Announcement

Collapse
No announcement yet.

Help with WinIsd speaker alignment options, please.

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

  • Help with WinIsd speaker alignment options, please.

    Hi Gents,
    For several yearsI have always used the C4 SC4 Chebyshev option when designing boxes for my small two way speakers. I've been doing this because I read somewhere that it was the best alignment. for high fidelity speakers of 1 CF or less size. What are the Qb3 Quasii-Butterworth, B4 SSB4 super boom box and other options for and when would it be appropriate to use them?

    Best,
    Jay

  • #2
    I too would enjoy an explanation or direction to an explanation of the differences/benefits of these different alignments.

    Comment


    • #3
      Alignments were created back in the days when slide rules and scientific calculators were used along with formulas to design boxes, to reduce the virtually unlimited number of possibilities to a manageable few. Speaker design software rendered them obsolete except as a starting point, since software allows you to create as many 'alignments' as you may wish, with charts you can overlay to see how they compare against each other.
      www.billfitzmaurice.com
      www.billfitzmaurice.info/forum

      Comment


      • #4
        What Bill said, there are an infinite number of ways to tune a bass reflex box, and it all comes down to taste in the end. I like to tune for the minimum group delay possible, some people (especially with tiny speakers) prefer a small hump in response and live with the extra group delay. Where the speakers go in the room affects things, too.
        Francis

        Comment


        • #5
          Just take your fav. woofer and run all 5 sims on the same screen (to see what's what).
          Prob. the way the box vol. and tunings change (whether a given "alignment" wants a larger/smaller Vb or a higher/lower Fb) are most closely related to varying values for Qts.
          SOMEtimes the (1st 3 / top 3) ISD alignments won't differ appreciably.
          Using Dayton's TCP115-4 (for example, using my OWN DATS derived T/S parms), the QB3 and BB4 boxes/tunings are virtually identical (.08cf tuned to 60Hz). The C4 wants a box 3/4ths the size, tuned 5Hz higher (w/a loss of 15-20Hz off the bottom end). The EBS alignments want a box about 2/3rds larger, tuned in the mid-40s (for EBS-3) and mid-30s (for EBS-6).

          One of the coolest things you can do in WinISD is to call up a ported box, then click your cursor INside the little ported box icon and drag your mouse around. You can "draw" nearly any rolloff curve you'd like that way !

          Comment


          • #6
            So for any given driver, by varying the box size and tuning fequency, one can create a bass frequency roll off identical to the one WinIsd would have provided if you had initially selected a different "alignment". Francis, please elaborate on your tuning approach, I don't know what "minimum group delay" means and a google search on the term turned up a lot of results unrelated in any way I can see to speaker building.
            Thanks!
            Jay

            Comment


            • #7
              Originally posted by Drjay View Post
              So for any given driver, by varying the box size and tuning fequency, one can create a bass frequency roll off identical to the one WinIsd would have provided if you had initially selected a different "alignment". Francis, please elaborate on your tuning approach, I don't know what "minimum group delay" means and a google search on the term turned up a lot of results unrelated in any way I can see to speaker building.
              Thanks!
              Jay
              A speaker can be considered as a high-pass device, especially a woofer. I tune for the least group delay, using the group delay plots in WinISD. I'm not sure I can really explain any better, but that's what I do. By varying the port length, box size etc. I try to achieve the least group delay, which conversely means the lowest polynomial order low frequency roll-off, as if the speaker was an actual high-pass filter. I'm an engineer, not a teacher. Maybe somebody else can explain it better than I did.

              In any event, once the speaker is in the room, and the response is further shaped by cabin gain, etc., things are different yet.

              As I said before, the least group delay method probably isn't useful for small speakers. The last ones I did that way were 10ft^3 boxes with four 8" woofers apiece, tuned to 28Hz. They sound quite nice, but small they aren't.
              Francis

              Comment


              • #8
                Originally posted by Chris Roemer View Post
                Just take your fav. woofer and run all 5 sims on the same screen (to see what's what).
                Because playing in WinISD is more fun than my day job, I couldn't resist.

                I ran all 5 pre-loaded alignments as well as a custom alignment ("Own Creation") using the T/S parameters for a DSA270-8 (Dayton Audio Designer Series Aluminum cone 10" woofer).
                The "Own Creation" has a 1 db hump before its roll-off. None of the named alignments do that, as they were all developed mathematically to achieve certain roll-offs, but many people design their own speaker to have up to a 3db hump.

                I don't understand Group Delay but I included that graph also.

                QB3 Quasi-Butterworth is green
                BB4_SBB4 Super Boom-Box is royal blue
                C4_SC4 (Sub) Chebychev is magenta
                EBS 3 Extended Bass Shelf -3db is red
                EBS 6 Extended Bass Shelf -6db is turquoise
                Own Creation is purple

                Click image for larger version

Name:	Transfer Function.PNG
Views:	134
Size:	39.4 KB
ID:	1444942


                Click image for larger version

Name:	Group Delay.PNG
Views:	134
Size:	119.2 KB
ID:	1444943

                Comment


                • #9
                  Originally posted by a4eaudio View Post

                  Because playing in WinISD is more fun than my day job, I couldn't resist.

                  I ran all 5 pre-loaded alignments as well as a custom alignment ("Own Creation") using the T/S parameters for a DSA270-8 (Dayton Audio Designer Series Aluminum cone 10" woofer).
                  The "Own Creation" has a 1 db hump before its roll-off. None of the named alignments do that, as they were all developed mathematically to achieve certain roll-offs, but many people design their own speaker to have up to a 3db hump.

                  I don't understand Group Delay but I included that graph also.

                  QB3 Quasi-Butterworth is green
                  BB4_SBB4 Super Boom-Box is royal blue
                  C4_SC4 (Sub) Chebychev is magenta
                  EBS 3 Extended Bass Shelf -3db is red
                  EBS 6 Extended Bass Shelf -6db is turquoise
                  Own Creation is purple

                  Click image for larger version

Name:	Transfer Function.PNG
Views:	134
Size:	39.4 KB
ID:	1444942


                  Click image for larger version

Name:	Group Delay.PNG
Views:	134
Size:	119.2 KB
ID:	1444943
                  OK, cool. Note the alignments with the "laziest" roll-off have the least group delay. I aim to optimize that, for large speakers at least.
                  Francis

                  Comment


                  • #10
                    Originally posted by fpitas View Post

                    A speaker can be considered as a high-pass device, especially a woofer. I tune for the least group delay, using the group delay plots in WinISD. I'm not sure I can really explain any better, but that's what I do. By varying the port length, box size etc. I try to achieve the least group delay, which conversely means the lowest polynomial order low frequency roll-off, as if the speaker was an actual high-pass filter. I'm an engineer, not a teacher. Maybe somebody else can explain it better than I did.

                    In any event, once the speaker is in the room, and the response is further shaped by cabin gain, etc., things are different yet.

                    As I said before, the least group delay method probably isn't useful for small speakers. The last ones I did that way were 10ft^3 boxes with four 8" woofers apiece, tuned to 28Hz. They sound quite nice, but small they aren't.
                    Sorry that is not what it means at all
                    craigk

                    " Voicing is often the term used for band aids to cover for initial design/planning errors " - Pallas

                    Comment


                    • Steve Lee
                      Steve Lee commented
                      Editing a comment
                      Very helpful.

                  • #11
                    Curt Campbell's site has an article about group delay.

                    Comment


                    • #12
                      Originally posted by craigk View Post

                      Sorry that is not what it means at all
                      Great! Doubtless you'll explain far better than I can, Craig.
                      Francis

                      Comment


                      • #13
                        I think that sealed is the holy grail for bass quality. Unfortunately, it has disadvantages in some situations. So, I try for minimum group delay from ported in an attempt to make it sound more "sealed", if that makes sense. It's partially successful, but adding phase-unwrapping DSP also helps.
                        Francis

                        Comment


                        • #14
                          FWIW I've never found group delay to be audible, within reason. In those cases where it was long enough to be audible other factors overshadowed it anyway.
                          www.billfitzmaurice.com
                          www.billfitzmaurice.info/forum

                          Comment


                          • #15
                            Originally posted by billfitzmaurice View Post
                            FWIW I've never found group delay to be audible, within reason. In those cases where it was long enough to be audible other factors overshadowed it anyway.
                            I was a little skeptical until a friend did an A-B with and without DSP phase unwrapping. It made a small, but detectable difference. Less overhang on low notes, if I had to name what I was hearing. In any event, all I'm really doing is tuning for the smoothest rolloff, which seems to work well in a larger room.
                            Francis

                            Comment

                            Working...
                            X