Hi all! I recently saw this post from @All_1 which prompted me to finally make a write up on this project. I started this over in the Speakerplans forum during Covid, but only recently managed to finish a build. I wanted to hopefully consolidate our knowledge on how to effectively sim boundary/baffle coupled style horns.
I've only had an hour to put sound through it, and got just one quick and dirty nearfield measurement so far. My laptop was then stolen so I lost some sim data and the measurement barring a few screenshots. I'll have more time in a few weeks which should hopefully include some outdoor measurements. Here's the build:
Here's the measured vs simmed response:
A few notes. I suspect the nearfield measurement is coloring things, hopefully making the response appear more like a 2pi measurement but tbd. Even so, a close response. The main thing for me was making sure that low corner was in fact there, since in Hornresp this required some guesswork (looking forward to diving more into this). The panel that's ratchet strapped on would start resonating like crazy at a certain point as well (I guess more reason to build the second soon) which more than likely also colored this not quiet measurement (UMM-6 tops out at 120dB I believe).
Here's the group delay measurement vs response:
As for subjective notes: I'm very happy so far, but much more testing to be done. The room I have this in is big with a tall ceiling, but the nearfield was a pleasant and "tight" massage. I'm starting real small and just have a 450 watt amp, but I was barely seeing the signal light. Simmed response suggests this cab will hit xvar at around 900watts.
I definitely want to shoutout everyone over in the speakerplans forum, especially Jo bg and citizensc for humoring my ramblings and running some great Akabak sims that convinced me to cut some wood and chance it, and all those that started threads here in diyaudio. I also need to thank several local sound crews in Seattle for their input (Milestone Productions and another crew to emerge soon...) And of course Tom Danley for all the inspiration and for creating this type of design (Lambda Labs Digital horn also deserves a shoutout, although I believe it's kind of a TH version of this). Oh and of course David McBean for creating Hornresp!
Let me know what y'all think and if you have any Qs. Hoping for more measurements soon.
I've only had an hour to put sound through it, and got just one quick and dirty nearfield measurement so far. My laptop was then stolen so I lost some sim data and the measurement barring a few screenshots. I'll have more time in a few weeks which should hopefully include some outdoor measurements. Here's the build:
Here's the measured vs simmed response:
A few notes. I suspect the nearfield measurement is coloring things, hopefully making the response appear more like a 2pi measurement but tbd. Even so, a close response. The main thing for me was making sure that low corner was in fact there, since in Hornresp this required some guesswork (looking forward to diving more into this). The panel that's ratchet strapped on would start resonating like crazy at a certain point as well (I guess more reason to build the second soon) which more than likely also colored this not quiet measurement (UMM-6 tops out at 120dB I believe).
Here's the group delay measurement vs response:
As for subjective notes: I'm very happy so far, but much more testing to be done. The room I have this in is big with a tall ceiling, but the nearfield was a pleasant and "tight" massage. I'm starting real small and just have a 450 watt amp, but I was barely seeing the signal light. Simmed response suggests this cab will hit xvar at around 900watts.
I definitely want to shoutout everyone over in the speakerplans forum, especially Jo bg and citizensc for humoring my ramblings and running some great Akabak sims that convinced me to cut some wood and chance it, and all those that started threads here in diyaudio. I also need to thank several local sound crews in Seattle for their input (Milestone Productions and another crew to emerge soon...) And of course Tom Danley for all the inspiration and for creating this type of design (Lambda Labs Digital horn also deserves a shoutout, although I believe it's kind of a TH version of this). Oh and of course David McBean for creating Hornresp!
Let me know what y'all think and if you have any Qs. Hoping for more measurements soon.
Last edited:
Looks large!
What are the external dimensions?
What drivers did you use?
What does HBH15 stand for?
Thanks Jason, would love to hear your rig someday. Looks beefy and love the tunes you play.
Hi Art! Was hoping you might chime in here. Dims are 1160mm x 757 x 518 (~46" x 29 x 20.4). I thought that was actually somewhat small compared to a lot of the other diy options, but it is still a horn sub. I went for the 15NW100 for a few reasons. Not the top of the line driver but in all my sims I found I was only able to get 1-2db additional output with say the 15SWs or even 15DS115 for a similar footprint. Those drivers are also more power hungry and expensive which was a key consideration for me. Having said that, the wood cost somewhere around 700 USD with CNCing so a few hundred more on a better driver isn't too much relatively. The SWs sim the closest to the NW in this cab, but I haven't found another driver I could really swap out yet.
As for "HBH" I'm thinking "Hallowell Baffle Horn", but if you read through the speakerplans post you'll see I'm liable to change names at any time 😅
My main takeaway and theory at this point is you really can get some "free lunch" in the form of additional path length/low corner by cutting a perpendicular mouth exit with standard FLHs (purpose of the speakerplans thread). It's not really free because you do increase group delay depending on the size of the opening but a worthy tradeoff in my opinion. Would love to throw some SynTripPs in the mix someday, but I'll probably be starting with some cheaper off the shelf mid-tops to get the ball rolling.
Hi Art! Was hoping you might chime in here. Dims are 1160mm x 757 x 518 (~46" x 29 x 20.4). I thought that was actually somewhat small compared to a lot of the other diy options, but it is still a horn sub. I went for the 15NW100 for a few reasons. Not the top of the line driver but in all my sims I found I was only able to get 1-2db additional output with say the 15SWs or even 15DS115 for a similar footprint. Those drivers are also more power hungry and expensive which was a key consideration for me. Having said that, the wood cost somewhere around 700 USD with CNCing so a few hundred more on a better driver isn't too much relatively. The SWs sim the closest to the NW in this cab, but I haven't found another driver I could really swap out yet.
As for "HBH" I'm thinking "Hallowell Baffle Horn", but if you read through the speakerplans post you'll see I'm liable to change names at any time 😅
My main takeaway and theory at this point is you really can get some "free lunch" in the form of additional path length/low corner by cutting a perpendicular mouth exit with standard FLHs (purpose of the speakerplans thread). It's not really free because you do increase group delay depending on the size of the opening but a worthy tradeoff in my opinion. Would love to throw some SynTripPs in the mix someday, but I'll probably be starting with some cheaper off the shelf mid-tops to get the ball rolling.
NICE build! Now you need to build 3 more to get that boundary effect.
I like https://forum.speakerplans.com/ too!
I like https://forum.speakerplans.com/ too!
Wow, this looks super interesting. We will also try to post more data in the future, they (4 BC118) will be used on first live event (DJ/EDM) next month together with some tapped horns.
Your results look quite good, probably perfect balance between sensitivity and extension for most music material 🙂
Your results look quite good, probably perfect balance between sensitivity and extension for most music material 🙂
Thanks @BP1Fanatic & @All_1! Plan is to definitely build at least one more and see how they act in a pair, but the target was always ~140db with an f3 of 33hz (low C) in a 4 stack, 2pi. Also under 150lbs per cab (still needs weighing). There's definitely some music that goes lower, but you can gain a bit more extension with more cabs and boundary loading. If lucky there may be some additional coupling to the baffle as well, but we shall see.
All_1 I'm super curious about the BC118! That's awesome you've already got plans to get them out. If they hit 150dB that's quite an order of magnitude above this cab, and plays lower too 😤
All_1 I'm super curious about the BC118! That's awesome you've already got plans to get them out. If they hit 150dB that's quite an order of magnitude above this cab, and plays lower too 😤
Last edited:
Quick little update here. I was finally able to have a proper session this last weekend but due to snow was unable to get any outdoor measurements. Until I get those it probably doesn't make too much sense to try and correct the sim, but here's hornresp exported into REW to compare a nearfield measurement in 1pi. I suspect the rear chamber volume may be a bit smaller than initially thought due to the faster roll off and slight bump at the low corner. I was also unable to really push it at all due to windows and other objects rattling, but I'm certainly quite content so far (I'm sure builder's bias is playing a part). I ran a 32hz HP and slight bell cut at 67hz and I was even getting some decent weight from "Not Like Us", a root note of 30hz/Low B (that's low for the Grammy's!) This should be approximately 3 watts:
Woah neat! Very similar response indeed. Without hearing them side by side I couldn't say. It does look like group delay is a bit higher at tuning in the tapped pipe, but honestly I don't know of any reliable studies showing that you could discern the difference at those frequencies. Scott Hinson has a good paper comparing the different designs here: https://drive.google.com/file/d/1NJ06zwwOFmJIcNhJP6DEPeMqa8pnnKbS/view
Interesting. I believe I said something similar in another thread comparing a 3ft3 sealed enclosure to a 3ft3 BP4 enclosure with the same driver.
