Announcement

Collapse
No announcement yet.

Help with a variation on Pete S. 3Way Peerless Design

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

  • #16
    Originally posted by jhollander View Post

    Yes
    Sorry poor grammar.

    Mr. OP - *did* you re-extract......

    Comment


    • #17
      Here is another box model, starting from scratch after a reboot so no data was left in PCD. This time it came out much better looking but I am surprised at the amount of overlap and the fact that I cannot get a nice deep null at the crossover. I am fairly certain that I shouldn't be worried about that, but when I try for it, it just doesn't happen.

      Does this model including box response look better now?

      To Dave B. I extracted -after- splicing in the box model and diffraction.
      Attached Files

      Comment


      • #18
        The woofers are running up too high. They should be -6dB at about 400hz.

        Comment


        • #19
          I tweaked it a little bit, but I see that as I try to make it more perfect the impedance in the upper bass begins dipping down to about 2.5 ohms, which worries me with the amp I current have (Akitika GT101).
          Attached Files

          Comment


          • #20
            It seems to me there is no baffle step compensation. You can see this in two places. The "midrange bulge" - where it looks like you have nice phase tracking, and the fact the midrange transfer function isn't 3+ dB below the woofer (level transfer).

            You should start with a net system sensitivity as a goal for baffle step compensation - then go from there. In this case I would target ~ 88dB assuming you've correctly modeled the woofer bass rolloff

            Comment


            • #21
              Ok edit - I was going by the graphs in post #15. your latest looks better

              Comment


              • #22
                It looks like you are shooting for 4th order woofer-midrange? Have you tried 2nd order acoustic? This might raise impedance. Lower your XO point (assuming your mid can handle it) to ~ 250Hz

                Comment


                • #23
                  Originally posted by shudson105 View Post
                  I tweaked it a little bit, but I see that as I try to make it more perfect the impedance in the upper bass begins dipping down to about 2.5 ohms, which worries me with the amp I current have (Akitika GT101).
                  My sim also drops low in the bass. I'm not sure there's a way around it.

                  Comment


                  • #24
                    I tried to model with different bass roll-offs but it turns out worse. I always end up with a resonance bump from the crossover which then requires notch filter and further complicates the crossover. Since I am using the same 830874 drivers as woofers and midrange, I don't think it's really a problem with all 3 drivers playing with so much overlap in this case. I think this model is probably the best its going to get without either a more complex crossover or at least building it and seeing what the real world project does.

                    Comment


                    • #25
                      Originally posted by shudson105 View Post
                      I tried to model with different bass roll-offs but it turns out worse. I always end up with a resonance bump from the crossover which then requires notch filter and further complicates the crossover. Since I am using the same 830874 drivers as woofers and midrange, I don't think it's really a problem with all 3 drivers playing with so much overlap in this case. I think this model is probably the best its going to get without either a more complex crossover or at least building it and seeing what the real world project does.
                      Do you already have the drivers?

                      Also, are you OK with a full baffle step compensated design? If the speakers are too close to the wall, the tonal balance might not be what you want.

                      The woofers could be wired in series.

                      l can show you my sim, but it's likely got the same issues.

                      Comment


                      • #26
                        Yes, already have the drivers. Yes, full BSC is expected as these are playing in an open basement well away from the walls. I see Pete's original design model (without box effects) also has a low 3 ohm impedance in the upper bass area.

                        Comment


                        • #27
                          Originally posted by shudson105 View Post
                          Yes, already have the drivers. Yes, full BSC is expected as these are playing in an open basement well away from the walls. I see Pete's original design model (without box effects) also has a low 3 ohm impedance in the upper bass area.
                          It's weird. It drops due to the woofer filter, not the mid, or tweeter. The tendency is for the filter to boost around 100hz also. Must be due to the relatively high resonance.

                          Comment


                          • #28
                            I'd try a damped second order. Lower the woofer cap and increase the inductor. Add a resistor after the woofer cap to damp down any peaking
                            John H

                            Synergy Horn, SLS-85, BMR-3L, Mini-TL, BR-2, Titan OB, B452, Udique, Vultus, Latus1, Seriatim, Aperivox,Pencil Tower

                            Comment


                            • #29
                              Everything is helpful, thanks for sharing.

                              Comment


                              • #30
                                I further revised the model, looks similar overall, but I think its a bit better blend of the drivers. I tried a damped second order from John's suggestion, but it lowers the sensitivity too much requiring high resistors throughout the rest of the circuit to lower everything down to the woofer level.
                                Attached Files

                                Comment

                                Working...
                                X