Announcement

Collapse
No announcement yet.

What is this (nearfield port response)?

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

  • What is this (nearfield port response)?

    Has anyone ever seen and thought about this before: why is the measured nearfield port response "broader" past the port resonance (27Hz) than that predicted by box modelling (see question mark on bottom plot).

    Here, the box model includes port loss (Qp=20) and enclosure loss (Qb=40), and the measured impedance curves (pink) are in complete agreement with the modelled ones. The driver response is also well-modelled past the port resonance.

    What is that?

    Action speaks louder than words but not nearly as often. -- Mark Twain

  • #2
    Re: What is this (nearfield port response)?

    I'm not sure... possibly due to the choice of internal damping material and the quantity that was used??? (or lack there of)

    What's inside this box in the way of damping?
    ~Marty

    Baby Eidolons
    Sapphos
    Cables (Post #54)
    Other speakers (Post #21)
    Design Thoughts (Posts: 6,10,13,33,35)
    Boundary Augmentation
    Dispersion/Interference

    Comment


    • #3
      Re: What is this (nearfield port response)?

      Originally posted by ReissM View Post
      I'm not sure... possibly due to the choice of internal damping material and the quantity that was used??? (or lack there of)

      What's inside this box in the way of damping?
      For those measurements, there was very little internal damping material. The enclosure is a 46" (internal) floorstander with the top 15" divided diagonally into a sealed mid chamber. Thus, the horizontal cross-section is rectangular for the bottom 31", and triangular for the top 15". I thought this was a sensible way to avoid a strong vertical resonant mode. The walls of the triangular top section are lined with OC703 rigid fiberglass. I have experimented with adding and subtracting polyfill and acoustic foam to the bottom 31" and it has no effect on the port response curve below about 100Hz. Above 100Hz the damping material has the expected effect of smoothing out the response.

      So, in short, most of the enclosure is unlined/unfilled. Only the top bit is lined with fiberglass.
      Action speaks louder than words but not nearly as often. -- Mark Twain

      Comment


      • #4
        Re: What is this (nearfield port response)?

        I'm not the right guy for this question, but I can throw ideas at you...

        1. What if your simulation box size and the actual box size are off just a tad. Any chance that could produce this effect?

        2. It seems like a "box-tuning" issue of some kind might be able to explain this? Is there any noticeable air leakage? Are the internal edges sealed well?

        3. What if the ports are flared vs non-flared. Which are you using and which is the simulation using? That can affect the port length and therefore the tuning ofthe box, right?

        4. If you don't mind experimenting you could try going to extremes: stuff the box significantly more than you typically would... and compare the two measurements specifically for the phenomenon that you're talking about.
        ~Marty

        Baby Eidolons
        Sapphos
        Cables (Post #54)
        Other speakers (Post #21)
        Design Thoughts (Posts: 6,10,13,33,35)
        Boundary Augmentation
        Dispersion/Interference

        Comment


        • #5
          Re: What is this (nearfield port response)?

          The model doesn't take into account resonant modes caused by the box dimensions and reflected waves exiting via the port, so that's the source of the discrepancy. Damping will smooth a lot of it above 100Hz, but damping is pretty ineffective below 100Hz unless you fully fill the box.
          www.billfitzmaurice.com
          www.billfitzmaurice.info/forum

          Comment


          • #6
            Re: What is this (nearfield port response)?

            Originally posted by billfitzmaurice View Post
            The model doesn't take into account resonant modes caused by the box dimensions and reflected waves exiting via the port, so that's the source of the discrepancy. Damping will smooth a lot of it above 100Hz, but damping is pretty ineffective below 100Hz unless you fully fill the box.
            Yes, I agree with that last sentence. My understanding is that the first enclosure resonance (vertical) occurs at about 150Hz. I believe this mode is evident from the measured SPL, and indeed it is reduced in severity with the addition of damping material.

            However, the issue is occuring in the 40-100Hz region. Isn't that well below the region where enclosure resonances are excited? It seems like the answer should be obvious but I can't see it.
            Action speaks louder than words but not nearly as often. -- Mark Twain

            Comment


            • #7
              Re: What is this (nearfield port response)?

              Originally posted by ReissM View Post
              I'm not the right guy for this question, but I can throw ideas at you...
              Hey, I appreciate your input!

              Originally posted by ReissM View Post
              1. What if your simulation box size and the actual box size are off just a tad. Any chance that could produce this effect?
              No. A reduction in box size will shift the impedance maxima upward in frequency beyond the measured values. In some sense, its the impedance curve that enables me to pinpoint the box volume.

              Originally posted by ReissM View Post
              2. It seems like a "box-tuning" issue of some kind might be able to explain this? Is there any noticeable air leakage? Are the internal edges sealed well?
              Its very well sealed (all seams are glued and internally caulked). Again, losses in general show up as reductions in impedance peaks. Increasing the losses (reducing box/port Qs) will lower the impedance peaks below measured.

              Originally posted by ReissM View Post
              3. What if the ports are flared vs non-flared. Which are you using and which is the simulation using? That can affect the port length and therefore the tuning ofthe box, right?
              The port is moderately flared at one end (and unflared inside the cabinet). I am assuming the appropriate end corrections (0.849 for the flared end and 0.613 for the unflared end).

              Originally posted by ReissM View Post
              4. If you don't mind experimenting you could try going to extremes: stuff the box significantly more than you typically would... and compare the two measurements specifically for the phenomenon that you're talking about.
              Yeah, I think that's what I am going to do.
              Action speaks louder than words but not nearly as often. -- Mark Twain

              Comment


              • #8
                Re: What is this (nearfield port response)?

                Hmmmm... well, as I stated previously I'm not an expert. Let me try one more thought...

                You're tuning the box to about 26 or 27 Hz. That might dictate a long-ish port length (depending on the diameter of the port and the box volume of course.) Is there any chance the "broadness" of the port's response is caused by some type of nonlinearity in the port? Maybe your simulation software isn't quite accounting for it?

                (I'm going with the idea that ideal ports fall into the same category as ideal capacitors. They always deviate from ideal due to some losses.)
                ~Marty

                Baby Eidolons
                Sapphos
                Cables (Post #54)
                Other speakers (Post #21)
                Design Thoughts (Posts: 6,10,13,33,35)
                Boundary Augmentation
                Dispersion/Interference

                Comment


                • #9
                  Re: What is this (nearfield port response)?

                  Originally posted by ReissM View Post
                  You're tuning the box to about 26 or 27 Hz. That might dictate a long-ish port length (depending on the diameter of the port and the box volume of course.) Is there any chance the "broadness" of the port's response is caused by some type of nonlinearity in the port? Maybe your simulation software isn't quite accounting for it?
                  The actual port is 7" long with 3" internal diameter. The modeling results suggest that Vb=78l and Fb=27Hz to match the measured impedance. The port length required to achieve this tuning turns out to be 7.2", with appropriate end corrections accounted for. So, everything is consistent EXCEPT the bloody resonance broadening.

                  Regarding nonlinearity, I can adjust the volume up and down with no change in the port SPL, so the process is linear.

                  Argh.
                  Action speaks louder than words but not nearly as often. -- Mark Twain

                  Comment


                  • #10
                    Re: What is this (nearfield port response)?

                    How close are you placing the microphone for the port measurement?
                    ~Marty

                    Baby Eidolons
                    Sapphos
                    Cables (Post #54)
                    Other speakers (Post #21)
                    Design Thoughts (Posts: 6,10,13,33,35)
                    Boundary Augmentation
                    Dispersion/Interference

                    Comment


                    • #11
                      Re: What is this (nearfield port response)?

                      Originally posted by ReissM View Post
                      How close are you placing the microphone for the port measurement?
                      The shape of the port response does not change below about 100Hz as the mic is moved from flush with baffle to, say, 1" inside port.
                      Action speaks louder than words but not nearly as often. -- Mark Twain

                      Comment


                      • #12
                        Re: What is this (nearfield port response)?

                        Try measuring the box with no damping material first vented, then with the port sealed at the inside end - or as close as you can get without removing the driver or changing anything.

                        Report the Fl and Fh impedance when vented and the frequency of the minimum in the diaphragm response. Also the frequency of the box resonance when sealed.

                        From these can be worked out the parameters in box as opposed to free air.

                        ~40" is L/4 for ~85 Hz

                        How long is your port? ~25"?

                        T/S theory assumes the dimensions of the box are less than ~1/6-1/10 wavelength.

                        Comment


                        • #13
                          Re: What is this (nearfield port response)?

                          The modeled port response is based purely on the helmholtz resonance.

                          But, the port will "leak" the sound from the back of the cone. You'll notice the model shows the port rolling off 2nd order like a perfect filter. But in reality, you see what you get, a whole slew of extra output across the band due to the rear wave from the woofer making its way out the port.

                          Damping is mandatory to keep the stuff above 200Hz from going anywhere, including back out through the woofer cone.
                          R = h/(2*pi*m*c) and don't you forget it! || Periodic Table as redrawn by Marshall Freerks and Ignatius Schumacher || King Crimson Radio
                          Byzantium Project & Build Thread || MiniByzy Build Thread || 3 x Peerless 850439 HDS 3-way || 8" 2-way - RS28A/B&C8BG51

                          95% of Climate Models Agree: The Observations Must be Wrong
                          "Gravitational systems are the ashes of prior electrical systems.". - Hannes Alfven, Nobel Laureate, Plasma physicist.

                          Comment

                          Working...
                          X