"Compare it to this, an 18" in a DSL box with a 22.5" exterior height"
Hence, the mouth = 21" wide x 28.5" high.
Hence, the mouth = 21" wide x 28.5" high.
The mouth of that DSL box is 45" x 22.5".
It is not the BC218, which has a mouth of 60" x 30".
The OP plans to make two single 18" boxes with a 30"x30" mouth to equal the BC218.
This image of a DSL DBH218 FLH throat using an 18" driver in a 22.5" (up/down dimension) has an S2 of ~2" x21", 42 square inches.
42 square inches does not include VTC:
The BC218 up/down dimension is 28.5", an S2 of 42 square inches would be 1.47" x 28.5".
42 square inches =270.97 square centimeters.
Proceed...
Art
My post didn't help, by not mentioning the second dimension must be estimated for a cabinet which has a different width.
Ricci's Othorn also uses high compression ratio, yes it's tapped horn and 21SW152 driver, but I believe maybe 18Sw115 would withstand even higher pressures since driver is smaller. Maybe I should try with 1:4 compression ratio with cone volume and 18mm plywood thickness also used in S2 calculation? So all together adds up to about 300cm2? For volume calculation I will use Solidworks to model cross section as I have accurate 3d model of 18SW115
TH and FLH are different, no back chamber in a TH.
With a sealed rear chamber, you can play around with reactance annulling by changing VRC (Volume of Rear Chamber).
https://kolbrek.hornspeakersystems.info/index.php/horns/my-approach-to-bass-horn-design
The 18SW115 has a slightly higher displacement to Mms ratio than the 21SW152, so probably can withstand slightly higher compression ratios.
That said, the Othorn's 1/4.2 compression ratio (S2/Sd) has been no problem at voltage levels that exceed Xmax in Hornresp simulations.
The Gjallerhorn has a similar compression ratio and the LMS Ultra 5400 (38mm Xmax) used in more than double the displacement of the 18SW115 , and no problems either.
You should enter an accurate VTC figure (around 4000), the cone cavity volume and 18mm plywood thickness.
Experiment with various compression ratios, (and VRC) probably don't bother with much over 1/5.
If the response gets too "droopy" above the low corner, the compression ratio is probably lower than in the BC218.
Last edited:
All_1 and I discussed "backstage" the possible layout for this subwoofer and how simulations turn out. The following is an update of our current research 🙂.
First try, variant 1, looks like the horn-fold we discussed already in this topic:
Hornresp input:
The simulations show decent results, but not the low frequency extension which is shown in the Danley datasheets. The 3D DWG on Danley's webpage showed a specific cut-out in a sheet which verified that the horn length must be longer than variant 1. Second layout, variant 2:
Hornresp input:
The simulations for this longer horn show the low frequency extension, but of course, efficiency at 25-35Hz is reduced.
I used ANSYS to get the acoustic impedance for 3 different setups (A, B and C), see below, and to get the transfer function between diaphragm velocity and pressure at different microphone positions. These results are imported in Matlab so we can calculate SPL, cone excursion etc.
For each configuration variant 1 and 2 are simulated, resulting in the following SPL plots (10m microphone distance, 28.3V RMS generator).
Only V1:
Including V2:
Cool to see the effect of the "Half moon" configuration, which gives you an additional 1-3dB for free.
First try, variant 1, looks like the horn-fold we discussed already in this topic:
Hornresp input:
The simulations show decent results, but not the low frequency extension which is shown in the Danley datasheets. The 3D DWG on Danley's webpage showed a specific cut-out in a sheet which verified that the horn length must be longer than variant 1. Second layout, variant 2:
Hornresp input:
The simulations for this longer horn show the low frequency extension, but of course, efficiency at 25-35Hz is reduced.
I used ANSYS to get the acoustic impedance for 3 different setups (A, B and C), see below, and to get the transfer function between diaphragm velocity and pressure at different microphone positions. These results are imported in Matlab so we can calculate SPL, cone excursion etc.
For each configuration variant 1 and 2 are simulated, resulting in the following SPL plots (10m microphone distance, 28.3V RMS generator).
Only V1:
Including V2:
Cool to see the effect of the "Half moon" configuration, which gives you an additional 1-3dB for free.
Yeah, that's some cool sim work!
The low end corner still looks like it might be a few Hz above the DSL BC218, and does not have the "saddle" depression between ~28Hz to 40Hz, indicating your horn path is a little shorter.
That said, half the BC218 interior horn width is ~74.3 centimeters, yours is 71.4 (1.14" less), and you cut the cabinet depth from 60", 159 centimeters to 152.4 centimeters, 59", so it would be a few inches shorter even if the horn fold was identical.
Any way you cut it, always a trade in output level for LF extension, I'd probably go for variant 1 for less parts, ease of build, and more 30-40Hz output.
I'd size dimensions for whatever the actual 5'x5' (1525mm?) Baltic Birch dimensions are.
Art
My thought was it has something to do with the overall length of the horn, and this could indeed be limited by the smaller overall cabinet size. However, to bring the saddle behavior into the response, and a more pronounced peak at ~28Hz, the rear chamber volume has to be decreased (I think, based on some tweaking in Hornresp). This brings the first impedance peak of the impedance curve close to the one in Danley's datasheet.
Next version will be a V2.1 with smaller rear chamber volume and see how this turns out. Horn length is what it is for now.
Next version will be a V2.1 with smaller rear chamber volume and see how this turns out. Horn length is what it is for now.
Rear chamber volume changed from 100L (V2) to 80L (V2.1), simulated flat (B) and half moon (C):
Very close I would say!
Very close I would say!
Maybe, but I think there is no more horn length in the BC218 (except for the couple inches because of the size of the enclosure) than we have now in this design.
For now my work is done 🙂, because All_1 is going for variant 1, am I correct?
Last edited:
I believe I will go for variant one, yes. It has a bit better sensitivity compared to V2 at 30hz and up, and it's easier to build. I am definitely preferring increased sensitivity over a bit of low end extension, Will build 4 cabinets with 100L rear chamber. Will be interesting to see how much power compression will closed rear chamber bring, Right now I am using 18SW115 in tapped horns and of course they get excellent cooling,
Will post thread updated with 3D model design and then build pics
Will post thread updated with 3D model design and then build pics
Making the driver access panel an aluminum heat sink like Cerwin Vega and B-52(used to) use on their FLH would improve the heat transfer to the outside air.
That said, the heat transfer of the B&C 18SW115 is a lot better than the drivers those cabinets use, and those FLH have smaller compression chambers.
I'm a big fan of the large frontal area concept, looking forward to seeing your build!
Art
I agree with you, I believe with 100L rear chamber would be better heat transfer since it has more surface area and also more volume than 80L version, that was part of the reason why I went with V1. What about making plates from 6mm aluminium maybe? Would it need additional bracing or it should be fine?
Also do you have an idea how much power should I put into 18SW115, I used them in bass refex and tapped horn with 145V peak limiter (8ohm version) with no issues running 10+ hours quite hard sometimes.
Also do you have an idea how much power should I put into 18SW115, I used them in bass refex and tapped horn with 145V peak limiter (8ohm version) with no issues running 10+ hours quite hard sometimes.
For the size needed for an access panel 6mm would probably be stiff enough with no additional bracing.
I don't know if it would "ring", but adding a few plywood braces would take care of that if it was a problem.
You must be running quite dynamic music!
My recommendations are more conservative, bottom of the OP, substitute your 8 ohm specs for the 4 ohm:
https://www.diyaudio.com/community/...g-18-15-12-inch-speakers.185588/#post-2512872
Here are some updates on project:
Any comments on the bracing of cabinet?
Still have some details to do, happy to hear your recommendations and ideas for bracing since I believe for this cabinet is really needed 😉
Also @weltersys thank you for info about 18SW115 power handling. Currently also working on simple system to monitor temperatures of driver when used at long parties,
- Currently finishing my 3d model (adding bracing)
- Ordered plywood 3000x1500 11 sheets needed for 4 cabinets with some leftover
- CNC cutting next week
Any comments on the bracing of cabinet?
Still have some details to do, happy to hear your recommendations and ideas for bracing since I believe for this cabinet is really needed 😉
Also @weltersys thank you for info about 18SW115 power handling. Currently also working on simple system to monitor temperatures of driver when used at long parties,
Does Hornresp do a simulation of step response - I'd be fascinated to see what such a long tract does to dynamics.
Most interesting sub project I've seen in a long time.
THX!
THX!