New Full Range Three Way Project - longish post

Collapse
This topic is closed.
X
X
 
  • Time
  • Show
Clear All
new posts
  • ChrisGraeff
    Junior Member
    • Sep 2005
    • 11

    New Full Range Three Way Project - longish post

    Over the past few years I have auditioned a number of speakers spanning the universe of prices. One of my favorites was the Eggleston Andra - version one. While not the most uncolored speaker I've heard it's performance from 1000 hz down was excellent. The Dynaudio 30W driver handled the lowest frequencies with authority and musicality. Further the dynamic range was impressive, easily handling live Diana Krall and Mickey Hart's Planet Drum at realistic play-back levels.

    Having read the Stereophile article, which was glowing, I was skeptical based on 1) their habit of loving exciting (read as: colored) sound, and 2) the freq response graphs showing a massive suck out from ~1500 - 3000 hz. To my surprise the human voice seemed very natural. On the other hand the piano on the Diana Krall disc sounded unnatural. When I commented to a friend accompanying me the best description we developed was plastic sounding. I know not very enlightening.

    Regardless of this limitation I have been contemplating a pseudo-clone. The mids are Morel MW166 - no problem acquiring them in 4 or 8 ohm variety. Dynaudio drivers are unavailable save for the eBay route so I'm excluding them as a possibility. The Morel Supremes received an extremely favorable report in Voice Coil and owe some of their lineage to Dynaudio. At this point they are my selection for the tweeter. The 30W is an undersized 12 inch driver with an Sd of 62 sq inches (400 sq. cm) according to Dynaudio literature. Morel does not have an equivalent. Eton's 11 inch looked promising but not from an economic perspective. The new Dayton RS 10 and 12 hi fidelity subwoofers look like they may fit the bill. Low Le and the first cone break up mode is above 1500 hz. This should allow it to work with a 2nd order low pass at 150 - 200 hz with either a notch filter or cascaded with another 2nd order low pass filter at ~600 - 800 hz.

    It is my contention an optimized XO can make this driver combination something special. While I have not modelled anything in LspCad my first thought is to 3rd order high pass the Supreme at ~1900. The upper MW166 will receive a 2nd order low pass at a similar frequency. The lower MW166 will be crossed in at 2nd order at ~400 - 600 effectively forming a 2.5 way speaker. Both MW166s will be high-passed at 150 -200 hz using a first or second filter. The Dayton 12HF will be XO'ed as mentioned above and loaded in an vented box. They will not be employed an isobaric arrangement.

    Any feedback regarding driver selection or XO choice would be appreciated. Even those loaded with caveats or negative comments! Note this project, other than enclosure and XO modeling, probably will not start until end the end of this year.
  • ThomasW
    Moderator Emeritus
    • Aug 2000
    • 10933

    #2
    Jon and I gave up on poly cone drivers about 10yrs ago. We evolved through the exotic plastic hybrids, and have settled on metal cones as our drivers of choice. As a result there's little input we can offer regarding your choice of upper range drivers.

    IMO the hallmark of the Eggleston design was their granite laminated cabinets. I think the lack of cabinet resonances was one of their strong suits.

    As usual the limitation of a 'fullrange' system comes into play. It's very difficult to get the lowest octaves from the woofer, when it's being ask to play up into the midbass. This is why we prefer to use a dedicated sub with mid-tower modules handling the higher frquencies

    IB subwoofer FAQ page


    "Complicated equipment and light reflectors and various other items of hardware are enough, to my mind, to prevent the birdie from coming out." ...... Henri Cartier-Bresson

    Comment

    • ChrisGraeff
      Junior Member
      • Sep 2005
      • 11

      #3
      Interesting that you "gave up on poly". I have not used a metal cone driver in a design however I have heard them in the Revel speakers in a direct comparison to Aerial Acoustics models. With all apologies to Revel's designer - Kevin Voecks? - I thought the Aerials ran neck-and-neck with the much more expensively priced Revel. No doubt the XO played a valuable part in Aerials performance.

      Admittedly the MW166 mids were designed a number of years ago - more than 10 since they were introduced but they still yield an excellent sonic presentation.

      Regarding using the RS12HF as my woofer - assuming you have seen the FR graph and specs on mobilediy, would you concur with my low pass XO design thoughts mentioned in the initial post?

      Comment

      • ChrisGraeff
        Junior Member
        • Sep 2005
        • 11

        #4
        One other note: it appears Jon has used a poly cone in a recent design, the Hi Vi D6.8 mated with Vifa XT tweeter.

        Comment

        • ThomasW
          Moderator Emeritus
          • Aug 2000
          • 10933

          #5
          Jon did a design using the D6.8 as a low cost DIY project for some of the children in the family.

          Our data regarding the new RS drivers comes from being involved in the beta testing.

          The Arvo schematic should provide insight as Jon's idea of a XO design for low distortion woofers in a 3-way system.

          IB subwoofer FAQ page


          "Complicated equipment and light reflectors and various other items of hardware are enough, to my mind, to prevent the birdie from coming out." ...... Henri Cartier-Bresson

          Comment

          • GrahamT
            Senior Member
            • Oct 2004
            • 378

            #6
            Originally posted by ChrisGraeff
            One other note: it appears Jon has used a poly cone in a recent design, the Hi Vi D6.8 mated with Vifa XT tweeter.
            The ExtremeMT aswell, but I sort of twisted his arm for that one. :B But that driver works very well in that specific application. It is probably the widest range, highest output (lowest efficiency) 2way 7" sealed bookshelf design.

            Comment

            • Scott Simonian
              Senior Member
              • Nov 2004
              • 216

              #7
              And has much more extention than any other 2-way bookshelf.
              My Sound Splinter 18's each in 25cuft boxes w/ EP2500

              Comment

              • AJINFLA
                Senior Member
                • Mar 2005
                • 681

                #8
                Here's some Dynaudio's: http://www.woofersetc.com/index.cfm?...roduct_ID=1573









                Cheers,

                AJ

                Manufacturer

                Comment

                • David G
                  Senior Member
                  • Jan 2005
                  • 170

                  #9
                  Originally posted by ChrisGraeff
                  Over the past few years I have auditioned a number of speakers spanning the universe of prices. One of my favorites was the Eggleston Andra - version one. While not the most uncolored speaker I've heard it's performance from 1000 hz down was excellent. The Dynaudio 30W driver handled the lowest frequencies with authority and musicality. Further the dynamic range was impressive, easily handling live Diana Krall and Mickey Hart's Planet Drum at realistic play-back levels.

                  Having read the Stereophile article, which was glowing, I was skeptical based on 1) their habit of loving exciting (read as: colored) sound, and 2) the freq response graphs showing a massive suck out from ~1500 - 3000 hz. To my surprise the human voice seemed very natural. On the other hand the piano on the Diana Krall disc sounded unnatural. When I commented to a friend accompanying me the best description we developed was plastic sounding. I know not very enlightening.

                  Regardless of this limitation I have been contemplating a pseudo-clone. The mids are Morel MW166 - no problem acquiring them in 4 or 8 ohm variety. Dynaudio drivers are unavailable save for the eBay route so I'm excluding them as a possibility. The Morel Supremes received an extremely favorable report in Voice Coil and owe some of their lineage to Dynaudio. At this point they are my selection for the tweeter. The 30W is an undersized 12 inch driver with an Sd of 62 sq inches (400 sq. cm) according to Dynaudio literature. Morel does not have an equivalent. Eton's 11 inch looked promising but not from an economic perspective. The new Dayton RS 10 and 12 hi fidelity subwoofers look like they may fit the bill. Low Le and the first cone break up mode is above 1500 hz. This should allow it to work with a 2nd order low pass at 150 - 200 hz with either a notch filter or cascaded with another 2nd order low pass filter at ~600 - 800 hz.

                  It is my contention an optimized XO can make this driver combination something special. While I have not modelled anything in LspCad my first thought is to 3rd order high pass the Supreme at ~1900. The upper MW166 will receive a 2nd order low pass at a similar frequency. The lower MW166 will be crossed in at 2nd order at ~400 - 600 effectively forming a 2.5 way speaker. Both MW166s will be high-passed at 150 -200 hz using a first or second filter. The Dayton 12HF will be XO'ed as mentioned above and loaded in an vented box. They will not be employed an isobaric arrangement.

                  Any feedback regarding driver selection or XO choice would be appreciated. Even those loaded with caveats or negative comments! Note this project, other than enclosure and XO modeling, probably will not start until end the end of this year.
                  Why use 2 midranges? I think it's unnecessary and you'd be better off using a single higher quality driver true midrange - e.g Audiotechnology 15h5206. it will have far less vertical lobing issues than 2 mids.The Mw166 has a fairly high moving mass and low efficency (86dB). It's at it's weakest in the 1-3kHz region.
                  Personally I prefer light poly or paper cones in the midrange over rigid cones, but that's due to my preference for shallow crossovers. If you are planning on using 24dB/oct filters then something like a Seas excel may be preferable.

                  I would not use a 1st order highpass on the mid, because the impedance resonance will cause a hump. Go for a low Q 2nd order highpass at around 300Hz which eliminates the hump issue, makes component values manageable and is easier in terms of baffle-step compensation.

                  The Dayton RS12 looks promising or 2xRS10s to bring up efficiency and overcome baffle step.

                  As for the tweeter, I use the Morel Supreme and it's very dynamic and natural, although I've heard others with slightly more air. Do you have a copy of that voicecoil article you could email me? I'm very interested in their findings.
                  Cheers
                  David

                  Comment

                  • ChrisGraeff
                    Junior Member
                    • Sep 2005
                    • 11

                    #10
                    Dave - As I mentioned so far nothing has started, i.e. it's in the thought experiment phase. It's my understanding a TMM in a 2.5 configuration greatly ameliorates the lobing issue. The second MW166 will keep the sensitivity up as it handles the losses due to baffle step. Employing a 1st order high-pass on the mids is based directly on my listening session with Andra V 1; the mids in that design are unfiltered and I was not able to detect any compression related artefacts so I'm surmising a 1st order filter should be adequate.

                    Using two RS 10s (RSS265HF) will not bring up the overall sensitivity as they are 4 ohm drivers and cannot be paralleled. Indeed as I review the RS 12 (RSS315HF) I realize it's sensitivity is 89 db/2.83v/1m. With full baffle step roll off that drops it down to 83 db. Unless the ground plane loading adds another 4 db I'm not certain this driver will work for me.

                    I can scan the Voice Coil article for you though it may take a few days as I'm on vacation tomorrow.

                    Comment

                    • cjd
                      Ultra Senior Member
                      • Dec 2004
                      • 5570

                      #11
                      Get a real amp.

                      You sure can parallel the RS10 or RS12 sub.



                      You can also bi-amp even with passive crossovers and adjust the gain levels to flat output.

                      C
                      diVine Sound - my DIY speaker designs at diVine Audio

                      Comment

                      • ChrisGraeff
                        Junior Member
                        • Sep 2005
                        • 11

                        #12
                        Sure you can parallel anything however most amps do not care to drive a 2 ohm impedance.

                        Bi-amping will work but not within my price constraints. Good idea, though.

                        Comment

                        • TacoD
                          Super Senior Member
                          • Feb 2004
                          • 1080

                          #13
                          Originally posted by David G
                          Why use 2 midranges? I think it's unnecessary and you'd be better off using a single higher quality driver true midrange - e.g Audiotechnology 15h5206. it will have far less vertical lobing issues than 2 mids.The Mw166 has a fairly high moving mass and low efficency (86dB). It's at it's weakest in the 1-3kHz region.
                          Personally I prefer light poly or paper cones in the midrange over rigid cones, but that's due to my preference for shallow crossovers. If you are planning on using 24dB/oct filters then something like a Seas excel may be preferable.

                          I would not use a 1st order highpass on the mid, because the impedance resonance will cause a hump. Go for a low Q 2nd order highpass at around 300Hz which eliminates the hump issue, makes component values manageable and is easier in terms of baffle-step compensation.

                          The Dayton RS12 looks promising or 2xRS10s to bring up efficiency and overcome baffle step.

                          As for the tweeter, I use the Morel Supreme and it's very dynamic and natural, although I've heard others with slightly more air. Do you have a copy of that voicecoil article you could email me? I'm very interested in their findings.
                          Cheers
                          David
                          www.gattiweb.com
                          I also have used the Morel Supereme 110 in two projects, it has great dynamics. In comparison with a Millenium it has more punch around 2 kHz, but due to this, it is less neutral than the Seas. A bit more foreward I would say.

                          The reason why I used the Supreme is that the match with Audiotechnology is very good. I did a 2-way and an 2.5-way both with 18H52. For the latter I paired the AT with an Eton 7" Hex. All poly cones are not stiff enough to handle the low frequencies without coloration.

                          I would go for a proper 3-way with a 10" (if you have the space for it), I have made several 2.5-way setups and for me the compromise is to big. You have the drawbacks of a 2-way (not extra bass extension, less airy mid) and not the extra resolution what a dedicated mid has.

                          Comment

                          • ChrisGraeff
                            Junior Member
                            • Sep 2005
                            • 11

                            #14
                            Hi TacoD - Not sure what you mean by "I would go for a proper 3-way ..."? How is it my design would not yield a 3-way? I'm using two mids in 2.5 fashion to handle the BSC and they will be bandpass filtered. And the low end will be covered by a driver dedicated to those frequencies.

                            Comment

                            • jdybnis
                              Senior Member
                              • Dec 2004
                              • 399

                              #15
                              How is it 2.5 way if you're bandpassing the mids? I think that's the definition of a 3-way.
                              -Josh

                              Comment

                              • ChrisGraeff
                                Junior Member
                                • Sep 2005
                                • 11

                                #16
                                The mids are not covering the same range from a low to high perspective. One will run from ~150-200 to 1900, the other from the same low to ~400 - 600 effectively acting as a BSC (baffle step compensation) unit. The system as a whole is a 3-way albeit a slightly unusual one.

                                Comment

                                • Dennis H
                                  Ultra Senior Member
                                  • Aug 2002
                                  • 3798

                                  #17
                                  That's called a 3.5-way.

                                  Comment

                                  • David G
                                    Senior Member
                                    • Jan 2005
                                    • 170

                                    #18
                                    Originally posted by ChrisGraeff
                                    Dave - As I mentioned so far nothing has started, i.e. it's in the thought experiment phase. It's my understanding a TMM in a 2.5 configuration greatly ameliorates the lobing issue. The second MW166 will keep the sensitivity up as it handles the losses due to baffle step. Employing a 1st order high-pass on the mids is based directly on my listening session with Andra V 1; the mids in that design are unfiltered and I was not able to detect any compression related artefacts so I'm surmising a 1st order filter should be adequate.

                                    Using two RS 10s (RSS265HF) will not bring up the overall sensitivity as they are 4 ohm drivers and cannot be paralleled. Indeed as I review the RS 12 (RSS315HF) I realize it's sensitivity is 89 db/2.83v/1m. With full baffle step roll off that drops it down to 83 db. Unless the ground plane loading adds another 4 db I'm not certain this driver will work for me.

                                    I can scan the Voice Coil article for you though it may take a few days as I'm on vacation tomorrow.
                                    Using the lower mid in such a narrow band (200-700Hz?) will cause quite a lot of bandpass gain, and as you've pointed out, having a single bass driver will bring your sensitivity way down to 83dB.
                                    I'd prefer to have 2 x 8 or 10" bass drivers giving you 89dB, meeting a 89dB mid at 300Hz or so.

                                    Comment

                                    Working...
                                    Searching...Please wait.
                                    An unexpected error was returned: 'Your submission could not be processed because you have logged in since the previous page was loaded.

                                    Please push the back button and reload the previous window.'
                                    An unexpected error was returned: 'Your submission could not be processed because the token has expired.

                                    Please push the back button and reload the previous window.'
                                    An internal error has occurred and the module cannot be displayed.
                                    There are no results that meet this criteria.
                                    Search Result for "|||"