Announcement

Collapse
No announcement yet.

Did I make a mistake…..

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

  • Did I make a mistake…..

    …when crossing the Dayton RS270-8 to the mid at 300 Hz 2nd order?

    I have recently, in connection with the ongoing prel design/thoughts on my next speaker project

    http://techtalk.parts-express.com/fo...in-early-phase

    learnt that this type of driver should be crossed fairly steep in order to avoid breakup modes in higher frequencies. My concern is now that my choise of 2nd order perhaps was not-so-wise.

    I supply some pictures of “the evidence” that I designed in good faith but perhaps made a mistake anyway:
    • RS270-8 response from datasheet
    • My raw woofer measurement in box with distortion levels
    • The overall design with the XO
    • A pic of the woofer in box

    I don´t see any breakup in the measurement plot. Do you?
    Should I consider redesign to a 4th order design?

    Best regards//lasse

    Click image for larger version  Name:	Measurement RS270-8 in box.JPG Views:	0 Size:	229.9 KB ID:	1454186
    Attached Files
    Perry Mason talking to his dentist:

    "Do you swear to take the tooth, the whole tooth and nothing but the tooth, so help you God?"

  • #2
    No such thing as a mistake unless it releases magic smoke.

    So, what you want to do is be sure any breakup artifacts never come into play. Frequency and slope of the crossover are two methods. I like to use a Zobel on woofers anyway, and it may help. Some use a notch filter. I try for 50dB down, but will accept 40 if it is giving difficulties with the crossover due to phase. You are about 10.

    Your third graph tells me you have not dealt with them well enough. I like the RS series for their low distortion, but a pain to tame. Worth it when done though. Don't know what mid and tweet you are using, but when I use a mid, I try to get up to around 2500 or 3K. Usually a bit cleaner as not working the tweeter so hard. Especially ones like the XT25 or HDS that are supurb above 3K.

    Comment


    • #3
      Originally posted by tvrgeek View Post
      No such thing as a mistake unless it releases magic smoke.

      So, what you want to do is be sure any breakup artifacts never come into play. Frequency and slope of the crossover are two methods. I like to use a Zobel on woofers anyway, and it may help. Some use a notch filter. I try for 50dB down, but will accept 40 if it is giving difficulties with the crossover due to phase. You are about 10.

      Your third graph tells me you have not dealt with them well enough. I like the RS series for their low distortion, but a pain to tame. Worth it when done though. Don't know what mid and tweet you are using, but when I use a mid, I try to get up to around 2500 or 3K. Usually a bit cleaner as not working the tweeter so hard. Especially ones like the XT25 or HDS that are supurb above 3K.
      Hi "tvrgeek",

      Sorry, I don´t understand. Where am I 10 dB down iso 40/50?

      Regards//lasse
      Perry Mason talking to his dentist:

      "Do you swear to take the tooth, the whole tooth and nothing but the tooth, so help you God?"

      Comment


      • #4
        Originally posted by tvrgeek View Post
        No such thing as a mistake unless it releases magic smoke.

        So, what you want to do is be sure any breakup artifacts never come into play. Frequency and slope of the crossover are two methods. I like to use a Zobel on woofers anyway, and it may help. Some use a notch filter. I try for 50dB down, but will accept 40 if it is giving difficulties with the crossover due to phase. You are about 10.

        Your third graph tells me you have not dealt with them well enough. I like the RS series for their low distortion, but a pain to tame. Worth it when done though. Don't know what mid and tweet you are using, but when I use a mid, I try to get up to around 2500 or 3K. Usually a bit cleaner as not working the tweeter so hard. Especially ones like the XT25 or HDS that are supurb above 3K.
        Gotta disagree.
        The RS270 shows some edge diffraction wiggles around 1K - 2K, but the actual breakup modes are 2k - 10 k.
        I recently finished a three way with an RS270-8 woofer, and crossed to a 2 inch dome mid around 800Hz.
        Initially thought that i had to use at least 3rd order, if not a notch for the breakup, but late in the crossover tuning
        found that the 2nd inductor on the woofer really had very little effect above my chosen Fc.

        I think I hear a difference - wow, it's amazing!" Ethan Winer: audio myths
        "As God is my witness I'll never be without a good pair of speakers!" Scarlett O'Hara

        High value, high quality RS150/TB28-537SH bookshelf - TARGAS NLA!
        SB13/Vifa BC25SC06 MTM DCR Galeons-SB13-MTM
        My Voxel min sub Yet-another-Voxel-build

        Tangband W6-sub

        Comment


        • #5
          Originally posted by donradick View Post

          Gotta disagree.
          The RS270 shows some edge diffraction wiggles around 1K - 2K, but the actual breakup modes are 2k - 10 k.
          I recently finished a three way with an RS270-8 woofer, and crossed to a 2 inch dome mid around 800Hz.
          Initially thought that i had to use at least 3rd order, if not a notch for the breakup, but late in the crossover tuning
          found that the 2nd inductor on the woofer really had very little effect above my chosen Fc.
          Hi,

          so your opinion is that 2nd order at 300 Hz could be OK?

          Regards//L
          Perry Mason talking to his dentist:

          "Do you swear to take the tooth, the whole tooth and nothing but the tooth, so help you God?"

          Comment


          • #6
            I'd say that 300Hz is unnecessarily low. I don't see any problem with 800Hz, which would greatly expand the choices for the mid driver.
            www.billfitzmaurice.com
            www.billfitzmaurice.info/forum

            Comment


            • #7
              What mid?

              Ok, if that is just edge diffraction, it still needs to be dealt with. It could also be box reflections. A steeper crossover would still help but how the drive is mounted, back relief cuts, felt moretite and all kinds of games. 300 and 3K are very good "nominal" 3 way starting values. Something like the RS100 would make a great matching mid and it can easily reach about any tweeter.

              Ignore data sheet responses once you have bought the drivers. They are ADVERTISEMENTS. Only use measured IMP and FRD. You seem to be rolling off way higher ( 65-ish) than that driver should in a cabinet sealed cabinent, closer to 50 Hz. I sure would not want that hump @ 50. I have just been "turned on" to band pass alignments. Might be another solution to both the woofer efficiency being too high and filtering that 500-750 ripple. Just a thought.

              I don't see the mid and tweeter, so I don't know where the ripples, just offset to the woofer in the green line are coming from. Is your mid cabinet of similar dimension to the woofer? What is the radius of the cabinet edge. Bigger the better. Radius of the cutout, sufficient stuffing etc. Are your boxes stuffed? What is the Qtc? What is the volume of your box?

              Comment


              • #8
                Model comes out at 45Hz in 2.8 foot HUGE box. Yours is big, but I doubt that big. STUFF! EBP around 45, so sealed seems logical. See what happens. Then try a steeper crossover.

                Comment


                • #9
                  Originally posted by tvrgeek View Post
                  What mid?

                  Ok, if that is just edge diffraction, it still needs to be dealt with. It could also be box reflections. A steeper crossover would still help but how the drive is mounted, back relief cuts, felt moretite and all kinds of games. 300 and 3K are very good "nominal" 3 way starting values. Something like the RS100 would make a great matching mid and it can easily reach about any tweeter.

                  Ignore data sheet responses once you have bought the drivers. They are ADVERTISEMENTS. Only use measured IMP and FRD. You seem to be rolling off way higher ( 65-ish) than that driver should in a cabinet sealed cabinent, closer to 50 Hz. I sure would not want that hump @ 50. I have just been "turned on" to band pass alignments. Might be another solution to both the woofer efficiency being too high and filtering that 500-750 ripple. Just a thought.

                  I don't see the mid and tweeter, so I don't know where the ripples, just offset to the woofer in the green line are coming from. Is your mid cabinet of similar dimension to the woofer? What is the radius of the cabinet edge. Bigger the better. Radius of the cutout, sufficient stuffing etc. Are your boxes stuffed? What is the Qtc? What is the volume of your box?
                  Hi,

                  Some answers:
                  Mid:Peerless 830991
                  Tweeter: SEAS 27TFFNCCG
                  Volume: Abt 28 litres
                  Fairly densely stuffed boxes
                  Cabinet edges rouded with a radius of abt 15mm

                  Pictures:
                  Stuffing
                  Mid/Tw unit
                  Drawing of woofer unit
                  Click image for larger version

Name:	IMG_0121.jpg
Views:	199
Size:	44.6 KB
ID:	1454239 Click image for larger version

Name:	IMG_0080.JPG
Views:	199
Size:	508.0 KB
ID:	1454240 Click image for larger version

Name:	IMG_0081.JPG
Views:	198
Size:	504.2 KB
ID:	1454241 Click image for larger version

Name:	IMG_0013.jpg
Views:	192
Size:	61.1 KB
ID:	1454242
                  Perry Mason talking to his dentist:

                  "Do you swear to take the tooth, the whole tooth and nothing but the tooth, so help you God?"

                  Comment


                  • djg
                    djg commented
                    Editing a comment
                    Nice dresser.

                • #10
                  Well, can't explain the ripple then. Did not flush mount the woofer? WinISD suggests Qtc .7 with 78L. ( based on the published spec, your plot suggest it may want even bigger!) That would explain the early rolloff as well as most of the hump. This is why I harp about measuring and prototyping. You could have gained a couple L in the base, and a bit of clever work could have gained a few more in the top. I would have turned the mids to get the tweeter closer and brought the woofers to the top of the box. Your sat box is probably way too big for the Peerless as a midrange. wasting more space your woofer needs. Really, that bugger needs to be a full size monkey coffin. Maybe lost a little with overkill reinforcing the baffle. ( or is that a theft deterrent as you can't lift them?)

                  Now, what to do. I might suggest a Linkwitz Transform for the woofer. I would try the various crossover mods, steeper, Zobel or notch for that ripple. Flush mount the woofer and knock off the edges on the inside. If you can get your midrange within a couple dB of ripple, you are good to go.

                  Not stuffed too tight behind blocking the pole vent are you?

                  Click image for larger version

Name:	Capture.JPG
Views:	197
Size:	121.2 KB
ID:	1454248

                  Comment


                  • #11
                    Kids, the "ripple" in the RS270-8 FR posted above is from the Parts Express data sheet! Should have been (and probably was) mounted in an IEC baffle, so not a problem with the box as built.
                    Lasse, I am sort of in agreement with Bill Fitzmaurice here. Let the woofer go a little higher, and take some stress off the mid. I'd look at about a 3.5mH and 80uF for the LP second order.
                    I put the RS270-8 in a 33 liter box, and have a lot of low end down into the 40s. F6 at 44 Hz. Sounds great. Using 600uF caps inline for a linkwitz transform.

                    I think I hear a difference - wow, it's amazing!" Ethan Winer: audio myths
                    "As God is my witness I'll never be without a good pair of speakers!" Scarlett O'Hara

                    High value, high quality RS150/TB28-537SH bookshelf - TARGAS NLA!
                    SB13/Vifa BC25SC06 MTM DCR Galeons-SB13-MTM
                    My Voxel min sub Yet-another-Voxel-build

                    Tangband W6-sub

                    Comment


                    • #12
                      I am not digging that 5 dB bump at 100Hz though. Maybe reduce the resistors on the mid and tweeter and bring them up 2-3dB. let your ears guide you, or do some 1/6 octave measurements just to determine overall octave to octave balance without getting lost in the wiggles.

                      I think I hear a difference - wow, it's amazing!" Ethan Winer: audio myths
                      "As God is my witness I'll never be without a good pair of speakers!" Scarlett O'Hara

                      High value, high quality RS150/TB28-537SH bookshelf - TARGAS NLA!
                      SB13/Vifa BC25SC06 MTM DCR Galeons-SB13-MTM
                      My Voxel min sub Yet-another-Voxel-build

                      Tangband W6-sub

                      Comment


                      • #13
                        I would say you hit the sweet spot with a 300hz xo point. That Peerless mid is very nice sounding, and crossed at 300hz , it should be coasting. With that GFC driver, I do prefer a steep xo to the tweeter, but what I see here should be fine. I've made a few 3-ways using that mid. They were all temporary affairs where I stacked boxes. My xo part count was excessive. My 2-way center channel uses it crossed very steep at about 2.7k. Any xo can potentially be improved, but at what cost. The GFC driver has a breakup around 5k and another around 10k. (Going from memory, so I may be off.) The more you suppress those, the better the driver sounds. With my center, I chose to go 4th order electrical. It's overkill, and will probably be temporary. Right now it's connected with jumpers.

                        Your roll-off is very close to matching mine. I suspect that speaker of yours sound pretty darn good.

                        Here it is.

                        Comment


                        • #14
                          Originally posted by donradick View Post
                          Kids, the "ripple" in the RS270-8 FR posted above is from the Parts Express data sheet! Should have been (and probably was) mounted in an IEC baffle, so not a problem with the box as built.
                          Lasse, I am sort of in agreement with Bill Fitzmaurice here. Let the woofer go a little higher, and take some stress off the mid. I'd look at about a 3.5mH and 80uF for the LP second order.
                          I put the RS270-8 in a 33 liter box, and have a lot of low end down into the 40s. F6 at 44 Hz. Sounds great. Using 600uF caps inline for a linkwitz transform.
                          Hello and thanks for att the feedback! Very encouraging words!

                          Just a question: How do I connect a 600uF cap in my XO, shown in a previous post, to get a passive Linkwitz transform??

                          Best regards// Lasse
                          Perry Mason talking to his dentist:

                          "Do you swear to take the tooth, the whole tooth and nothing but the tooth, so help you God?"

                          Comment


                          • #15
                            Hi Lasse -

                            Based on a Tony Gee design - the Black Box. Actually, I'm not sure that a Linkwitz transform would work for you. He purposefully put the RS270 in a too small box to create a hump around 80-90Hz, then used 600uF inline with the woofer to flatten that out and extend the low end a bit. For your design, the RS270 is in a pretty big box already. Your impedance plot shows the Fb around 45 Hz, which should be "good enough" - also the bump you have at 100 Hz may be a measurement artifact. Perhaps you can do a near field woofer measurement and splice that to the far field measurement.
                            http://www.humblehomemadehifi.com/archive.html

                            I think I hear a difference - wow, it's amazing!" Ethan Winer: audio myths
                            "As God is my witness I'll never be without a good pair of speakers!" Scarlett O'Hara

                            High value, high quality RS150/TB28-537SH bookshelf - TARGAS NLA!
                            SB13/Vifa BC25SC06 MTM DCR Galeons-SB13-MTM
                            My Voxel min sub Yet-another-Voxel-build

                            Tangband W6-sub

                            Comment

                            Working...
                            X