Post by: Jared Scott on December 03, 2008, 03:47:45 AM
I have stuff available in both CAD viewable files and jpegs, as well as the driver specs.
Post by: Art Welter on December 03, 2008, 12:46:21 PM
Post it up! JPEG works for me.
Art Welter
Post by: Iain_Macdonald on December 03, 2008, 12:50:07 PM
Have you modeled it in Horn Response? If you have, please post a jpg of the input parameters page. Lots of people use it, and you are likely to get a reply.
Iain
Post by: Jared Scott on December 03, 2008, 04:43:38 PM
External cabinet dimentions are 45"x45"x20"
Estimated program power rating will be 2000-3000 watts rms.
Driver: http://www.diycable.com/main/product_info.php?cPath=24_93_15 0&products_id=681
Modeled number, impedence, phase, and group delay
Wire-frame side view
3-D model
Predicted output for 1, 2, 4, and 8 horns at 2.83 volts
Predicted output for 1, 2, 4, and 8 horns at 2000 watts rms per cabinet
Post by: Marjan Milosevic on December 03, 2008, 05:52:41 PM
First of all you have to learn which drivers are suitable for horn loading.
That one is not!
That driver is more suitable for a car subwoofer or a home cinema.
It has too low Fs, too high Qts and too low BL and way too low db/1m.
Also i dont think it will survive the compression made inside the horn and will tear apart in mater of hours (if not minutes) with 2000W applied to it.
As for the box it self. Your idea in general is ok but the firs part of the throat is too complicated for build. That is totally unnecessary.
For 360 US$ you can find much better drivers that are made for horn loading.
What is you final goal with this?
Post by: Marjan Milosevic on December 03, 2008, 05:54:06 PM
How did you come up with a same number for ATC and VTC?
Post by: Jared Scott on December 03, 2008, 07:12:06 PM
As for the ATC/VTC thing, my picture doesn't include one part that would be inside the throat chamber (too hard to see it inserted in the model). With the piece installed, it would provide 1182 square centimeters of chamber area. And I designed the chamber so that when the driver is at peak outward excursion there is 1182 cubic centimeters between the driver and the chamber walls. I'm guessing this was an incorrect assumption on how to calculate those numbers. And just curious as to why it would tear itself apart so quickly? The construction materials are also almost identical to the LAB12 driver.
As for the driver, I selected it because of it's similarities to the LAB12 driver. I assumed the Fs wasn't the best. But the Qts is almost the same as the LAB12 driver (0.40 vs 0.38), it's got 3.6T-M more BL than the LAB12, and only 0.8 dB less sensitivity. To me, it looked quite capable in comparison to a driver that I know performs very well in horn loaded situations.
As for the complexity of the enclosure, I designed it with ideal curvature/angles. To build it, I would obviously modify the design to use flat panels instead of so many complex curves.
This isn't meant to be argumentative, just an explanation of where my thinking was coming from.
Post by: Elliot Thompson on December 03, 2008, 09:52:51 PM
| Jared Borchardt wrote on Thu, 04 December 2008 00:12 |
As for the driver, I selected it because of it's similarities to the LAB12 driver. I assumed the Fs wasn't the best. But the Qts is almost the same as the LAB12 driver (0.40 vs 0.38), it's got 3.6T-M more BL than the LAB12, and only 0.8 dB less sensitivity. To me, it looked quite capable in comparison to a driver that I know performs very well in horn loaded situations. |
The size of the driver makes a huge difference when you are comparing QTS numbers.
For an 18-inch driver it is a QTS ranging from 0.2 - 0.3 that is needed for horn loading. The Maelstrom is 0.4
The Tesla meters should also be around 35 - 24 whereas, the Maelstrom is only 18. That is 1 TM lower than the JBL 2241. That is not a large motor for an 18-inch driver in terms of Horn-Loading. However, 14.9 TM is huge for a 12-inch woofer. Remember, you need a strong motor for you firing through a small opening in which a lot of air pressure build up will occur between the cone and the wood or should I say compression chamber.
If you are in need of an 18-inch driver that is focused on horn loading, you can visit www.speakerplans.com for that's all the designers focus on (18 inch drivers/Folded Horn bass cabinets) over there. And they know lots of them.
I was actually looking for an old Lab 12 response chart so you can compare your design to the Lab Sub since both designs share the same outer dimensions. Maybe a Lab Sub user can compare their simulation to your design.
Best Regards,
Post by: Art Welter on December 04, 2008, 01:33:55 AM
The first throat area “funny looking” part of your horn will end up being a band pass lossy thingy, rather than extending your horn length. Eliminate all that rounded stuff and mount your speaker to the first portion of the horn, and extend that to the front right of the cabinet internally.
Your speaker cone should be as close to the upper right hand side of the cabinet as the magnet structure will allow. You may want to have the horn go a little Hypex at the throat end, particularly if you are only planning to use one or two.
Having tried a throat area like you drew to get a little more length, after comparing the much simpler to build, slightly shorter path length, the shorter version had more low end output.
No curved corners are needed, flat panels approximating the curves will be good.
Post by: Jared Scott on December 04, 2008, 06:24:55 PM
Would an 18" driver with these numbers be a better fit for horn loading?:
Fs: 30.8
Qts: .28
BL: 27.1
SPL: 97.3dB
Post by: Marjan Milosevic on December 04, 2008, 06:59:52 PM
Post by: Jared Scott on December 04, 2008, 07:42:24 PM
I'll run some numbers, model it up, and post what I get back here to see if I'm making progress with this stuff.
I appreciate the help guys.
Post by: Elliot Thompson on December 04, 2008, 09:27:28 PM
| Jared Borchardt wrote on Thu, 04 December 2008 23:24 |
Lets see if I'm heading in the right direction. Would an 18" driver with these numbers be a better fit for horn loading?: Fs: 30.8 Qts: .28 BL: 27.1 SPL: 97.3dB |
That is a good starting point. You need to have various drivers so you can compare. Going closer to 0.2 than 0.3 will offer more mid-bass (80 Hertz - Up) while, closer to 0.3 while offer more bass (80 Hertz - Down)
Keep in mind horns are more finicky to drivers than a reflex bin when it comes down to performance. You may want to look more on the European market for they specialize in horn-loaded design eighteens more than American manufactures.
Best Regards,
Post by: Marjan Milosevic on December 05, 2008, 07:18:20 AM
But you have access to 18Sound, B&C, RCF, P-Audio... that also has good drivers for horns.
My bigger concern is that Jared is not entirely aware of the physical size of that box.
That is a Lab sub dimensions box.
Here is one picture that will give an idea of how that box will be.
That box will also be heavy so i will suggest to go with a neo magnet driver.
Regards
Marjan
Post by: Jared Scott on December 05, 2008, 07:33:19 AM
Part of the reason my design was so complex is because I intended on making the internal parts out of a fiberglass/plywood laminate that I can make and form to any shape I desire. The idea was to save a bit of weight using the fiberglass instead of plywood inside. My idea was to maintain truck-pack dimentions on the height and depth for transport, but it is 2.5" narrower as well.
Post by: Casey McDonald on December 05, 2008, 11:14:45 AM
I play around with hornresp a fair bit and always get massive peaks and dips (esp 2pi), and have trouble getting my head around the fact hornresp says a box itself can have nearly 0.1s of delay within itself at certain frequencies.
I've done some searching around and found a rule of thumb that says 1/f is the point at which it becomes audible. I've built horn subs based on hornresp frequency response alone (before I found out / knew about all the other graphs) and they seem to sound okay but the GD plots are horrible, so how much does it really matter? i.e. when designing these should one pay as much attention to the GD plot as to say, frequency response?
To keep my post on track with the thread - yeah, seems the OP's selected driver is also a heavy mms...
Post by: Marjan Milosevic on December 05, 2008, 11:19:11 AM
i think you should be looking in a folding like in the punisher. See the picture.
You dont need to complicate anything. For low frequencies it would not have any effect on the response.
Post by: Marjan Milosevic on December 05, 2008, 11:26:14 AM
Post by: Elliot Thompson on December 05, 2008, 03:51:20 PM
| Marjan Milosevic(MarjanM) wrote on Fri, 05 December 2008 12:18 |
Elliot, that is surely true if you think PD and Void. But you have access to 18Sound, B&C, RCF, P-Audio... that also has good drivers for horns. |
I wasn't talking about availability but more on a manufacturing standpoint.
18Sound, RCF and, B&C are European drivers. P Audio is Asian. There are lots of companies not widely used like BMS and, Faital Pro that offer 18 inch horn drivers as well available in the States. The only European Manufacture that does not focus on horn-loaded designs is Beyma.
I agree on the overall size.
One of the things many assume is bigger speaker = bigger/better bass. Well, from a horn's concept you need a huge cabinet to not only compensate the driver, but the horn as well achieve the low 40 - 30's that is easier with a reflex using a smaller thumbprint.
I remember reading someone building a single 30-Hertz straight horn using an 18-inch driver and it was bigger than his house.
We can learn a lot from just looking at a compression driver versus horn ratio in terms of overall size. Especially, those that are capable in achieving 500 - 300 Hz full-space.
Best Regards,
Post by: Art Welter on December 05, 2008, 10:56:37 PM
You said:
"Especially, those that are capable in achieving 500 - 300 Hz full-space"
Don't know what you were going for, but 200 Hz bandwidth was probably not it..
Post by: Elliot Thompson on December 06, 2008, 01:45:49 PM
| Art Welter wrote on Sat, 06 December 2008 03:56 |
Elliot, You said: "Especially, those that are capable in achieving 500 - 300 Hz full-space" Don't know what you were going for, but 200 Hz bandwidth was probably not it.. |
I wasn't going for anything. My comments you quoted was based on how large of a horn is needed to provide the frequency and SPL using a compression driver when going down to 5 - 300 Hertz at full space.
http://www.lansingheritage.org/images/jbl/catalogs/1971-pro/ page10.jpg
If the horn's mouth is not big enough to play the desired frequency at a high sound pressure level, it won't do it. That is why you need multiple horns grouped together to create one big horn if you are aiming for high SPL bass ranging from 40 Hz - down.
Best Regards,
Best Regards,
Post by: Tim McCulloch on December 06, 2008, 03:49:12 PM
Art was referring to the typo in your post (500 - 300).
Have fun, good luck.
Tim Mc
Post by: Elliot Thompson on December 07, 2008, 10:10:33 PM
| Tim McCulloch wrote on Sat, 06 December 2008 20:49 |
Hi Elliot- Art was referring to the typo in your post (500 - 300). Have fun, good luck. Tim Mc |
Thanks Tim.
Best Regards,
Post by: Art Welter on December 07, 2008, 11:34:00 PM
5, 500, whatever it takes...
Post by: Peter Morris on December 08, 2008, 12:16:36 AM
- Your design looks fundamentally fine.
- I would use a lower compression ratio so there is less stress on the cone. The 3:1 you have used is getting up there for “standard” cone materials and designs.
- You don’t need all of those fancy curves for low frequency horns.
- I would select a different driver eg. 18sound 18WL1400, RCF LF18X400, B&C (what ever the equivalent is)
- In most situations the speaker would be used on the ground so I would use 1/2 space loading ( pi) when I model it.
- If you are only going to build 2 of these boxes I would build some good double 18 reflex cabs instead. If it was only 4 cabinets I would shorten the horn to about 8 feet.
- Use a lot of bracing – there is not enough in the LAB.
Post by: Jared Scott on December 08, 2008, 08:40:38 AM
| Peter Morris wrote on Sun, 07 December 2008 23:16 |
General comments - Your design looks fundamentally fine. - I would use a lower compression ratio so there is less stress on the cone. The 3:1 you have used is getting up there for “standard” cone materials and designs. - You don’t need all of those fancy curves for low frequency horns. - I would select a different driver eg. 18sound 18WL1400, RCF LF18X400, B&C (what ever the equivalent is) - In most situations the speaker would be used on the ground so I would use 1/2 space loading ( pi) when I model it. - If you are only going to build 2 of these boxes I would build some good double 18 reflex cabs instead. If it was only 4 cabinets I would shorten the horn to about 8 feet. - Use a lot of bracing – there is not enough in the LAB. |
I am currently re-doing the design with a different more horn friendly design.
The complex curves are gone. The horn bends are still smooth curves, but those will be simple to build.
It is designed as half space loading, pi is quarter space loading.
If I ever actually build it (as stated before, this is more gears toward a learning expreience than actual build project), I plan on building 8 of them.
Post by: Jared Scott on December 08, 2008, 01:51:02 PM
This one measures 45x45x18" external dimentions
It was modeled with this driver: http://pasub.com/PASUB390.pdf
I understand that this driver is not available yet, just experimenting.
Horn numbers
Cross section view
3-D View
1, 2, 4, and 8 cabinets at 2.83 volts
1, 2, 4, and 8 cabinets at program power
Post by: Johan Diettrich on December 08, 2008, 03:10:41 PM
Group delay figures from Hornresp are difficult interpretted, exporting as an AkAbak script gives easier digestible gd figures.
Those corner deflextors aren't strictly nessecary for sub usage.
Best regards Johan
Post by: Peter Morris on December 09, 2008, 08:05:04 PM
I would have a look at what happens if you shorten the horn a bit. You may gain more useful output than you loose in LF extension.
FWIW have a look at these links.
http://www.billfitzmaurice.com/T48.html
http://billfitzmaurice.net/phpBB3/viewforum.php?f=12&sid =34588dfcfa53949660a2d51ad54170e1
http://billfitzmaurice.net/phpBB3/viewtopic.php?f=12&t=4 293
http://billfitzmaurice.net/phpBB3/viewtopic.php?f=1&t=71 63