Announcement

Collapse
No announcement yet.

A Design for Line Arrays with a 3D Printed Enclosure

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • A Design for Line Arrays with a 3D Printed Enclosure

    I've been lurking on this forum off and on for a while and have recently been inspired to take the plunge and build some speakers. I hadn’t quite found a project that made we want to dive right in and build something before. Recently learning about the 3D printing equipment available to me at work has rekindled my desire to make something interesting and I’ve decided to see how I can incorporate this into a set of speakers. I'm posting here as I would love to hear what everyone thinks of this plan and if there any suggested improvements/changes/etc. I’m quite a novice when it comes to speaker building and I’m sure there is plenty that I have overlooked.

    Project Goals
    To create a set of slim wall-mounted line-array speakers to use as the left and right channels of my home theater setup while using the 3D printing equipment available to me to make the cabinets. I would like these speakers to be able to play at a reasonably loud level sufficient for an average home theater. Since I will be using a subwoofer, they will not need to play particularly low, however. I have been fascinated with line arrays and am kind of nervous about designing a good crossover, so I decided to come up with a sort of lesser version of the CBT24K which uses full range drivers.

    Driver Selection
    I took a look at a few smallish full-range drivers and landed on the Dayton Audio ND-64-4 2.5” that is used in the CBT24Ks. This driver looks to offer everything I would want in a full range and can reach fairly low in the small cabinets that this project dictates. I have also decided to use the matching Dayton ND65-PR Passive Radiators as fitting a port in these cabinets would be tough. I plan on using 6 drivers, wired series/parallel for a 6 ohm load, along with 6 of the radiators, each in their own enclosure, per side.

    Enclosure Design and Material
    After some research, I found a design for a modular 3D printed soundbar on Thingiverse. This design consists of individual enclosures that can be stacked. I believe that this could work well with a few modifications. I plan to increase the overall size of the enclosure with my initial designs having a volume of about .006 ft3. I also plan on increasing the thickness of the walls to at least 5 mm or maybe more. In addition, I will probably look to stack a few of the enclosures on top of each other so they can be printed in one piece, which means I will have less enclosures to join down the road. As far as what material I will be printing these with is concerned, I have decided to go with wood filled PLA plastic. I haven’t decided on the specific type I will be using yet, but wood filled PLA usually consists of around 30% wood fibers mixed in with the plastic. This material can be sanded and stained and should be dense enough to help tamp down on any unwanted vibrations.


    With all of that said, what do you think? Am I off my rocker? Would I be better off building a 2 way or or some other established design? Your feedback on this is greatly appreciated!
    Last edited by 4kn3jansent; 10-12-2019, 11:19 AM. Reason: Edited to make clear that the plan is for 12 enclosures (6 drivers, 6 radiators) per side.

  • #2
    I think this sounds like a cool way to build some speakers. I'm not a whole lotta help, but have a few thoughts.

    5mm wall thickness seems pretty thin, and maybe subject to resonance, but then again, I'm not familiar with the properties of your material. Just a thought - take a look at the project in the link. A totally different construction technique, but there's some similarities to what you're proposing that might be applicable. Namely, the internal structure of the cabinets looks like it's designed to break up reflections inside the enclosure (and it would add some "meat" to the walls.) Also, the front baffle corners are rounded to reduce diffraction. It seems like both of these concepts could be implemented with 3D printing pretty easily.

    https://www.diyaudio.com/forums/full...ine-array.html

    Comment


    • #3
      Resonance is definitely one of the things I'm worried about with this design. I can go thicker than 5mm for the enclosure walls which will probably help, although I will have to look into the impacts this would have on the amount of printing material I would need. I've also been looking into the many different printing materials available. My initial assumption was that one of the types of wood filled PLA composite would be the best for resonance, but there are other options out there like carbon fiber filled, metal filled, glass filled, or plastics other than PLA. I'll definitely be doing more research in this area to see what is best while balancing cost effectiveness. Would it be a safe assumption that I should be printing with the densest and least flexible material possible? Are there other material properties I should be looking for?

      Also, thanks for the link and shape suggestions. I'm sure I can work rounded baffle edges and irregular internal structure into this design. If I'm going to 3D print these I might as well take advantage of the design flexibility it gives me.

      Comment


      • #4
        Generally the passive radiator area exceeds the driver area in the designs I've seen.

        Comment


        • #5
          Based on the driver areas, you would need about ten of the ND-65 drivers to match the surface area of a 6.5" driver. So, if you also consider x-max, about 20 per speaker would have SPL output comparable to a more typical 6.5" 2-way. I don't think that's very cost effective. In addition, the drivers will struggle to reach the 80hz x-over to a sub, unless you plan to EQ them.

          Comment


          • #6
            I think you guys misinterpreted his speaker selection. He's using a 2.5" driver and 2.5" passive per enclosure. Total of 6 enclosures per side. At least that's how I see it.
            A mains
            The Ventures
            Open Invit8tions
            RSR
            Sound Troopers
            Acorns
            442
            DGBG's
            The Monuments

            Comment


            • #7
              "...I can go thicker than 5mm for the enclosure walls which will probably help, although I will have to look into..."

              This is an interesting project, it's good of your employer to encourage it.

              If the printer produces a honeycomb core you may be able to increase the core thickness and rigidity without increasing the material use and print time too unreasonably. Have you checked the estimated print time and material required?

              Comment


              • #8
                I also see 6 ND64s, 2 paralleled series strings of 3 each.
                Modeling using 6 in 0.036cf, tuning w/6 PRs (no added mass) yields an F3 of around 160Hz (Fb near 130, which looks about "optimum").
                About 108dB? at 80wRMS (or is it only 101?).

                Comment


                • #9
                  Originally posted by 6thplanet View Post
                  I think you guys misinterpreted his speaker selection. He's using a 2.5" driver and 2.5" passive per enclosure. Total of 6 enclosures per side. At least that's how I see it.
                  I was definitely not very clear in this part of the write up. The plan is to have each single enclosure house either a driver or a passive radiator. I plan on having 6 drivers and 6 radiators for a total of 12 enclosures per side. Per my initial calculations this would give me the enclosure volume that I need to use the radiators.

                  Comment


                  • #10
                    Originally posted by djg View Post
                    Generally the passive radiator area exceeds the driver area in the designs I've seen.
                    I chose the number of passive radiators based on the general rule I've seen that radiators need to have at least double the displacement of air (Vd) of the matching drivers. The drivers have a Vd of 29.64 with 6 of them combined having a Vd of 177.84. The radiators have a much higher Vd of 78 due to their larger Xmax, with the total of 6 radiators coming out to 468. This lead me to believe that 6 radiators would work well. Would it be wise for me to work more radiators into the design?

                    Comment


                    • #11
                      Originally posted by rpb View Post
                      Based on the driver areas, you would need about ten of the ND-65 drivers to match the surface area of a 6.5" driver. So, if you also consider x-max, about 20 per speaker would have SPL output comparable to a more typical 6.5" 2-way. I don't think that's very cost effective. In addition, the drivers will struggle to reach the 80hz x-over to a sub, unless you plan to EQ them.
                      I do plan on EQing them. Since I'm new to speaker building I had considered 2 way design, but I really wanted something that could fit into the long and thin form factor. I also have a bit of a desire to do something a little bit different. If it ends up not working as well as I'd hoped it would still have been an interesting project that helped me learn a bit more about 3D printing.

                      Comment


                      • #12
                        Originally posted by TN Allen View Post
                        "...I can go thicker than 5mm for the enclosure walls which will probably help, although I will have to look into..."

                        This is an interesting project, it's good of your employer to encourage it.

                        If the printer produces a honeycomb core you may be able to increase the core thickness and rigidity without increasing the material use and print time too unreasonably. Have you checked the estimated print time and material required?
                        I am very lucky with what I have available to me at work. I work for a technical college that is a part of the United States Fab Lab Network. This means that they have a lab with 3D Printers, CNC machines, Milling Machines, 3D Scanners, laser engravers, and computer controlled vinyl cutters all available for student, staff, and public use. As far as the walls go, I believe I can produce a honeycomb core, but I'm a bit worried that having voids within the walls could contribute to resonance. I have not figured out the print time and material needed quite yet. Once I have finalized what I would like the design to be I will definitely be getting to that.

                        Comment


                        • #13
                          Originally posted by Chris Roemer View Post
                          I also see 6 ND64s, 2 paralleled series strings of 3 each.
                          Modeling using 6 in 0.036cf, tuning w/6 PRs (no added mass) yields an F3 of around 160Hz (Fb near 130, which looks about "optimum").
                          About 108dB? at 80wRMS (or is it only 101?).
                          Thanks for the modeling. The way I have things drawn up currently I would have 0.072cf as the radiators would each have their own enclosure. How would that look?

                          Comment


                          • #14
                            Originally posted by 4kn3jansent View Post

                            I am very lucky with what I have available to me at work. I work for a technical college that is a part of the United States Fab Lab Network. This means that they have a lab with 3D Printers, CNC machines, Milling Machines, 3D Scanners, laser engravers, and computer controlled vinyl cutters all available for student, staff, and public use. As far as the walls go, I believe I can produce a honeycomb core, but I'm a bit worried that having voids within the walls could contribute to resonance. I have not figured out the print time and material needed quite yet. Once I have finalized what I would like the design to be I will definitely be getting to that.
                            You are fortunate. I'll look up the Fab Lab Network, sounds interesting.

                            A honeycomb core is the default on the printers I've used. You can design a thicker wall and probably will have a choice of density. Perhaps you have access to FEA software to analyze the rigidity of the wall for different thicknesses, and can design for the optimum?

                            I'm curious about the software you use for design.

                            Comment


                            • #15
                              Originally posted by 4kn3jansent View Post

                              I was definitely not very clear in this part of the write up. The plan is to have each single enclosure house either a driver or a passive radiator. I plan on having 6 drivers and 6 radiators for a total of 12 enclosures per side. Per my initial calculations this would give me the enclosure volume that I need to use the radiators.
                              The Sd of the driver or passive, is 15.6. So 12 of these drivers would have an Sd of 187.2 centimeters square. A 6.5" Peerless HDS driver has an Sd of 143.

                              Lets look at swept volume of the two at x-max. The array would be 187.2 x 1.9= about 374. The 6.5" would be 143 x 5.3= about 757. Roughly twice that of the array.

                              Maybe this is not a good way to compare potential output, but the bass has to be generated by displacement of the driver, or the passive radiator.

                              Question. Are the passive radiators going on the side, or between the ND-64 drivers?

                              Comment

                              Working...
                              X