Announcement

Collapse

Midwest Audio Fest

It’s that time audio enthusiasts! Registration for the 2019 Speaker Design Competition is now open! Visit midwestaudiofest.com for details and to list your speaker project. We are excited to see all returning participants, and look forward to meeting some new designers this year, as well! Be sure your plans include a visit to the Parts Express Tent Sale for the lowest prices of the year, and the Audio Swap Meet where you can buy and trade with other audio fans. We hope to see you this summer! Vivian and Jill
See more
See less

Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

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

  • #76
    Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

    Oh, it was for a coincident design mounted in a spherical composite enclosure.
    The tweeter is mounted in the base of the woofer. There is another portion on the rear side that acts as a sort of guide for the midrange which vents out the slots around the waveguide.


    Jay,

    I'm a little bit confused, so there is no audible benefit achieved by physical time alignment that can't be achieved by
    filter topology?
    Attached Files

    Comment


    • #77
      Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

      Originally posted by Cncbydesign View Post
      Jay,

      I'm a little bit confused, so there is no audible benefit achieved by physical time alignment that can't be achieved by filter topology?
      Yes, that is true in theory and in practice. Physical time alignment is strictly required only by an acoustic 1st order, transient-perfect filter system, which is rare in practice (for example, Thiel and Vandersteen). All other crossover types exhibit a certain amount of phase distortion, anyway, whether drivers are time aligned or not. Time-aligned LR4 and correctly implemented non-time-aligned LR4 (with a relaxed rolloff on the midwoofer) cannot be distinguished from each other when you measure/hear finished systems. It's not only my understanding or opinion. If you'd like, I can ask other crossover gurus to chime in on this.

      jAy

      Comment


      • #78
        Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

        Hi jAy,

        How are you keeping, I am going to buy 8 of these drivers. I would love to build the MTMMM with the SB neo. Any Chance of a design in this config? Hope so

        Regards,

        Stu

        Comment


        • #79
          Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

          Originally posted by Stu View Post
          Hi jAy,

          How are you keeping, I am going to buy 8 of these drivers. I would love to build the MTMMM with the SB neo. Any Chance of a design in this config? Hope so

          Regards,

          Stu
          Hi Stu,

          I can model it up for you, but won't post it here---too many options already. I email you. If not, please give me a reminder in a week or two. I'm sorry you missed the sale, but I believe this midbass still has its value at $43 each ($39 if you order more than 3). I remember it was more expensive before.

          jAy

          Comment


          • #80
            Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

            Originally posted by jkim View Post
            This is not true. In MOst cases, physical time aligment has nothing to do with a speaker's performance in the time domain. With analog crossovers, time alignment of drivers is required only by a transient-perfect system with acoustic 1st order filters. It is useful for acoustic 2nd order filers, too, but it is mainly to obtain good on-axis phase tracking (and thus flat summation of drivers), not to improve the finished system's performance in time domain---...
            I'm not talking about phase alignment, I'm talking about time alignemnt.

            A friend of mine did controlled listening tests between time aligned and non time aligned systems and determined that there was a noticable improvement in perceived sound quality in the time aligned systems.

            If anything, the sound from the tweeter should arrive slightly before the woofer to give the most realistic reproduction in sound. I've mentioned before, I don't find most speakers enjoyable to listen to, other than for background music - the sound they produce is too disimilar (somehow unatural sounding?) to the actual instruments I think. I have yet to do sufficient testing to put forward anything other than guesses as to what might be the top few most significant shortcomings of modern speakers...

            But, I do know, just having a flat frequency response and low distortion is not sufficient to make a great sounding speaker IMHO.
            "...this is not a subwoofer" - Jeff Bagby ;)

            Comment


            • #81
              Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

              Originally posted by critofur View Post
              I'm not talking about phase alignment, I'm talking about time alignemnt.

              A friend of mine did controlled listening tests between time aligned and non time aligned systems and determined that there was a noticable improvement in perceived sound quality in the time aligned systems.

              If anything, the sound from the tweeter should arrive slightly before the woofer to give the most realistic reproduction in sound. I've mentioned before, I don't find most speakers enjoyable to listen to, other than for background music - the sound they produce is too disimilar (somehow unatural sounding?) to the actual instruments I think. I have yet to do sufficient testing to put forward anything other than guesses as to what might be the top few most significant shortcomings of modern speakers...

              But, I do know, just having a flat frequency response and low distortion is not sufficient to make a great sounding speaker IMHO.
              when you add a high pass filter into the mix, you add delay to the signal going to the tweeter. Time alignment counts for very little in anything but a design using TP approach. Any other XO topology negates any benefit to aligning acoustic centers. Aligning acoustic centers can help the XO design, but it does nothing for accurately reconstructing a square wave, other than in a 1st order (acoustic) filter. A dozen mm here or there can be accommodated in the filter.

              All that really matters around the XO point is that the phase relationship between the drivers is constructive to a smooth forward lobe, preferably with a smooth power response as well. Getting the phase correct is done with the XO in all but 1st order designs.

              If you think you can tell the difference between a time aligned LR4 setup and a non-time aligned version, be my guest. But please document your analysis so that you can truly convince yourself that time alignment is audible and beneficial in other than 1st order systems.

              One question about the "careful test" mentioned above. Were the designs under evaluation in similarly constructed cabinets using identical drivers?
              R = h/(2*pi*m*c) and don't you forget it! || Periodic Table as redrawn by Marshall Freerks and Ignatius Schumacher || King Crimson Radio

              Byzantium Project & Build Thread || MiniByzy Build Thread || 3 x Peerless 850439 HDS 3-way || 8" 2-way - RS28A/B&C8BG51


              95% of Climate Models Agree: The Observations Must be Wrong
              "Gravitational systems are the ashes of prior electrical systems.". - Hannes Alfven, Nobel Laureate, Plasma physicist.

              Comment


              • #82
                Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                Originally posted by jkim View Post
                This is not true. In MOst cases, physical time aligment has nothing to do with a speaker's performance in the time domain. With analog crossovers, time alignment of drivers is required only by a transient-perfect system with acoustic 1st order filters. It is useful for acoustic 2nd order filers, too, but it is mainly to obtain good on-axis phase tracking (and thus flat summation of drivers), not to improve the finished system's performance in time domain---the sytem's total phase wrap is 180 degrees, anyway. The most popular, easy-to-use crossover type is the Linkwitz-Riley acoustic 4th order filter, which must be employed by over 90% of commercial and DIY speaker designs with drivers mounted on a flat baffle (w/ 5.25" or larger midrange and no-wave-guided dome tweeter). In this case, we typically use what we call asymmetric filter to obtain drivers' good phase tracking. If it is implemented correctly, it exhibits virtually no difference in terms of the system's phase behavior from a time-aligned LR 4th order system.
                I agree with one exception, well, maybe two. There are a few other T-P crossovers that are not acoustic first order. Time-alignment is necessary for any of them with the caveat that the phase tracking is ideal. The optimal results, not perfect that is, may require a balance of time- and phase-alignment. That will always be case-dependent.

                As for phase behavior, remember that an ideal LR4 and the relaxed slope variant may have rather different off-axis behavior, hence power response, due to phase in the crossover region.

                Active crossovers that can adjust delay on the design axis suffer more in the off-axis, even if time-aligned on the design axis, so in that case it's still probably better to physically align (for phase) if possible. John K. compared T-P systems to non-T-P he constructed, I believe, and found that there was little if any significant benefit even to a T-P system. They are, after all, only T-P on a single axis for non-coincident systems, virtually all of those that are multi-driver ones.

                Originally posted by Cncbydesign View Post
                I'm a little bit confused, so there is no audible benefit achieved by physical time alignment that can't be achieved by filter topology?
                That's one I don't think that anyone can answer with authority, though there can benefit to aligning (phase alignment, that is) due to possibly improved off-axis response. The problem is that to make any definitive statement, all else must be equal. For speakers and various topologies/alignments, it's not possible to have "all else equal". We can have a lot of empirical evidence, but one would have to do the level of research and testing beyond our means. There's a lot of evidence that it is not audibly significant. Certainly not anything dramatic.

                Originally posted by critofur View Post
                I'm not talking about phase alignment, I'm talking about time alignemnt.

                A friend of mine did controlled listening tests between time aligned and non time aligned systems and determined that there was a noticable improvement in perceived sound quality in the time aligned systems.
                This would require "all else equal". I doubt that this could have been the case. Power response, distortion profile, possibly even compression, etc., all would have to be taken into account. I don't find this to be convincing.

                If anything, the sound from the tweeter should arrive slightly before the woofer to give the most realistic reproduction in sound.
                On what do you base this and in what type of system(s)? As noted elsewhere in the thread, unless it's a T-P system, there will be significant delays between drivers that are orders of magnitude larger than any small adjustments to align, especially for a tweeter. A square wave output will look very little like a square wave, on-axis or not, if it's not a T-P system.

                But, I do know, just having a flat frequency response and low distortion is not sufficient to make a great sounding speaker IMHO.
                On that we can all strongly agree. Jay is certainly taking more than that into account from what I've read, such as power response that is very high on the list.

                dlr
                WinPCD - Windows .NET Passive Crossover Designer

                Dave's Speaker Pages

                Comment


                • #83
                  Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                  Thanks for your reply, Dave. Yes, of course, there is a 2nd order quasi T-P system (acutally M-P in effect). In fact, as you know, all real-world implementations of BW1 essentially have 2nd order slopes at extremes. But the emphasis in my above reply was on "T-P," not on "1st order." To obtain a T-P response on the design axis, time alignment is required. That was my point.

                  And about the differences between time-aligned LR4 and non-time-aligned LR4 with asymmetric rolloffs, I agree that if you go off-axis there can be some differences. But if it's strictly on axis, we won't have any easy way to distinguish between them when we measure only finished systems' (not individual drivers') on-axis responses. Also, Im pretty sure that, depending on the amount of drivers' relative acoustic offset, vertical distance between drivers on the baffle, and how we impelement the asymmetrical LR4, etc, there should be a situation in which it is virtually impossible to distinguish between them by measuring two finished systems' on- and off-axis responses. In fact, I believe that in most cases with a 6.5" or a smaller midbass driver, the off-axis behavior of a well-implemented asymmetric LR4 2-way system with drivers mounted on a flat baffle is not very different in any significant ways from that we would obtain from a time-aligned LR4 system.

                  And Pete made a good point. Other than T-P crossovers, any other crossover types causes each individual driver to have selectively different amounts of phase shift over the frequency, which are essentially time delays in the time domain. Time alignment of drivers' acoustic centers does not affect this fact to any significant degree. It only helps two drivers to be phase aligned on the design axis so that flat summation is obtained on that axis, which can also (and usually) be achieved without time alignment. Distortion in the time domain, caused by the filters, still remains. In fact, I mentioned this in one of my previous replies, but did not elaborate enough.

                  jAy

                  Comment


                  • #84
                    Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                    Originally posted by jkim View Post
                    And about the differences between time-aligned LR4 and non-time-aligned LR4 with asymmetric rolloffs, I agree that if you go off-axis there can be some differences. But if it's strictly on axis, we won't have any easy way to distinguish between them when we measure only finished systems' (not individual drivers') on-axis responses.
                    That may be, but what good is it to limit one's investigation to on-axis only? That's one small part of the sound of a system. That's a prime example of the limitations of any system.

                    Also, Im pretty sure that, depending on the amount of drivers' relative acoustic offset, vertical distance between drivers on the baffle, and how we impelement the asymmetrical LR4, etc, there should be a situation in which it is virtually impossible to distinguish between them by measuring two finished systems' on- and off-axis responses.
                    Yes, but you're setting artificial restraints, "...there should be a situation...". One can easily set up conditions in which differences are difficult to detect. It's never impossible unless limitations are imposed.

                    In fact, I believe that in most cases with a 6.5" or a smaller midbass driver, the off-axis behavior of a well-implemented asymmetric LR4 2-way system with drivers mounted on a flat baffle is not very different in any significant ways from that we would obtain from a time-aligned LR4 system.
                    Possibly. Not guaranteed, of course. As always, it's case-dependent.

                    dlr
                    WinPCD - Windows .NET Passive Crossover Designer

                    Dave's Speaker Pages

                    Comment


                    • #85
                      Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                      Thanks again for the reply, Dave. I agree that there is no definite answer if we include all the aspects of a system. But I think we made the topic more complicated than necessary. Also maybe beyond the scope of the original questioners' concern. ;) Their question seemed to be about the direct effect of time alignment upon a speaker's performance in time domain, not about its secondary effects like off-axis behavior, which must be minute especially for higher order filters like LR4. I think their questions have been addressed well in Pete's and my replies. For example, the questioners' posts say:

                      "If anything, the sound from the tweeter should arrive slightly before the woofer to give the most realistic reproduction in sound."

                      "the lack of time alignment seems to plague most DIY speakers and most commercial speakers."

                      "Lack of time alignment seems to be one of the most universal ones chosen, to some people I've talked to, though, they would not even want to bother to listen to a system that was not properly time aligned."

                      In what sense specifically? We can certainly clarify one thing. That's what I and Pete did, because we thought that's what their question was about.

                      -jAy

                      Comment


                      • #86
                        Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                        Originally posted by jkim View Post
                        Thanks again for the reply, Dave. I agree that there is no definite answer if we include all the aspects of a system.
                        How else to analyze a speaker system? We don't hear just the design axis response.

                        Their question seemed to be about the direct effect of time alignment upon a speaker's performance in time domain, not about its secondary effects like off-axis behavior, which must be minute especially for higher order filters like LR4.
                        Time-domain performance is not a single axis issue or at least should not be. Our perceptions related to the time-domain also encompass all axes. There's no general rule that says it's a minute issue as I see it. That completely ignores aspects that are arguably far more important, driver directionality and diffraction, the former having more serious affect on the time-domain performance in the off-axis. Perfect time-alignment on a single design axis is possibly the least important issue and may be almost insignificant compared to other issues. That's the answer I would offer.

                        Going back to what I think is your first reply on the topic, "In MOst cases, physical time aligment has nothing to do with a speaker's performance in the time domain.", I can't agree with this unless one is artificially imposing a restriction to a single design axis. Time deltas between drivers due to offsets in 3 dimensions creates other lobing issues that can strongly affect power response. Optimizing on a single axis to adjust phase tracking will likely yield a different power response since the corrections are essentially for the lobing. This is altering the lobe summed response on that axis that in turn alters the power response since the rest of the lobing is also affected. Physically aligning obviates the need for correction for phase (not time) and may simultaneously improve (maybe I should just say "alter") the power response. Small changes, no, but again, it's always case-dependent. The smaller the midwoofer in a 2-way, the less important in general simply due to lesser acoustic offset.

                        dlr
                        WinPCD - Windows .NET Passive Crossover Designer

                        Dave's Speaker Pages

                        Comment


                        • #87
                          Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                          Originally posted by dlr View Post
                          Optimizing on a single axis to adjust phase tracking will likely yield a different power response since the corrections are essentially for the lobing. This is altering the lobe summed response on that axis that in turn alters the power response since the rest of the lobing is also affected. Physically aligning obviates the need for correction for phase (not time) and may simultaneously improve (maybe I should just say "alter") the power response.
                          In your statements, it seems that you're saying theoretical LR4 with time alignment has all good properties. I know that you don't, but although you put qualifier "maybe I should just say 'alter'," other people may misinterpret. With TM layout, as you know, even with time alignment, we optimize only on a single axis vertically, and in this sense, we can view that time alignment is, just like using asymmetric rolloffs, simply a means to phase align drivers on a single axis. Horizontally, of course, a time aligned system will stay that way over a wider angle, but with the beaming of real-world midwoofers factored in, I suspect that the effect of time alignment versus no time alignment won't make a night and day difference on horizontal axes, either.

                          But I already agreed that quasi-LR4 with asymmetric rolloffs could exhibit different off-axis behavior than LR4 with time alignment. But which would result in better perceived sound quality? That's a totally different question. Like you say, I think it's case-dependent. In fact, a special case of asymmetric LR4, with woofer rolloff being very shallow up to 2nd order and tweeter rolloff steep up to 5th or 6th order, has a benefit in smoothing out the power response dip and hump combination that is notoriously associated with the harshness of a 2-way system. Just like LR2 (a caveat, though, because they are not precisely the same), this mitigates the effect of the hump in LR4 power response, which usually occurs an octave above Fc.

                          Comment


                          • #88
                            Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                            I have a question before starting this interesting build. How much of crossover design change would be needed if the cabinet had massive roundovers?

                            Or say if the spacing between the drivers was the same as the original design, but each driver was inside it's own "sphere"? Maybe set up in a 'figure 8' pattern, each with big roundovers, and a fairly deep elongated enclosure.

                            Any ideas?

                            This seems to be a cheap enough design to experiment with, but I'd be going in blind with regards to the crossover. Would there still be a need for the baffle step?

                            I've got some JVC 3x5" line arrays to finish up first, but this design will be next.

                            Comment


                            • #89
                              Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                              How big roundovers? You can have up to 1" roundovers with 9" wide baffles. You won't need any crossover modification unless the cabinet width is greater than 10".

                              jAy

                              Comment


                              • #90
                                Re: Willing to do some modeling for the Tang Band W6-789E 6.5" midbass (BF Specials)

                                Originally posted by jkim View Post
                                TM with SB29RDCN-C000-04

                                Cabinet

                                It turns out that this midbass driver does not like anything smaller than 0.8 cu ft (22.5 liters) for a vented alignment. For this TM modeling, I used 0.8 cu ft net volume with 46 Hz tuning, but 1.0 cu ft with 44 Hz tuning looks even better. This means that the cabinets have to be either tall standmounters, or floorstanders with a false chamber at the bottom.

                                With 0.8 cu ft volume, a 2.5" ID tube cut to 5.75" length (including the flare) will provide 46 Hz tuning. With 1.0 cu ft volume, a 2.5" ID, 5" long tube will give 44 Hz tuning. It can be front- or rear-ported. The cabinet's inner walls should be lined with acoustic foam, and the bottom 4" to 6" be filled with dense polyfill.

                                A sealed cabinet requires 0.5 to 0.6 cu ft, and will provide a system -3 dB point of 80 Hz. This is perfect for the speakers to be used with a subwoofer, integrated by an AV receiver's active low and high pass filters. A sealed cabinet should be filled 80% with polyfill.

                                Baffle width should 8.5" to 9". The tweeter's center is placed 3" away from the baffle's top edge with no horizontal offset. The midwoofer's center is 5.125" (5-1/8") below the tweeter's center, that is, 8.125" away from the baffle's top edge. All drivers should be flush mounted in properly recessed cutouts.

                                I do not provide cabinets' external dimensions. You should be able to come up with your own based on the requirements given above.

                                I recommend rounding over the baffle's top and sides using at least 1/2" round over bit (rounding over the bottom edge is not needed). In case the baffle is 1" thick or thicker, you should chamfer the midwoofer holes from inside to ensure good airflow behind the drivers.

                                Crossover

                                Below are the schematic---all component specs are included---, and its predicted responses:






                                Don't use expensive caps or foil inductors. Parts Express Dayton standard poly caps are perfect, and Jantzen or Erse coils will work great. For L1 (2.5 mH), Parts Express steel laminate 18 gauge is a great choice. On the crossover board, inductors need to be placed carefully to avoid their interference. If you post a picture of your initial layout in this thread, people will give you good feedback.

                                When soldering components, if you don't have much experience, it's always a good idea to use heat sinks like alligator clips between joints being soldered and component bodies not to damage them by overheating. 14 gauge (even 16 gauge) cable is thick enough for internal wiring.

                                Options

                                Tweeter fine-tuning: I strongly recommend fine-tuning the tweeter response by adjusting R9 and C9 values. Default values are 4.7 ohm and 7.8 uF (6.8 uF and 1.0 uF caps wired in parallel). To tweak C9, buy 6.8 uF and two 1.0 uF caps per speaker. Then you'll be able to try 6.8, 7.8, and 8.8 uF by wiring them in parallel.

                                -jAy
                                Alright, I have all the parts for this project. I decided on a sealed TM, with the SB Acoustics tweeter, and I took your advice and got the additional parts to test the different values.

                                What is a good method to swap between them easily? I was thinking of initially keeping the crossover outside, and running the wire through a temporarily sealed hole that would eventually be one of the wire terminals, and using a basic toggle switch, but was concerned of its ability to carry a clean signal through. Can anyone recommended a good switch for this purpose?


                                Has anyone else that took advantage of the sale started on a project using them?
                                My projects
                                https://picasaweb.google.com/112307725038877176664

                                Comment

                                Working...
                                X