Announcement

Collapse
No announcement yet.

PC Speakers (T Line Box)

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

  • #31
    Originally posted by Chris Roemer View Post
    Yeah ... part of the FUN of all this speaker design/building hobby is running simulations. WinISD ( from linearteam.org ) uses T/S (Thiele/Small) parameters (mainly Qts, Fs, and Vas) to determine the proper box style (closed or ported), internal volume, and tuning (if vented). The "designers" on here don't just take a guess at what might work (although that kinda WAS S.O.P. back 40 yrs ago), it uses up too much time, material, and money. A box sim tells you how low your selected woofer will be able to play, and a crossover simulator helps in the design of a crossover (mostly for 2-ways and 3-ways). It's often POSSIBLE to run such an accurate sim that if you build it, then measure it, it ends up running within a dB or 2 of your design sim.
    Ah ok....I got some spreadsheet which name elludes me at the moment, where I think it was geared more for subs, but asked for information like le2 and re2 I think, which wasn't on a spec sheet. I'll see if I can grab a copy of winsd and have a go, but think some of this may be over my head, maybe I'm over simplying this but was half expecting to see just volume requirement and maybe a length of port...I have no idea how to tune anything as well...might have a date with Dr Google and search some pre done designs.

    Comment


    • #32
      Chris, i had a go setting up the tectronics in WiniSD and to be honest looks like there is a bit of data that one who knows what they are doing would generally overlook.

      so i ran multiple options with the vented box and closed similar volumes to that on PE's site, which was done in BassBox.

      The box size for the option with a passive radiator seemed quite large at 1.7ft3, The vented option went off the charts lol but was 0.15ft3. Not sure really what to look for in the graph, but im gathering we want to avoid the spikes and dips shown in green and maybe either the two blues which are both sealed, just altering the "ripple response" which honestly i dont know what that actually is.

      Do i get brownie points for trying? which are probably cancelled out with my dumb response.....

      so i short have no idea what i'm looking at, no what is actually a good curve.

      Comment


      • #33
        There are 2 designs I know of using the Tectonics woofer you are using. One, the BMR-3L uses a 3 liter box, but should use a 3.5 liter box so port could go inside. That design is discussed here: http://techtalk.parts-express.com/fo...nic-tebm65c20f

        The other, the Mariposas is detailed here: http://techtalk.parts-express.com/fo...fa-hibm65c20-8

        Both are designs from individuals who know much, much more than I do. And both designs have received good reviews.

        Given my rather limited speaker building skills, if I were going to use that woofer in a design, I would be tempted to use the box from the BMR-3L and the tweeter and crossover from the Mariposas. The Mariposas box is sealed so will not have as deep of bass as if you use the BMR-3L box.

        The problem with designing a box for the Tectonics is, based on what I have read, the actual T/S specs do not match the published specs. So unless you can measure the specs, any box you design is likely to be off.

        Comment


        • #34
          Your sim looks reasonable, and you're right. THOSE vented box curves have too much "roller coaster" stuff going on (which IS the same as ripple - btw).
          You can see that the closed boxes can barely reach 80Hz.

          Try this: on one of your vented box curves, plug in a volume of 0.15 cu.ft., and enter a box tuning of 61Hz. That would be a 1"id port tube that was 3" long. Should be easily doable! You end up with a pretty nice curve that gives you about +4dB of baffle step compensation (the part of the curve above the "0 dB" line), and an "F3" (considered the limit of bass extension) near 50Hz. That's not an octave below 80, but nearly so. 0.15cf is about the size of 2 kleenex boxes taped together. Not tiny, be can still be used as computer speakers. A PR box probably can't be made to go any lower, and there's no need since a 1"x3" port tube won't be hard to fit in the box.

          Comment


          • #35
            Thanks Chris, I'll give it a go after work. I probably didn't explain myself before but this sudden interest again was I bought 4 of these drivers but only ended up using 2 on the PC speakers build. I intend to make another project for portable use outside and around the joint with the other two.

            I'm sure I was doing something wrong in winisd, but promising to hear I was actually progressing towards the right track. So for me the end goal is to get that Hz down low and a nice smooth curve? Ie ported tuned to 61Hz

            edit: ok i ran it and i can see what you mention, although there is a slight sag the crest curve it isnt as wild as the first simulation. Out of curiosity i did continue to drop the Hz until i got a smooth curve, which i can say i got at 8hz, but it required a 213in long port, i do like a challenge, but that is quite silly
            Attached Files
            Last edited by 3rutu5; 10-17-2018, 06:04 AM.

            Comment


            • #36
              Originally posted by Paul K. View Post
              I modeled the line as built by "3rutu5", but before you view the modeling results I need to make a few comments. When designing a TL the first thing to choose, after choosing a driver, is to estimate what system tuning frequency is likely to be appropriate for that driver such that the overall response shape is reasonably smooth. In the "olden" days that was the driver's fs, but that's only true if the driver's Qts is equal or close to 0.4. When Qts is lower than 0.4, the system tuning frequency will need to be higher than fs, and if Qts is higher than 0.4, the system tuning frequency will need to be lower than fs. This driver's Qts is 1.07, and the system tuning frequency will need to significantly lower that the driver's fs of 92 Hz. So, I've attached some modeling results which I hope are adequately clear. For the record, I prefer drivers for TLs to have a Qts of 0.35 to 0.50.
              Paul
              [ATTACH]n1389354[/ATTACH]
              Paul, just thinking (talking) out loud and forgive my ignorance, based on that modelling you did of the TL and the particular driver I was about half the length of what would have been required to produce a better result. Does that mean i could technically introduce a few more folds to extend the length and get better performance out of a different project? What I'm thinking is similar internal to that of one of those horns (sort of looks like intestines to be crude)

              Comment


              • #37
                If you add more folds you'd increase the length of the line, thereby lowering its 1/4-wavelength resonant frequency. The longer the line, the lower its resonant frequency. What you might achieve with a different driver and different line configuration depends on the driver's characteristics and how optimum the line's design is for that driver.
                Paul

                Originally posted by 3rutu5 View Post

                Paul, just thinking (talking) out loud and forgive my ignorance, based on that modelling you did of the TL and the particular driver I was about half the length of what would have been required to produce a better result. Does that mean i could technically introduce a few more folds to extend the length and get better performance out of a different project? What I'm thinking is similar internal to that of one of those horns (sort of looks like intestines to be crude)

                Comment


                • #38
                  Originally posted by djg View Post
                  Using eyeballs, not cad, the drawing is fine and material thickness is easily determined. The two inner piece lengths are not called out, but are easily interpreted. At least to me.

                  your speakers look very nice. I have a pair of Tyger 23's Mariposas with that woofer and little Dayton tweeter, in solid aspen. Sound nice.
                  I have just watched a video where a guy used the same tweeters and as he called it installed a very basic crossover which was just a capacitor on the tweeter. Not understanding the crossover process, but understanding that it is meant to limit frequencies under 60hz, how do that sound to the experts in audio? They also wired the tweeter up as a mono in a sound bar, but left the drivers as their individual channels. He also mentions that it was he first project (which he has done plenty more) and was trying to keep it all simple and as a learning experience in speaker building and woodworking.

                  for me starting out it sounds simple enough for me to use in my next build.

                  Comment


                  • #39
                    That guy (in the video) is completely lost.
                    A good book for you is "Speakerbuilding 201".

                    Comment


                    • #40
                      Originally posted by Chris Roemer View Post
                      That guy (in the video) is completely lost.
                      A good book for you is "Speakerbuilding 201".
                      Cheers, I had a sneaking suspicion that was going to be the response

                      Comment

                      Working...
                      X