Announcement

Collapse
No announcement yet.

Jims 3 way speaker build

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

  • I don't have an omnimic but I was looking at the owners manual, and speculating that a setting could alter the measurements. There's many adjustments. Maybe a high-pass? Didn't see one. What I saw on page 69 mentioned an auto gain function that should be unchecked.

    With HOLM.... You can start a sweep at a higher frequency so that the woofer doesn't unload below tuning. Perhaps Omnimic has that feature, and a factory default setting.

    If that mic works with any computer, I'd download HOLM, and see what differences, if any, show up.

    Comment


    • jimbones
      jimbones commented
      Editing a comment
      auto is unchecked but I dont think I can change the sweep, its a prerecorded cd

    • rpb
      rpb commented
      Editing a comment
      Are there multiple tracks of the same type to choose from?

    • rpb
      rpb commented
      Editing a comment
      Is there a high-pass setting? Sweeping a ported speaker might require one to protect the woofer.

  • DeZZar this is a 4 inch flare one side only. attached are the exported zma for the system. the screenshot of it with phase (scary low impedance) and exported frd for woofer and port as requested.
    I have a suspicion that the braces are interfering with internals of the cabinet. Maybe I am way off here.....
    Attached Files

    Comment


    • DeZZar
      DeZZar commented
      Editing a comment
      Mate I have to be honest and say I have never seen a raw driver in cabinet impedance sweep look like that! Something's not right. This is without crossover parts in place?

      It should essentially look just like the example.one I posted earlier.

    • rpb
      rpb commented
      Editing a comment
      I thought the holes in your brace looked small, but someone said they were fine.

  • Somethings off....This post here: https://techtalk.parts-express.com/f...01#post1492201 has a normal looking impedance for a ported woofer. Its absolutely uncorrelatable with the above result and it should be very similar - just with a change in where the deep null is positioned and perhaps some small changes to that resonance spike.
    Constructions: Dayton+SB 2-Way v1 | Dayton+SB 2-Way v2 | Fabios (SB Monitors)
    Refurbs: KLH 2 | Rega Ela Mk1

    Comment


    • jimbones
      jimbones commented
      Editing a comment
      The stuffing did that

    • DeZZar
      DeZZar commented
      Editing a comment
      No, not at all Jim - the overall shape of the impedance curve will be the same - the differences will be fairly minor - a shift in the tuning point, change in peaks etc - not what we're seeing here. You could 100% stuff the box and still not see this sort of change. There's no chance stuffing alone is going to suddenly introduce a 1.5ohm minimum dip in the woofers response where it was perfectly smooth and showing 9ohms without the stuffing.

      You said 'bypass' the crossover parts - are they still connected to the driver?

    • jimbones
      jimbones commented
      Editing a comment
      Dezzar I reconnected the XO, I will do it again woofer alone

  • zma and jpg woofer no xo
    Attached Files

    Comment


    • Originally posted by jimbones View Post
      zma and jpg woofer no xo
      There we go. You can't really bypass the crossover parts - if they are still connected to the woofer in any way - they'll have an affect.

      So we can see now that the port length with all final stuffing in place is tuning the box to what looks like 38-39hz - a little too high.

      You should run the woofer and port frequency response again with the crossover parts completely disconnected. We're looking for a response pattern that shows a deep null in the woofer response around 38hz. With this tuning in a 70L net volume box we should see an f3 around the mid to high 30's with a fairly humped up response around 50hz.

      If we see an f3 much much higher than this then it might be reasonable to conclude that either the woofer needs A LOT of break in to reduce compliance and fs OR for whatever reason its seeing a smaller volume, in effect, than intended. Maybe the size of the holes in the brace although I'm not so sure about that - you could try increasing the number of holes or the size of the existing ones.
      Constructions: Dayton+SB 2-Way v1 | Dayton+SB 2-Way v2 | Fabios (SB Monitors)
      Refurbs: KLH 2 | Rega Ela Mk1

      Comment


      • jimbones I'm not sure what the Omnimic options are but see if you can display down to 10hz or when 5hz. It won't matter for crossover work but will show more of the woofer and port roll off.

        Comment


        • OK so I was able to create more free flowing space by opening up the windows on the braces while not totally mutilating them ha ha. I did the zma over with the recommendation of setting the low f mark lower. I will do the frd next Click image for larger version

Name:	zma remove braces stuff.jpg
Views:	123
Size:	129.2 KB
ID:	1492945

          Comment


          • things look like they are going in the opposite direction lol The zma is woofer only (no XO) the frd include a 8.4 mh and 80uf 2nd order
            Attached Files

            Comment


            • DeZZar
              DeZZar commented
              Editing a comment
              Well....I don't know what to make of this. Impedance looked better in post 214 above where that peak @ 150hz was mostly subdued. Now its back in full force. The aligned sum between the attached port and woofer response shows an f3 all he way up at 55hz and a steep roll-off from there. Port tuning looks to be on or just over 40hz. I'm really struggling to trust these measurements and I have no experience with OmniMic...

              From here all I can recommend is:
              - Ensure the box is sealed up tight, no leaks and that the internal surfaces are all padded with felt. Top and bottom surfaces should have a decent amount of wadding in place.
              - Re-confirm your NET internal volume is in fact 70L and if it is, lengthen the port until your impedance shows box tuning is at 32Hz.
              - Trust that the SPL of the drivers is per their spec and try the crossover I suggested which assumes they are playing at expected levels and that the woofer will result in an SPL target of 85db and see if it creates a better balanced sound overall (if you just plug my values into your sim it'll look strange because I have scaled your files to match an expected output level)

          • My bad, when I modified the braces I didn't put the stuffing back. I spoke to madisound and they said with that amount of bass energy I should have a decent amount of stuffing behind the woofer. I am sick today but will correct and measure

            Comment


            • I stuffed the area behind the woofer pretty good. I measured the woofer and port again. Also I am trying to break the woofer in by playing some tones on repeat on the cd player.
              Attached Files

              Comment


              • zma bypassing XO
                Attached Files

                Comment


                • I could be wrong, but it looks like your output at 20hz has increased quite a bit! Like maybe 6dB.

                  Comment


                  • Originally posted by rpb View Post
                    I could be wrong, but it looks like your output at 20hz has increased quite a bit! Like maybe 6dB.
                    yea the graphs look quite different. Is there anything else I need to do WRT the box and tuning before I even consider moving on to the XO??

                    Comment


                    • BTW I tried summing the 2 curves to see how they would sum but I am not getting it. any instruction is appreciated.

                      Comment


                      • I was going to stay out of the way while you and DeZZar got the tuning, and bass figured out. It looks to me like you are tuned to about 33hz. Is that where you want to be?

                        You have a filter on the woofer, mid, and tweeter, right?

                        The mid, and tweeter were padded down more, right? Check your sim. The mid should be padded about 4dB lower than its raw measurement. Tweeter is then padded to match the mid. How does the impedance look? Safe?

                        How does the mid response look now? Smooth like before I hope.

                        I'd check to see how the woofer and mid sum together. Measure from about 12", with the mic half way between them. Try switching polarity on one driver. You want flat one way, and a null the other polarity. The null would be at the xo point, but can be very wide. The theory is, if they are in phase, and sum well, when you reverse the polarity, they will cancel where they overlap. This is not an exact procedure, but it's what I do to get in the ballpark.

                        Comment


                        • jimbones
                          jimbones commented
                          Editing a comment
                          33 is what I was aiming for. The things that affected it the most as far as I can see was the woofer break-in and stuffing. Yes the woofer has the XO on it. I measured it both ways. I did have to pad the mid and tweet quite a bit. I think I padded just a bit too much now but thats OK I will back some out. I will also do as you suggest with the mid and the woofer and reverse polarity looking for a null. Happy Thanksgiving All!
                      Working...
                      X