Announcement

Collapse
No announcement yet.

First attempt at crossover design

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

  • #16
    If two eights would be of interest, check these out. Similar size box as the single twelve.

    https://www.parts-express.com/dayton...oofer--295-484

    Comment


    • #17
      Honestly I'm torn between a nice, sleek modern design with smaller footprint and a 'vintage' giant box. The vintage one would more closely resemble the stolen ones so I guess I should shoot for that, if for nothing else but nostalgia. Plus it would be nice to show the audiosnob at work that I can do something besides bookshelf speakers. He says I have a small speaker fetish. Nice enough guy who actually does know a few things. He actually told the guys at Best Buy that a certain pair of speakers they had were wired in reverse polarity and was actually correct. Heard about that for months.

      Comment


      • #18
        Originally posted by BTRaulerson View Post
        Honestly I'm torn between a nice, sleek modern design with smaller footprint and a 'vintage' giant box. The vintage one would more closely resemble the stolen ones so I guess I should shoot for that, if for nothing else but nostalgia. Plus it would be nice to show the audiosnob at work that I can do something besides bookshelf speakers. He says I have a small speaker fetish. Nice enough guy who actually does know a few things. He actually told the guys at Best Buy that a certain pair of speakers they had were wired in reverse polarity and was actually correct. Heard about that for months.
        I'm lazy when it comes to wood work, so sat / sub systems are my chosen system type now, but I can envision a 12" 3-way tower that is 16" wide at the base, and about 6" wide at the top. Maybe 45" tall. That would just be cool. Sloped baffle too, about 1" in 10" slope. One inch round-overs on the sides, and maybe an open back mid-range chamber. Then some days, I get the urge to make an old school monkey coffin of about 5 cu-ft. Maybe someday!

        Comment


        • #19
          It is decided. I will build a monkey coffin.

          Comment


          • #20
            How about the Tarkus? Sounds like what you are looking for. https://sites.google.com/site/undefinition/tarkus

            Comment


            • #21
              I think the "Rockin’ It Old School Alternate" project is closer to what I'm looking for. Those are monkey coffins for sure! Thank you for the suggestion though. I picked out some Dayton drivers, all 4 ohm, but I'm struggling quite a bit with the crossover. I'm going to read some more and see if I can figure it out.

              Comment


              • #22
                I'm stuck. Where does the nasty spike come from at around 300 hertz? What am I missing? The sub is the Dayton Audio 15" 4 ohm DCS380-4, the mid is the DA RS150P-4 6" 4 ohm, and the tweeter is a DA RST28F-4 1 and 1/8th inch 4 ohm tweeter. I think I used a 2nd order high pass for the tweeter, 2nd order high/low pass for the mid and a 3rd order low pass for the woofer.

                Comment


                • #23
                  Are you using more than 3 drivers? These 4 ohm drivers will likely yield a 4 ohm speaker. Will the amplifier tolerate that load?

                  Your sim is too small to read, but I'd look at the impedance files of the woofer, and mid, and see if there is a odd jump in the impedance, that in reality, is not there.
                  Look at the frd files too.

                  How big is this speaker going to be?

                  Comment


                  • #24
                    3 drivers, all rated at 4 ohm. I'm ........... pretty certain the arc welder will tolerate 4 ohm. The speakers will be gigantic. Roughly 24"x24"x36". Sorry about the tiny screenshot. Not sure how to change it but I'll try.

                    Edit: I tried to change the screenshot properties.
                    Click image for larger version

Name:	Monkey Coffin.jpg
Views:	0
Size:	617.4 KB
ID:	1432178
                    Attached Files

                    Comment


                    • #25
                      That's big! Is the room big? Since they won't be tall, are you going to slope the front baffle, possibly? The x-over sims assume listening on the tweeter axis. If the tweeter is going to be 30" off the floor, the response will be different at higher positions, especially up close. If the listening position is 16 feet away, it's less of an issue. I assume these are more for hard rock, than audiophile listening, but sometimes you can accomplish both.

                      I can read the graphs now.

                      Comment


                      • #26
                        Your drivers all need their "offsets" (X, Y, Z) numbers filled in, or your phase will be all wonky (throwing your "summing" off).
                        (That means that you need to create "minimum phase" files. And I think the spike is in your woofer's .frd file.)
                        I built some 5cf speakers once. They were just TOO HUGE to be given a place in my friend's living room ! !

                        You have NO "baffle-step" in these; but... it prob'ly wounldn't be needed if the woofer's near the floor (which I'd assume it would be)?
                        You should be able to get rid of the 2k "dip" by making the mid's LP filter values slightly smaller (or the tweeter's HP filter values a bit larger).

                        Try just a 2nd order on the woofer (6mH coil & 200uF shunt cap).

                        Comment


                        • #27
                          Originally posted by BTRaulerson View Post
                          I'm stuck. Where does the nasty spike come from at around 300 hertz? What am I missing? The sub is the Dayton Audio 15" 4 ohm DCS380-4, the mid is the DA RS150P-4 6" 4 ohm, and the tweeter is a DA RST28F-4 1 and 1/8th inch 4 ohm tweeter. I think I used a 2nd order high pass for the tweeter, 2nd order high/low pass for the mid and a 3rd order low pass for the woofer.
                          Phase spikes such as those can't exist in drivers. This can be verified simply by looking at the smoothness of the raw response of the drivers at the same point as the phase anomalies.

                          I would say it's due to one (or more) of several possibilities: the data is spliced from near-field and far-field measurements that have not had phase re-generated for minimum-phase response after splicing to properly align the phase of both measurements, anomalies from interpolation after importation of the raw files into the software or graphing anomalies due to the type of curve fitting used for the continuous display of the discrete data. I've seen the third one in both the PCD and my WinPCD software. You can safely ignore those. However, I would suggest processing your raw files to re-generate minimum-phase response. Just be sure that you have reasonable driver acoustic offset estimates, required when using minimum-phase files.

                          dlr
                          WinPCD - Windows .NET Passive Crossover Designer

                          Dave's Speaker Pages

                          Comment


                          • #28
                            Use a really sturdy work table!

                            I agree with Chris on the baffle step. These drivers seem to match well with no padding, yet my concern would be that the bass might just be too overwhelming for the tonal balance that I like. A lot depends on the room placement, and personal preference.

                            Comment


                            • #29
                              I'd suggest crossing the mid at 4k, and notch the bump at 1.8k if it's still significant.

                              Comment


                              • #30
                                The bump on the woofer @ 1.7k is down by over -40dB, not to worry.
                                If you cross the mid even higher, IT's "bump" @ 5.5k will probably become a problem.

                                Comment

                                Working...
                                X