Announcement

Collapse
No announcement yet.

Please critique this crossover

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

  • Please critique this crossover

    So after watching online instruction with Vituix on designed a crossover with it. This is my first attempt at this. It is an MTM with a Dayton NDFA20-6 and two Dayton TCP115-8 in series. This is what i got after 6 parts. It looks reasonably flat from a distance. Opinions and advice? Thank you.


  • #2
    Go 2nd order on the woofers. You'll be able to fill in the midrange dip a bit and suppress that breakup better.
    Electronics engineer, woofer enthusiast, and musician.
    Wogg Music
    Published projects: PPA100 Bass Guitar Amp, ISO El-Cheapo Sub, Indy 8 2.1 powered sub, MicroSat, SuperNova Minimus

    Comment


    • #3
      If/when you go second-order on the woofers' crossover, you'll probably want to increase the value of the woofer's inductor a little and use a resistor+capacitor (together in series) to ground as the second-order part instead of just a capacitor by itself....just because that woofer tends to get a spike in the frequency without the resistor when adding a capacitor.

      You might also end up wanting to angle the woofers frequency downward somewhat and quiet the tweeter by 4-6db overall instead of aiming for flat if you're using the FRD and ZMA files from PartsExpress. IF you're using your own in-box measurements then flat is a good aim, but most company measurements (like from Dayton and PartsExpress) are made with the woofer mounted flush "inside a wall" so the measurement doesn't show the bass-rolloff of "BaffleStep" while your own situation with the woofer mounted in a box and probably a little forward of the wall will likely have some BaffleStep Diffraction which can roll-off the bass around 3-6db. Basically, instead of a flat frequency you end up aiming for something like this to compensate:
      Click image for larger version

Name:	VituixCAD_Diffraction.png
Views:	341
Size:	9.1 KB
ID:	1470909
      That slope's position will move lower and lower as the box-size and woofer/s get larger and larger (this is a simulation with a really small box+woofer), but VituixCAD has this built-in if you click the "Tools" and then select "Diffraction"....it's also in the Enclosure tool section (middle tab in some version) which is a cool feature by itself, kind of like WinISD in many ways.

      Something similar to the above slope (but using your own input box width/height and speaker placement) may be a good option if you feel like flat lacks bass or lacks low-mids.
      My first 2way build

      Comment


      • #4
        2nd order on the woofers, and 3rd order on the tweeter. That takes phase/polarity concerns off the table, and both increases tweeter protection and lower excursion sourced tweeter THD. I'd also be inclined to wire the woofers parallel. If your amp won't handle a 4 ohm load I'd use 4 ohm drivers in series. BTW, your picture shows the woofers in parallel, not series.
        www.billfitzmaurice.com
        www.billfitzmaurice.info/forum

        Comment


        • #5
          Originally posted by Jicafold View Post
          So after watching online instruction with Vituix on designed a crossover with it. This is my first attempt at this. It is an MTM with a Dayton NDFA20-6 and two Dayton TCP115-8 in series. This is what i got after 6 parts. It looks reasonably flat from a distance. Opinions and advice? Thank you.
          You're just practicing, right? Try a 3rd order filter on the woofer. Try to get a relatively sharp knee at about 2.5k.

          And..... Like LOUT said, pull down the mids, and top end a bit. The coil(s) for the woofer need to be larger.

          The ND20 tweeter is not going to cross low enough for an mtm. (Not ideal anyway.) If you are just practicing, try the RST28.

          Comment


          • #6
            Originally posted by rpb View Post

            You're just practicing, right? Try a 3rd order filter on the woofer. Try to get a relatively sharp knee at about 2.5k.

            And..... Like LOUT said, pull down the mids, and top end a bit. The coil(s) for the woofer need to be larger.

            The ND20 tweeter is not going to cross low enough for an mtm. (Not ideal anyway.) If you are just practicing, try the RST28.
            Yes, just practicing with this finished product in mind. I chose the ND20 beacause I already have one here on the coffee table. Thank you all for the advice to try a third order. I hadn't ventured that far yet thinking that less is more.

            My amp should handle the low impedence. It is an NAD T773 receiver at the moment.

            Lout, thank you and I need to read about that. I understand most of what you are saying but some is over my head. Just trying to learn as much as I can and do things the right way.

            Comment


            • #7
              Originally posted by rpb View Post

              You're just practicing, right? Try a 3rd order filter on the woofer. Try to get a relatively sharp knee at about 2.5k.

              And..... Like LOUT said, pull down the mids, and top end a bit. The coil(s) for the woofer need to be larger.

              The ND20 tweeter is not going to cross low enough for an mtm. (Not ideal anyway.) If you are just practicing, try the RST28.
              I see what you are saying. I had a hard time with that NDFA 20 – 6 tweeter getting the crossover point low enough to avoid the cone breakup problems with the woofer. The
              RST28 Is out of stock at the moment, not that I am in a hurry. However they do have the Dayton Audio DC28FS-8 which also goes down to 1300. Maybe that would be a good choice. I’m working on that 2nd and 3rd order crossover. It’s looking better but I wanted to put a resistor on the tweeter prior to the crossover to lower the highs but the program, PCD, won’t let me do that. Maybe I need to erase some things and start over with the tweeter. Thank you.

              Comment


              • #8
                So with second-order on the woofer and 3rd order on the tweeter with a resistor on the woofer and tweeter both I ended up with this. It’s a little better. I wish the mid range was nicer. Maybe I should stop messing with this design and try it with a different tweeter as suggested.

                Comment


                • #9
                  Originally posted by Jicafold View Post
                  So with second-order on the woofer and 3rd order on the tweeter with a resistor on the woofer and tweeter both I ended up with this. It’s a little better. I wish the mid range was nicer. Maybe I should stop messing with this design and try it with a different tweeter as suggested.
                  As has been mentioned, it will probably sound nicer with a little more downward slope on the woofer for BSC. That *might* also fill the hole around 1kHz a little.
                  Francis

                  Comment


                  • #10
                    Originally posted by fpitas View Post

                    As has been mentioned, it will probably sound nicer with a little more downward slope on the woofer for BSC. That *might* also fill the hole around 1kHz a little.
                    What do you mean by BSC?

                    Comment


                    • #11
                      Originally posted by Jicafold View Post

                      What do you mean by BSC?
                      Baffle step correction. Re-read post #3 above.
                      Francis

                      Comment


                      • #12
                        I don't see anything all that bad about that plot. If I was to change anything it would be the midbasses, a higher Fs would give better results above 800Hz, but I'd be using the cabs with subs to handle the low end. If you're using these without subs a less than perfect midrange is what you trade off in order to go lower than the average four inch woofer is capable of.
                        www.billfitzmaurice.com
                        www.billfitzmaurice.info/forum

                        Comment


                        • #13
                          Thank you. I have to read over meassge #3 a few times and get my head around that.

                          billfitzmaurice, I was going to use this in a ported box as an MTM. Here the size of 1/2 the box according the calculator. So I was going to double this, 2 ports, and add the tweeter.



                          Ported Box
                          Vas = 2.52 lts
                          fs = 59.2 Hz
                          Qts = 0.43
                          D = 93 mm
                          Dimensions

                          Vb = 0.11 ft3 = 3.11 lts
                          = L x W x H
                          f3 = 53.96 Hz
                          fb = 59.2 Hz
                          Dv = 1 in = 2.5 cm
                          Lv = 4.67 in = 11.45 cm
                          Vb: Speaker Box Internal Volume
                          f3: 3dB Cutoff Frequency
                          fb: Enclosure Resonant Frequency
                          Dv: Port Diameter
                          Lv: Port Length

                          Comment


                          • #14
                            Fine, if you're running it full range and want to get a 54Hz F3. But if you're using subs and cross to them in the 80-100Hz range you don't need a 54Hz F3, so you can use a driver that doesn't go that low and has better mids.
                            Baffle step addresses the loss of sensitivity where the baffle is less than one wavelength in dimension, ending up 6dB down an octave or so below that frequency. IMO too much attention is paid to it, and not enough to speaker placement. The reason for the sensitivity loss is the frequencies above the baffle step all fire forward, with a maximum dispersion of 180 degrees, while those below the baffle step don't all fire forward, with a maximum dispersion of 360 degrees. BSC, baffle step compensation, attenuates the frequencies above the baffle step, for flatter overall response. However, unless the speaker is outdoors the frequencies below the baffle step aren't lost. They first wrap around the cab, then they reflect back into the listening space. Depending on the speaker placement most of what wraps around the cab may end up reflected back to the listening position. That may open another can of worms due to boundary reflection sourced cancellation, explained here: http://arqen.com/acoustics-101/speak...-interference/

                            Between the one and the other it's best to measure the speaker response in position and then add BSC if required. If you can flush mount in the wall that eliminates concern over both BSC and boundary sourced cancellation, as there is no reflected wave off the wall.
                            www.billfitzmaurice.com
                            www.billfitzmaurice.info/forum

                            Comment


                            • #15
                              Originally posted by Jicafold View Post

                              I see what you are saying. I had a hard time with that NDFA 20 – 6 tweeter getting the crossover point low enough to avoid the cone breakup problems with the woofer. The
                              RST28 Is out of stock at the moment, not that I am in a hurry. However they do have the Dayton Audio DC28FS-8 which also goes down to 1300. Maybe that would be a good choice. I’m working on that 2nd and 3rd order crossover. It’s looking better but I wanted to put a resistor on the tweeter prior to the crossover to lower the highs but the program, PCD, won’t let me do that. Maybe I need to erase some things and start over with the tweeter. Thank you.
                              Look at the schematic tab of PCD. The resistor for in front of the tweeter is over on the left side. Check the label, and then look for it in the tweeter resistors box. My computer is put away, or I'd show you. Also, not a good idea to add a series resistor to the woofer.

                              Comment

                              Working...
                              X