDF/Ply constructed cabinets - for which the best technique is CLD.)

So I am thinking, that covering the outer surface of the horn profile with 4mm thick bitumen damping product could help reduce unwanted resonances and lower distortion further.

I hope this, combined with the sorbothane gaskets - which should minimize direct energy transfer and mechanical excitation of the horns - should give us a very clean output.

Good idea, but you can take it further by covering the whole exterior side of the hard plastic with sand loaded caulking or other viscous adhesive. Use it to fill gaps between the wooden board and the horn. I have heard of people covering plastic horns with dampening caulking to great success. If you tap it and it rings there may be an issue there. Once the four wooden boards are attached and holes drilled the mode of vibration will probably change (for the better as less "bell like").
 
Founder of XSA-Labs
Joined 2012
Paid Member
OK Xrk, these are the best plans I could get online - using screenshot etc.
What further measurements specifically would be helpful?

So along 5 points on axis of middle picture measure correspond max and min elliptical diameter (Max is horiz and min is vertical) and label it in a sketch similar to third picture on right. See how it starts as 36mm round (actually now modified by you) then goes egg shape elliptical? I basically need these to calculate the WG cross sectional area as a function of axial distance from the 36 mm throat.

You can sketch the drawing and save as PNG or whatever.

Thanks.

Tip : it is much more precise if you cut cardboard ellipse to fit and measure cardboard. Do this for 4 points since throat is round.