Announcement

Collapse
No announcement yet.

How to measure total system with real XO designs vs. driver responses for XO sims

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

  • How to measure total system with real XO designs vs. driver responses for XO sims

    Hi Guys,

    I'm working through my SevenSixTwo coaxial design, and I've got my first prototype of the XO design put together. This brings me to a new question in my speaker building and designing career. Do I measure the system differently when evaluating the system response with the XO vs. measuring individual response of the drivers to start my XO designs? My goal would be toward final system characterization where folks measure on-axis and then, 15, 30, 45, 60 deg off axis. I think I've traditionally seen single measurements of each case (far field only?) as opposed to taking near field and far field of each driver and combining them like we do for XO designs.

    I raise this question because I took measurements last night of the system response, and I'm a bit confused by the 350Hz dip. This wasn't predicted in my initial PCD work, and I'm wondering if my measurement methods are at play. The image below is my OmniMic measured response at 30 inches from the speaker, on-axis with the Coax/tweeter. I loaded the FRD into PCD as an overlay since I'm at work and don't have the full simulation file from PCD to compare against right now.

    So ultimately, my question is this: What's the right way to do these "final measurements" of a system response?

    Thanks!

    Click image for larger version

Name:	762_SystemResponse_Edited.png
Views:	1
Size:	64.4 KB
ID:	1368427
    Voxel Down Firing with Dayton SA70
    Translam Subwoofers - The Jedi Mind Tricks
    The Super Bees - Garage 2 way
    SevenSixTwo - InDIYana 2018 Coax

  • #2
    What is your time gate for this measurement?

    Comment


    • #3
      Originally posted by 1100xxben View Post
      What is your time gate for this measurement?
      As I was fiddling around with OmniMic last night, I used the blended setting. I believe the gating window was 5ms, so I don't exactly trust the results below 500 - 800 Hz. That's why I'm asking how most people usually do this.
      Voxel Down Firing with Dayton SA70
      Translam Subwoofers - The Jedi Mind Tricks
      The Super Bees - Garage 2 way
      SevenSixTwo - InDIYana 2018 Coax

      Comment


      • #4
        I suppose it's worth calling myself lazy too I was kind of hoping that there was an accepted method to measure far-field only and have good results for the whole frequency range. If I were doing near field, I'd have to point at the woofer instead of just at the tweeter. That's part of what's tripping me up, since how can I measure farfield for the whole system then measure nearfield at just the woofer to get low frequency response correctly?

        Thanks for all the help and for trying to guide me to water. I'm sure I'm over complicating this topic.
        Voxel Down Firing with Dayton SA70
        Translam Subwoofers - The Jedi Mind Tricks
        The Super Bees - Garage 2 way
        SevenSixTwo - InDIYana 2018 Coax

        Comment


        • #5
          How high was the speaker off the floor? That dip looks like floor bounce cancellation.
          Craig

          Comment


          • #6
            Originally posted by PWR RYD View Post
            How high was the speaker off the floor? That dip looks like floor bounce cancellation.
            Hey Craig, good thought. The coaxial driver axis was ~64 inches off the floor. There's 8 inches between woofer axis and coaxial driver axis, so the woofer is at 56" off the floor.

            Click image for larger version

Name:	20180306_221449.jpg
Views:	2
Size:	452.5 KB
ID:	1368435
            Voxel Down Firing with Dayton SA70
            Translam Subwoofers - The Jedi Mind Tricks
            The Super Bees - Garage 2 way
            SevenSixTwo - InDIYana 2018 Coax

            Comment


            • #7
              Also your measured SPL level looks awefully low. I'd expect the noice floor of your house (furnace, fridge, etc) would cause some measurement issues. Put in a pair of ear plugs and crank it up to 95 dB @ 30 inches and throw a big couch pillow on the floor between the speaker and the mic.
              Craig

              Comment


              • #8
                Originally posted by PWR RYD View Post
                Also your measured SPL level looks awefully low. I'd expect the noice floor of your house (furnace, fridge, etc) would cause some measurement issues. Put in a pair of ear plugs and crank it up to 95 dB @ 30 inches and throw a big couch pillow on the floor between the speaker and the mic.
                Funny you mention that. I cranked the total system response up to 85 dB last night with OmniMic pink noise. There's a significant sensitivity difference between the coax and the esoteric woofer, which required me to put plenty of resistance in the mid and high frequency parts of the coax crossover circuit to even things out. I had my SMSL Q5 Pro amplifier set to volume "40", which is higher than I've ever had to push a speaker to get some output. I'd be interested to measure final system sensitivity when this is all said and done.
                Voxel Down Firing with Dayton SA70
                Translam Subwoofers - The Jedi Mind Tricks
                The Super Bees - Garage 2 way
                SevenSixTwo - InDIYana 2018 Coax

                Comment


                • #9
                  For grins, I repeated the measurement with the SMSL Q5 Pro set to "53" volume, or 95 dB on pseudo-noise sweep. I also stuck some pillows on the floor near the microphone, and made sure to adjust my gate to the latest point that still excluded any impulse irregularities. I still see that ~5 dB dip at 350 Hz. Here's the full results.

                  Click image for larger version

Name:	Full range FRD - as bought inductors - SMSL at 53 - 95dB.PNG
Views:	1
Size:	97.8 KB
ID:	1368458 Click image for larger version

Name:	20180308_203743.jpg
Views:	1
Size:	512.4 KB
ID:	1368459
                  Voxel Down Firing with Dayton SA70
                  Translam Subwoofers - The Jedi Mind Tricks
                  The Super Bees - Garage 2 way
                  SevenSixTwo - InDIYana 2018 Coax

                  Comment


                  • #10
                    Try to reproduce your original measurement conditions, mic distance, speaker height, track to check your sim. I'm not too concerned with dips below 400Hz as I always have room issues. You may want to look off axis to see if you made any spikes up.
                    John H

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

                    Comment


                    • #11
                      From the pictures, it looks like your woofer is about the same distance from the floor,and 2 walls. I'd move the setup a few feet to the right, and see if anything changes. I assume that your ceiling is about 10'. ????

                      Comment


                      • #12
                        Just noticed the blip in you inpulse window at 3.75 ms?
                        John H

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

                        Comment


                        • #13
                          Originally posted by jhollander View Post
                          Just noticed the blip in you inpulse window at 3.75 ms?
                          Me too. That looks like a reflection.
                          Electronics engineer, woofer enthusiast, and musician.
                          Wogg Music

                          Comment


                          • #14
                            Thanks John, Wogg, and rpb! I did see that impulse blip at 3.75 ms, but I thought that was outside the window. I can scoot the gate time over to make sure it's not included in my next iterations. Is there a rule of thumb you guys use to decide how close to the gate is good enough? I'm trying to get as much resolution in the measurement below 1 kHz as possible without getting many room influences either.

                            As for the room size and test location questions, I had not considered whether the speaker is placed at equal lengths from the rear and side walls before. I'll check that tonight. The ceiling in this basement is ~8.5 feet tall. Trying to get the drivers exactly 1/2 way between the ceiling and the floor is a challenge without doing a custom height measurement stand, so doubling up on the Dayton stands and using a bigger footprint piece of plywood as the base was the best I could quickly cobble together.
                            Voxel Down Firing with Dayton SA70
                            Translam Subwoofers - The Jedi Mind Tricks
                            The Super Bees - Garage 2 way
                            SevenSixTwo - InDIYana 2018 Coax

                            Comment


                            • #15
                              The only way I've been able to get good low frequency measurements (needed for a low xo in a 3 way) is by taking measurements either outside in my backyard or in our Church sanctuary which has an 18' ceiling. In both cases I placed the speaker up on a 6 foot ladder.
                              Craig

                              Comment

                              Working...
                              X