Announcement

Collapse
No announcement yet.

Anarchy and rs28f m/t my first speaker design

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

  • #31
    when i did the measurements i didn't move the mic or speaker and i didn't adjust the amp. xsim doesn't let you adjust x and y, would that affect the offset?

    Comment


    • #32
      Of course it does! A driver 6.75" below the tweeter is further from the mic than one that is straight ahead. What you are measuring and determining is the time of flight from the speaker to mic which is included in the phase of your measurement ( or rather, the difference in time of flight between the drivers). This would explain why you determined a longer offset if you haven't included the y axis dimension. This in itself isn't so much of a problem for your simulated crossover on axis, but without the x and y speaker coordinates you won't be able to simulate off- axis response
      "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
      exactly a Fc 3K for both drivers, anybody can do it." -Xmax

      Comment


      • #33
        Originally posted by joe_bernardi43 View Post
        i retook measurements and merged port and near-field with the far-field. is there any way to verify the phase tracking besides the reverse null? when i was figuring out the relative acoustic offset i came up with 2.75 inches does that sound right?
        When you merge near and far field measurements, you'll want to complete a baffle diffraction simulation add the baffle diffraction to your nearfield measurement so that the baffle is included in your result. This is necessary to account for the cabinet in the crossover design.

        If you use FRD Response blender to merge the measurements, it can apply the baffle diffraction right in the spreadsheet as part of the process.
        "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
        exactly a Fc 3K for both drivers, anybody can do it." -Xmax

        Comment


        • #34
          Originally posted by dcibel View Post
          Using the files provided, I found your z-axis offset to be -1.7" on the woofer. For finding the acoustic offset, you would place the mic at the tweeter axis, about 20" is probably okay for distance. Then without moving the mic, measure the tweeter and woofer separately (include phase), then measure both together. Import the separate measurements to your software, enter in the known values for mic location and driver x and y axis offsets. Then adjust the z-axis offset until the simulated combined response matches your measurement. If you've done this correctly in your measurements, then the acoustic offset is -1.7" by my measure. In my simulation, the individual measurement for the woofer appears to be at slightly higher amplitude than the combined response, which makes me question the measurement method a bit. [ATTACH=CONFIG]n1323216[/ATTACH]




          As I said, since you have the mic and it appears that you've already constructed a crossover, verifying your offset should be as easy as taking measurements and verifying them against your simulation.

          In your post above (#23) where you are trying to show the acoustic offset measurement, I have to wonder why this chart is so smoothed, as the smoothing is not evident in your measurements. If there is any smoothing applied by xsim, you should disable it.

          Since you already posted your findings, I won't bother posting mine as they are identical. Same offset I came up with.
          OP hope this helps.
          https://www.facebook.com/Mosaic-Audi...7373763888294/

          Comment


          • #35
            Your sim looks pretty good to me, but try these changes as an option to consider. Change the 2.5mH to 2mH. Up the 12 ohm resistor to 22 ohms. On the tweeter, change 8 ohms to 7 ohms. Change the tweeter to 3rd order using 6 uf, a .45mH, and a 15uf cap. The x-over point moves closer to 2k.

            I arbitrarily used -.040 for the offset.

            Comment


            • #36
              Here's a sim I did using the other woofer frd file. For some unknown reason, I can't get the other frd file to work.

              About 12 parts. Several are 1.5mH coils. Just pretend the response below 500hz looks like yours.
              Attached Files

              Comment


              • #37
                i am having an issue with pcd when i load the woofer frd it shows up on the system response but when i load the zma the fr goes away. that is why i was using xsim. can i use pcd to determine the offset then use that offset in xsim?

                Comment


                • #38
                  Originally posted by joe_bernardi43 View Post
                  i am having an issue with pcd when i load the woofer frd it shows up on the system response but when i load the zma the fr goes away. that is why i was using xsim. can i use pcd to determine the offset then use that offset in xsim?
                  If you're using xsim you should continue to use the offset if 2.75" that you've already determined, since this is the offset distance that includes the y-axis.

                  I fixed your zma files for you so they will load in PCD (see attached), however as I said the file where you've combined near and far field measurements is missing the phase information, so you'll have issues using this data regardless. If you want accurate results you will want to follow my instruction in post #33 above.

                  If using PCD, I entered a mic distance of 0.508m, woofer y-axis offset of -0.17, and woofer z-axis offset of -0.043. You will find that the offset response will match what I posted previously if you enter these values. For crossover off-axis simulation you may enter the mic distance at any value that you wish, but be sure to enter the piston diameters.
                  Attached Files
                  "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
                  exactly a Fc 3K for both drivers, anybody can do it." -Xmax

                  Comment


                  • #39
                    thank you to everyone i appreciate all the help. what was wrong with the zma files? the blended fr did have the baffle diffraction add using the blender spread sheet. i also extracted minimum phase using the blender spread sheet before saving it to an frd file

                    Comment


                    • #40
                      If you've included the baffle step in there then great, however your blended file that you posted here does not contain phase. Open the file and look for yourself, it contains only 2 columns. You need to extract minimum phase from this response before it is useful. Luckily for you, I've nothing better to do at the moment so I've done this for you (see attached).

                      Open your zma files and look at line 25. They were missing a comment "*" at the beginning of this line to tell PCD to disregard this line.
                      Attached Files
                      "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
                      exactly a Fc 3K for both drivers, anybody can do it." -Xmax

                      Comment


                      • #41
                        Here's a quick crossover I did for you. You may tweak it to your hearts desire.
                        https://drive.google.com/file/d/0B-G...ew?usp=sharing
                        "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
                        exactly a Fc 3K for both drivers, anybody can do it." -Xmax

                        Comment


                        • #42
                          I tried to use the phase extractor spread sheet to extract the phase from that frd file but it would get to 1% and then freeze and excel would not respond.

                          Comment


                          • #43
                            Originally posted by joe_bernardi43 View Post
                            I tried to use the phase extractor spread sheet to extract the phase from that frd file but it would get to 1% and then freeze and excel would not respond.
                            Ah...the truth comes out! ;)

                            I used the FRD response blender to extract phase.
                            "I just use off the shelf textbook filters designed for a resistor of 8 ohms with
                            exactly a Fc 3K for both drivers, anybody can do it." -Xmax

                            Comment


                            • #44
                              Here's a 2.5k x-over. C Click image for larger version

Name:	image_65417.jpg
Views:	60
Size:	329.4 KB
ID:	1323312 Click image for larger version

Name:	image_65418.jpg
Views:	63
Size:	352.3 KB
ID:	1323313 an the anarchy go that high cleanly?
                              Attached Files

                              Comment


                              • #45
                                If you are still using your original x-over, try this tweak. The first cap on the tweeter can be increased by paralleling another. It looks like a total of 6.5uf to 7 uf gives a better response than the 4.7uf shown. Also, the 12 ohm padding resistor could be tweaked to about 6 to 8 ohms. The tweeter x-over is rather low, so don't push it too hard, even though it can probably handle it. These changes should make a large change in the tonal balance.

                                Comment

                                Working...
                                X