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

MTM Disadvantages

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

  • #31
    High(er) order topologies aren't necessarily high "Q". A high Q filter can promote ringing.

    Comment


    • #32
      Originally posted by craigk View Post
      I have built several mtm designs. Most with 3 rd on tweeter, 2 nd on woofers. With good drivers 2 nd / 2 nd is not a problem. I have never seen a 6 / 6 design on any two way. Please name or show the designs you are talking about, because someone either is using totally junk drivers, or don't have a clue as how to design a crossover.
      Hey Craig. You didn't read my previous post carefully. Dynaudio made a kit with their top of the line drivers called the GEMINI. It uses first order crossovers (6/6) for the tweeter and mid bass. It's a mtm but also a D'Appolito alignment. To get the correct lobing of +/- 15 degrees you have to use odd order crossovers. Mixing even and odd order crossovers may sound fine but does not fulfill the requirements for a D'Appolito alignment. I do not consider Dynaudio drivers or Focal drivers Junk !! You can do a google search for "Dynaudio Gemini speaker" and find a wealth of information and reviews.You can also check Dynaudio archive at Gattiweb.com for full specs

      Comment


      • #33
        I'll reiterate that long ago D'Appolito abandoned the odd-order quadrature phase alignment using BW3 in favor of an even-order in-phase crossover, specifically LR4.
        Francis

        Comment


        • #34
          Originally posted by fpitas View Post
          I'll reiterate that long ago D'Appolito abandoned the odd-order quadrature phase alignment using BW3 in favor of an even-order in-phase crossover, specifically LR4.
          Pretty much, though not necessarily LR4. He came around to realizing that a strict formulaic approach didn't always give the best results, and he went to doing whatever did give the best results in a particular design. That 'new' philosophy can be seen at least as far back as the 'Thor' in 2002.

          www.billfitzmaurice.com
          www.billfitzmaurice.info/forum

          Comment


          • #35
            Originally posted by billfitzmaurice View Post
            Pretty much, though not necessarily LR4. He came around to realizing that a strict formulaic approach didn't always give the best results, and he went to doing whatever did give the best results in a particular design. That 'new' philosophy can be seen at least as far back as the 'Thor' in 2002.
            Well I can't fault his choice of woofers; I use the W18E001s in my system, a pair above and a pair below the horn. For those interested in the Thor, there's a link to a design article near the bottom of the page.

            http://www.seas.no/index.php?option=...=66&Itemid=365
            Francis

            Comment


            • #36
              The Thor is a compromise design by Joe to get those Seas magnesium cone drivers to behave properly. They have a severe resonant peak in the passband that must be dealt with using a LCR filter. Even with the filter many users can still hear the resonate peak as a metallic sound. The general consensus among the many reviews I've read is the midrange is too forward sounding. The bass goes deep with the TL design but is not strong enough to keep up with the midrange. I've read other reviews that say they're the perfect loudspeaker. We all hear things differently so I'm fine with that. In my previous experiences using metal cone drivers I've noticed they have a signature all their own. The Thor uses odd order crossovers just like the the original design but the woofer is now 6db/ octave. The tweeter is still 18db/ octave electrically but with the acoustic roll off is claimed to be a Linkwitz-Riley. The original design used a 18db crossover at 2.2khz. The D-28 is already starting it's 6db/octave rolloff at the crossover frequency so combined with the 18db crossover would it not also be 24db/octave acoustically? Seas hired Joe because of name recognition and wanted to promote their new line of drivers. As far as I can see the two designs are very similar electrically. They both use odd order crossovers at the same crossover frequency and incorporate the natural rolloff of the tweeter in their design. My original build using top of the line drivers is 95db/watt. The Thor is a large speaker but its efficiency is only about 88db/watt. I wonder why that is? It would take over twice the power to achieve the same SPL levels as my original 1984 D'Appolito build.

              Comment


              • #37
                OK. I just thought it was interesting to read about his design process. I don't have stock in SEAS or anything like that.
                Francis

                Comment


                • #38
                  Originally posted by civit View Post
                  I think the trade-off we are identifying is between power response and vertical polar response. This website shows the consequences of the MTM design pretty well:
                  http://www.musicanddesign.com/Power5.html

                  The MTM has a tightly controlled and symmetrical vertical polar response, but the reverberant field will have a big suck out around Fc. I'm not sure this is a terrible thing, but I tend to prioritize smooth power response in my designs.
                  I'd like to weigh in on the topic of MTMs. First, let me point out something about John K's study at the link above: it's for a crossover frequency of 2kHz (only). In the opening paragraph he states:
                  The polar response, at the crossover frequency, is much smoother for the MTM, as was originally demonstrated by D'Appolito. however, we can expect symmetric dips in the polar response below the crossover frequency, again, due to the interference between the M drivers.
                  I did not see any information about driver size (diameter) and placement (separation), but this generally seems to match what I have seen for your "typical" MTM constructed from two 6" midwoofers and a tweeter with a 4" diameter faceplate. You need to interpret John's study in terms of the system under study, what that means for the results, and how that would change if the MTM were conceived differently.

                  With that said, the main problem with this type of MTM system that is made clear by John's study is that the spacing between the two midwoofers is too large to accommodate the crossover point without causing destructive interference that narrows the response pattern and causes the dip in power response. But since the tweeter is a finite diameter you cannot locate the driver centers arbitrarily close together, or close enough to avoid (quoting John's text again) "symmetric dips in the polar response below the crossover frequency... due to the interference between the M drivers". If you intend to use a "tweeter" then you simply cannot escape this consequence of the MTM design unless you find a very small diameter tweeter that can be crossed very low. Smaller (4"-5") midwoofers could also reduce the interference since they can be located closer to each other, but they will act more like a point source so it's not as much of a benefit as it would seem and small woofers tend not to have much bass capability. In the end it is a losing battle and most traditional MTMs will have the power dip below the crossover point.

                  The solution is to throw tradition out the window and instead of a tweeter use a small "full range" type driver with a cone that is around 2"-3" in diameter. Any larger and the response from 10k-20kHz will be beaming, and any smaller and there is not enough cone Sd. Choosing the proper full-range driver will allow the crossover point to be reduced to about 700-800Hz. Since the fullranger's frame diameter is not much larger than 4", you can still place the midwoofers about the same distance apart on the baffle, but with the relatively low crossover point they will no longer produce power response dips below the crossover point, or at least that effect will be much reduced. Now the main challenge is to find a small fullranger with both high sensitivity and good distortion performance, and that can handle some power input (because the bandwidth has been increased).

                  I designed and built an MTM a few years ago (now dismantled) around this principle using the Tang-Band W2-800SL fullranger as the "tweeter" and was very impressed with the performance. I am hoping to re-build this system using (per speaker) a pair of Anarchy 708 midwoofers plus a Tang-Band W3-2141 fullranger sometime this summer. The cab is about 1.5cuft and will be ported at around 35Hz. I used the two Anarchy midwoofers in a different test system in this same cabinet volume and they had great bass but I did not like the response at 1kHZ and above, so the low-crossover-point MTM should be a perfect application for them.
                  Charlie's Audio Pages: http://audio.claub.net

                  Comment


                  • #39
                    Read the 'Thor' article. In it Joe D notes that even he doesn't use use his own original formula for M-M spacing, as he found it unnecessary to have close spacing at normal listening distances. You want to argue about M-T-M with the guy who invented it?
                    www.billfitzmaurice.com
                    www.billfitzmaurice.info/forum

                    Comment


                    • #40
                      Originally posted by billfitzmaurice View Post
                      Read the 'Thor' article. In it Joe D notes that even he doesn't use use his own original formula for M-M spacing, as he found it unnecessary to have close spacing at normal listening distances. You want to argue about M-T-M with the guy who invented it?
                      Let's look at what Mr D'Appolito says in his own article.

                      From JA's Thor design article, in the "crossover design criteria" section:
                      Even if the on-axis response is flat, poor off-axis response can produce a perceived colored frequency balance.
                      So, JA seems to have a firm understanding about how power response problems can lead to a "colored" sound, which typically means "bad" in my book.

                      Later in the section on Crossover Design and Optimization JA remarks:
                      In the case of the MTM configuration they also limit off-axis response in the vertical which greatly reduces floor and ceiling reflections.
                      The suckout and dip in power response is explained away as a "benefit"? The flawed vertical polars that will result with an MTM design like this, with its driver spacing and 2.5kHz crossover point, will only lead to a dip in the power reponse like John KJ. shows with his simulations. I did notice that while a plot of the horizontal polar response is shown, no vertical polars are shown in the Thor article. That's because they would not look so good, and it would be clear that the system will have a power response dip!

                      Unfortunately, I find this to be quite common. The clever and seasoned designer knows that it is nearly impossible to make a traditional MTM without some vertical suckout below the crossover point, so they wave their hands and say it's actually a feature. But this will only lead to that "colored" sound JA mentions. I don't see how this is a benefit.

                      The real advantage of the MTM is its relatively flat power response and symmetric vertical response in the crossover region. There is no MTM magic that will solve the problems of the midwoofers being too far apart below the crossover point - the MTM arrangement doesn't help that in any way whatsoever. The only good solutions are to
                      (A) move the drivers much closer together, which leaves no room for the tweeter and the system can't be constructed, or
                      (B) use a lower crossover point, however, there is no tweeter (e.g. circa 1" dome, etc.) that can handle a crossover point that is moved below 1kHz, where the response hole in the vertical polars will start to disappear.

                      This is why I feel that the best MTM systems will be MFM and will have a crossover point around 700-800 Hz with closely spaced drivers.
                      Charlie's Audio Pages: http://audio.claub.net

                      Comment


                      • #41
                        Originally posted by charlielaub View Post
                        The real advantage of the MTM is its relatively flat power response and symmetric vertical response in the crossover region. There is no MTM magic that will solve the problems of the midwoofers being too far apart below the crossover point - the MTM arrangement doesn't help that in any way whatsoever. The only good solutions are to
                        (A) move the drivers much closer together, which leaves no room for the tweeter and the system can't be constructed, or
                        (B) use a lower crossover point, however, there is no tweeter (e.g. circa 1" dome, etc.) that can handle a crossover point that is moved below 1kHz, where the response hole in the vertical polars will start to disappear.

                        This is why I feel that the best MTM systems will be MFM and will have a crossover point around 700-800 Hz with closely spaced drivers.
                        In my simulations of center channel MTM speakers, even using a very small and very robust tweeter doesn't help much. In fact, all the typical meausres to reduce CTC spacing do very little - small woofers, offset tweeter, whatever. At issue really is the directivity mismatch between a big speaker array playing really high and a small tweeter playing too low. Using a very robust 32mm tweeter at 1000hz or something doesn't help matters much because the tweeter's directivity at this frequency is incredibly wide, and the woofer array is very narrow.

                        I think you are correct about the MFM, it's an interesting idea to sacrifice a little upper treble linearity for smooth power response. I think it also makes sense to do a MCM, with a coaxial, or even look into MTTM, with two tweeters in a vertical array. Dynaudio does this on their bigger speakers.

                        Comment


                        • #42
                          I couldn't stand the Seas L15 aluminium driver (poor cousin to the W15) crossed at 2.7Khz due to the 3rd order HD peak at the same. I ended up at 2300Hz at the top end of what I could bear, and low end of what I thought the tweeter was capable (still don't like the tweeter). The 2500 XO of The Thor is above its W18 3rd order peak. I would cross this at 1800 or lower, which would require another tweeter / waveguide (and I don't know how nicely a waveguide would play between 2 midwoofers, plus increases to CTC).

                          Comment


                          • #43
                            I tend to agree that crossing the W18s higher than maybe 1.5kHz is asking for difficulties. I cross mine at 800Hz LR4 to the horn, so the SEAS resonant peaks are firmly in the stopband.
                            Francis

                            Comment


                            • #44
                              Every time I wonder what I’m giving up by listening to a full range driver from 300-Hz up, one of these threads pops up and reminds me why I went this way.

                              Comment


                              • #45
                                Can we all agree to shun the MTM, or is legislative action called for?

                                And I'm stuck with 3 S2000 MTMs. One sideways even.

                                Comment

                                Working...
                                X