BG Horn Mk1 - 36hz Front loaded horn
Printed From: Speakerplans.com
Category: Plans
Forum Name: Other plans
Forum Description: Discussion / Questions about all the other plans
URL: https://forum.speakerplans.com/forum_posts.asp?TID=105410
Printed Date: 26 March 2026 at 8:33pm
Software Version: Web Wiz Forums 12.08 - https://www.webwizforums.com
Topic: BG Horn Mk1 - 36hz Front loaded horn
Posted By: citizensc
Subject: BG Horn Mk1 - 36hz Front loaded horn
Date Posted: 22 June 2020 at 3:46am
|
Hi Guys I have made this thread as a place to access the plans and ask questions about the BG Horn. PM me if either of the below links go dead. https://www.dropbox.com/s/q7okm9x5utpp8gt/BG%20Horn%20Mk1.pdf?dl=0" rel="nofollow - PDF https://www.dropbox.com/s/n7d04zrc3088ewx/BG%20Horn%20V1.0.skp?dl=0" rel="nofollow - Sketchup What is the BG Horn? It is a 15 inch front loaded horn covering ~36-90hz. It is designed for the 15BG100. It has similar folding to a Punisher but is larger, more sensitive, plays lower, and should take more power. Why should you build the BG Horn? -You need a FLH that plays below 40hz -You want to build a FLH but other designs look too complex and hard to build -You want a FLH that that is a manageable size and weight, the BG Horn is about 75kg loaded -You would rather not spend big money on they types of drivers that are popular in DIY horns (pd.1850, V18-1000, 18sw115 etc.) -You want to build a horn with a driver that is widely available and is likely to stay that way for some time -You don't live in Europe or the US and find it difficult to get good quality 18 inch drivers due to air freight restrictions. Why shouldn't you build a BG Horn? -You don't feel comfortable with building untested design. If you would like to give this plan a try, hit me up via PM or comment in this thread, id love to hear about it! -You plan to build less than 4 cabs, this may work okay with two cabs but is designed to work in groups of 4 or more. -You need a cab that is a one man lift (build single reflex cabs if you want this) -You are looking for a one-way bass solution that plays above 100hz. Here are some sims show where it fits - keep in mind that speaker design is about compromise, these cabs are all different size, efficiency, bandwidth and cost compromises. All sims are 2.83v @ 8 ohm (1 watt), 4 cabs grouped, 0.25 watt per cab. Punisher (12.00sw) Black, BG horn grey  1850 horn (pd.1850) Black, BG Horn Grey  WSX (V18-1000) black, BG Horn Grey  SBH (PD.1850) Black, BG Horn Grey  TH18 (18sw115-8) Black, BG Horn Grey  |
Replies:
Posted By: KaphaSound
Date Posted: 23 July 2020 at 3:03pm
| Very new to horn design and genuinely just curious, but what would happen if you cut the top sheet in the same style as the “eye shape” in the Danley BC218s and put 4 together in the same eye configuration? Would you be able to slightly increase the length of the horn this way and gain a bit of extra directivity due to the increased size of the ‘face’ of the stack? |
Posted By: citizensc
Date Posted: 24 July 2020 at 2:20pm
|
This is an interesting question, I understand your thinking, I suspect directivity would be improved and you might be right about the slightly increased horn length. This is something I would like to try and measure. I guess I will have to add it to the list of things I would love to do if I wasn't a poor student locked down in an apartment away from my home city. Would be interested to hear from anyone who has used the BC218. |
Posted By: KaphaSound
Date Posted: 25 July 2020 at 4:17pm
| I’d love to help I’m just not 100% confident in hornresp yet, but to get close I assume you’d use 1pi to approximate the large face, then a bit of extra length to get around the final bend at the mouth, but here’s where I get stuck . Would you essentially need to calculate the total horn with 2 or 4 drivers so the combined S1 (S2 for OD) for the compression ratio, then I guess combined S3/4 to keep the expansion rate consistent? Then the total combined mouth size (S5), and then are there any other paramaters like VTC/ATC VRC/LRC that need to double/quadruple according to the number of drivers? I’m really hoping to pick up 4 15s to put under some EV ZLX 15ps as a first project and the B&C BG15 drivers were looking quite appealing. The goal was to make something capable of hitting 30hz, but after some reading I was realizing that’s near impossible in any kind of semi compact/lighterweight format, until I saw your design. So I’d basically love to see if the boundary compliant multiple horns one mouth method could even push f3 a bit lower if only by just a few hz. |
Posted By: citizensc
Date Posted: 02 August 2020 at 10:03am
|
Sorry for the delayed response, been a little bit busy.
I would keep it simple, I would sim it as 4 seperate horns, this is quite easy, just enter it as one horn, click calculate then click tools > multiple speakers > drag the sliders, for 4 horns I would go 2 parallel, 2 series, this way it will be an 8 ohm, this will give you 1 watt with the default 2.83v for Eg. You dont need to do anything fancy with changing chamber volume sizes to simulate multiple horns. The tricky part would be correctly modelling the last part of the horn, I think I know how I would get the correct length for the horn. I would just draw the path in sketchup and measure it. The mouth is a little trickier. Modelling the mouth area of this would be simple, it would just be the cross sectional area (LxW).  But modelling this isn't so straight forward.  The mouth actually looks smaller and I suspect it is smaller but the reduced mouth area compared to the last picture may be acoustically invisible at the frequencies we are talking about. Similar to hanging a compression driver and horn flare in mouth of a kick bin like an xtro. The horn flare may take up 30% of the cross sectional area at the mouth but does not have a significant effect on the output.
Feel free to try sim it in hornresp and post it, will be happy to give my 2 cents. Might even get some input from other people. |
KaphaSound wrote:Posted By: Jo bg
Date Posted: 02 August 2020 at 10:53am
|
The mouth on the BC subs is actually smaller as you noticed, the horn path restricts at mouth before suddendly increasing with the boundary loading. So it's not easy to have a regular (expanding) flh act as a Bc sub when stacked, as there would be no such restriction. The stack will still load better for the bigger mouth and size, but just stacking and a round exit will not make a BC sub.
|
Posted By: KaphaSound
Date Posted: 15 August 2020 at 9:20am
|
Sorry for taking a minute to get back here and I don’t mean to turn this into a BC thread because the original design looks great! I just was also looking at the BG driver and it seems like you really could just add an extra cut to the top sheet, potentially lose a bit of weight and gain a bit of extra F3 on the low end. I actually did end up doing this in hornresp to the best of my knowledge and will post it soon to see what you guys think, but honestly I think the secret sauce to the BC series could be in just the extra surface area and slight extra length from the last bend. If you think about taking a scoop for example notice how you could essentially turn it on its side and gain all that extra width which would bring you closer to 1 pi although of course the mouth would need to be reoriented somehow (and I’m assuming this is why most designs might work best in 4 or more simply because you’re approaching 1 pi space at that point?). Again I’m super new to this but I won’t be surprised if this type of folding becomes more popular. Peep the labda labs DH-18 for example ( https://www.lambda-labs.com/en/products/dh-18 - https://www.lambda-labs.com/en/products/dh-18). They tout the same significant improvements from that particular arrangement but also offer the different stacking options which have different response curves.
Also in terms of necking down at the mouth in the ‘eye’ config, I think I’ve read and definitely noticed in sims that this can help improve sensitivity in the lowest frequencies.
|
Posted By: KaphaSound
Date Posted: 17 August 2020 at 3:03am
Alright so here was the hornresp design. The one thing I noticed is a slight increase in the group delay up to about 33ms but for the improved sensitivity that could still be worth it? Let me know what you think. I basically quadrupled the necessary parameters as I read on the speakerplans site the best way to model a shared mouth horn was to model the entire horn first and then split into 4 sections. I'd really love to hear if there's anything we can do to improve this. I also completely guessed for the extra length/total mouth size so we can dial those in but this is simmed in 2pi so in reality the response might be halfway between 2pi and 1pi.    Also if anyone know of a good way to directly upload images here that would be great. I've read through before but all the suggested websites to grab the image url from are paid only. (Edit: thanks guys)
|
Posted By: DMorison
Date Posted: 17 August 2020 at 2:41pm
|
You can upload images direct to the forum; if replying to an existing thread remember to use the full reply editor accessed by the arrow icon (7th button along in the row of controls immediately above the reply window). That works best if your images are less than IIRC 1000 pixels wide, it does compress them a bit though even when they're smaller to start with. If you don't want that compression or want to preserve larger sizes then off-site hosts are where it's at. I use imgur, which is free and you don't even need to create an account, you can upload directly there and get links to include in your message here. I've seen others use imgBB and postimage well enough too.
|
Posted By: citizensc
Date Posted: 18 August 2020 at 12:48am
I see you have simmed it as a single horn containing 4 drivers, I strongly recommend you simulate it as 4 horns, one driver each. I think this is a more accurate representation of what you are trying to model. It is also much simpler.  The best way to upload images here is to use https://imgur.com/upload?beta" rel="nofollow - https://imgur.com/upload?beta , Its really simple, you don't even need to make an account. Once you have uploaded it there, wait like 5 seconds then right click the image and click 'copy image address' (assuming you use google chrome). Then just paste that address between {img} {/img} but with square brackets. |
Posted By: citizensc
Date Posted: 19 August 2020 at 8:37am
|
Okay, my curiosity got the better of me and I gave it a go. I think these sims are imperfect to say the least, I don't think a boundary extending 700mm out of mouth constitutes 1pi, especially when we are dealing with wavelengths greater than 8 meters in length. I also think its quite hard to capture the geometry of the last segment of the horn with two hornresp segments. The measurement for horn-length for the last segment may be a bit pessimistic, I made an arc from the middle of the previous segment to the middle of the cutout in the cab (see image 1.). Maybe it should be to the far corner of the cab as this would be the centre of the total mouth if you coupled four cabs together (see image 2.) Let me know what you think.   Here is what the path looks like - open to suggestions on how to improve it.   4 x regular BG horn in grey, 4 x BG horn with BC mod in black (2pi)  4 x regular BG horn in grey, 4 x BG horn with BC mod in black (1pi)  I suspect this wouldn't be flat to 30hz with an f3 in the high 20s like the sim suggests but I also suspect you would gain some LF response and sensitivity, I think you would have to build 4 cabs and test it to find out for sure. Let me know if you have any suggestions to improve the sims. |
Posted By: doller
Date Posted: 19 August 2020 at 10:05am
|
Great work you two, very interesting stuff. sorry I can't help with the sims bit dumb on that front sorry. Correct me if I am wrong but isn't the point of the bc system focused bass. I.e. not so omnidirectional on the dance floor so to speak. So how different would a bass beam sound to bass everywhere if you get what I am trying to say. Can't sim that I guess. This would be an expensive experiment. I am tempted to take a jig saw to my labs and see what happens. maybe a router. Did the old Celestion boxes with barn doors use the same concept?
|
Posted By: KaphaSound
Date Posted: 20 August 2020 at 3:41am
|
Man holy cow that’s one sexy looking design! You’re using Sketchup correct? Seriously unreal even just the 2pi output looks like it could have some serious extension. Last I checked there was a sharp change in phase around 80hz or so. Is this important or not at all? Also how was the delay response looking here? I’ll be giving this a go by the end of the year without a doubt and will definitely give some different hornresp charts before committing to a final design. I was hoping to have a slightly wider design potentially to have semi decent stereo separation on some mid tops that would rest on top of the subs. Also I see that you still have some dividers up between the two stacks whereas in Danley’s design the combined mouth has no division (this one looks like two horns next to each other rather than one big combination). Any thoughts on how that might differ? I have no doubt the BC format will lead to some very interesting designs all around and again citizensc thanks for all you’ve done here. Bare with me for the next couple days as work has picked up quite a bit, but I’ll post some more photos and thoughts on this. Maybe one final thought to leave you with is how the curve in the cut of the top sheet lines up perfectly with the mouth flare in the BC218. I could imagine air getting potentially trapped in the current design in the corner right underneath the access panel which would maybe add unnecessary force on the driver (again this is the first plan I’ve contributed to or even loaded into hornresp so take everything I say with a big grain of salt).
|
Posted By: citizensc
Date Posted: 21 August 2020 at 6:57am
I think directivity is a big part of it, I would be surprised if this didn't decrease noise on stage and by doing that it would increase noise on the dance floor. I am not familiar with the Celestion you are talking about but I can imagine they would have a similar effect. It would be an expensive experiment, on my never ending list of projects id like to try but cant due to my current circumstances id love to build another sound system, if I did i would probably try this and measure it. If you do end up cutting your labs up, id love to see the results, my mouth is watering thinking about the graphs, no pressure though, don't do anything you will regret!
Yep, I have tried about 6 different software packages for designing speakers, I keep going back to sketchup because its so simple and fast. It may lack features but I can draw a bass horn in about 10 min if you give me a 2d drawing. I am a rapid iteration kinda guy, rather than get something right the first time I will iterate it about 20 times and zone in on what I think is optimal.
If you are talking about rapid changes from -180 to 180 degrees, this isn't actually a rapid change, it is just the graph wrapping around from bottom to top rather than continuing to extend downwards, it just makes the graph easier to read. See the images below for more detail, I also always post my hornresp inputs so if you are curious about any data I haven't shown, you can always plug the numbers in and have a look. As far as this design goes, I am literally just cutting a chunk out of the standard BG horn, you can see I haven't changed the bracing at all, if you were to actually build this I would cut curves in the bracing and the panels where the cabs meet to make it a bit more 'aesthetic'. After some playing in hornresp, I have decided the hole I cut is too small, 1500cm^2 per BG horn seems to be the point of diminishing returns. Here are some sims with the bigger hole.   BG horn grey, with the BC mod black, both 2pi  BG horn Grey, BC mod black, BG = 2pi, BC mod = 1pi  BG horn Grey, BC mod black, group delay may look bad with the BG mod but in the 40-50hz range its actually reduced and i'm not sure you can actually hear 35ms of group delay at 35hz  BG horn grey, BC mod black, as I said at 80hz that isnt a sudden change in phase, the plot is just going off the bottom of the graph and wrapping around to the top.  Like I have said previously, I think 1pi is a bit optimistic, I suspect reality is some where between 1pi and 2pi. |
Posted By: snowflake
Date Posted: 21 August 2020 at 12:53pm
|
to sim the effect of the four speakers stacked together you can add a final horn section that expands to 4m2 or whatever it is over a very short distance and model it in 2pi. still not exact unless horn mouth is on the floor but close. as you have already used all the hornresp sections you would have to merge some of them together to free up section 4. |
Posted By: citizensc
Date Posted: 21 August 2020 at 1:36pm
I was thinking about this, I have already done some merging to free up a segment for the BC hole thing. I might free up another and try it tomorrow. |
Posted By: bass*en*mass
Date Posted: 24 August 2020 at 10:50pm
|
..if things go well this week i might finally manage to finish the proto extensions of our new custom 18" FLHs.. :) similar to yours, 4x 36hz f3 horns coupled to an extension of more or less unknown tuning.. 250x180cm mouth in our case+, substantial cross section area/volume - curious myself what exactly might happen :) will report back asap! |
Posted By: KaphaSound
Date Posted: 25 August 2020 at 5:45am
| I think David McBean and others have shown that aggregating multiple sections into one has a negligible effect on the total response curve, so I think hornresp for these designs could still be quite accurate. The only thing that isn’t certain is where the response lies between 1 and 2pi. Correct me if I’m wrong but even if a face only reflects 60 percent of a wavelength those gains should still be significant at that frequency shouldn’t they? Or is it an all or nothing scenario? Either way with the current response curve it actually looks like the higher sub frequencies could use a boost for once so these frequencies should definitely see some improved sensitivity from the size of the face. Also citizensc is there a reason you are using conical segments? I’ve definitely also read that parabolic is better at simming folded horns due to two walls remaining constant/ not expanding (until the end of this particular design at least). |
Posted By: citizensc
Date Posted: 25 August 2020 at 11:29am
Sounds interesting, you should start a thread for the project, id love to read it! And I would love to what impact the extensions have. As far as simming this in hornresp goes, I will sim it in AKABAK 3 and report back when I get time. I think this will give a far more realistic result. |
Posted By: citizensc
Date Posted: 03 September 2020 at 4:31am
|
I have moved the BC conversation to another thread, just to keep this thread a little more focused on the BG horn. https://forum.speakerplans.com/bc-horns-from-standard-flhs_topic105665.html" rel="nofollow - https://forum.speakerplans.com/bc-horns-from-standard-flhs_topic105665.html And for those that are interested, here is how the BG horn compares to the Punisher in AKABAK 3. Please keep in mind that the Punisher is a more compact cab. Both are 1.41v @ 2 ohm, 4 horns coupled in 2 pi measured from 1m. Punisher in red, BG in blue 
|
Posted By: dnbpc
Date Posted: 04 February 2021 at 4:06pm
|
Has anyone built a prototype of this design yet? Would be interested in hearing the real life results as eyeing it up for a future build. ------------- Pick & Mix Soundsystem |
Posted By: citizensc
Date Posted: 05 February 2021 at 7:33am
I've had multiple people people express interest in building the BG horn but I don't know of anyone following through. I am pretty confident it would make a lot of bass when used correctly (in multiples) as this has been shown in both conventional sims and BEM sims which work very differently. It's also unlikely to be a driver killer due to lowish compression and a tough driver with a heavy cone. The only thing I cant be completely confident about is the tonal characteristic, id imagine it would have a relatively warm low end due to using a driver with a qts above 0.4. I also speculate that it would be quite punchy simply out of virtue of being a front loaded horn. If you do end up building some, don't hesitate to hit me up, I'm happy to help with any issues you might encounter in the build. |
Posted By: dnbpc
Date Posted: 04 March 2021 at 7:21pm
|
Thanks I’m still thinking it over at the minute, the thought of events coming back has kicked me into gear again. Did I read on the earlier thread the the b&c 15ps100 was a potential driver or did I imagine it? I’ve got two sat around so may be a good fit. ------------- Pick & Mix Soundsystem |
Posted By: citizensc
Date Posted: 05 March 2021 at 1:38am
|
The 15PS100 could work well. You wont hit the SPL of a 15BG100 as it only has 6.5mm of Xvar compared to the 14mm of the 15BG100 but it should still give you a good idea of what they sound like. Here are some AKABAK 3 sims of some different drivers in the BG horn, please keep in mind these are on axis measurements, not power response measurements like hornresp outputs. I added in the 15SW115-8 to show why putting a driver with lowish Qts in this horn may be a bad idea. I also have no idea if any of these drivers will actually fit in the chamber, will leave that to others to figure out. All measurements are 4 cabs with 1.41v @ 2ohm measured from 10m, scaled to 1m in 2pi
 |