Announcement

Collapse
No announcement yet.

Crossover help part 2

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

  • Crossover help part 2

    Hi Gentlemen,

    I hope starting this as a new thread is OK. I have done this because the original thread is somewhat buried and I am at a different stage ot the build. Since I still have not figured out how to load files into Xsim and also have not learned how to use it, I decided to try a more old school approach. I built a crossover using the woofer circuit of the Carmody Speedsters and the tweeter leg of JR's Tashas. I then played with the values of the components measuring the frequency response changes with REW. After about 8 iterations I was able to come up with somethjing that seems like a pretty good first approximation. The speaker actually sounds quite good with very nice detail, impressive, clean bass, and airy highs, but I am sure it could be greatly improved if the frequency response could be flattened more.
    So, given that I need to keep things as simple as possible due to my limited crossover building skills, are there changes in value of the components in the existing layout which might improve things? Or are there simple other changes that are worth trying? Since the crossover is outside the cabinet and I have a good variety or LRC components on hand it is quick and easy to make physical changes and measure the result.
    Pictures of the crossover I am currently using and the resulting in cabinet response are attached.

    Thanks in advance,

    Jay

  • #2
    You could go bigger on the woofer L1 (up to 1.5mH), smaller on L2 (down to .25mH), and/or add 1or 2 ohms in line with the 8.2 shunt cap. This would add BSC and soften the knee. If you do that, more tweeter padding might be needed.

    Comment


    • #3
      I don't think there's anything "wrong" w/what you've got there.

      I generally try to use an L-pad on a tweeter rather than just series resistance. It helps suppress the Z-peak at Fs.
      For something similar, I ended up w/an 8uF series cap, a 0.50mH shunt coil (so... just 2nd order), and an "L" w/: SR=4n(ohms) and PR=2n
      I also went 2nd order on the woofer: 1.5mH series coil, and a 6uF shunt cap.
      Reverse polarity is indicated on the tweeter.

      Comment


      • #4
        Thanks a lot guys. I am away from home until Jan 3 but I will try these changes when I get back. What I'm most concerned by is the dip between 3.5K and 5K. Is there any fairly easy way to correct this?

        Thanks!
        Jay

        Comment


        • #5
          Originally posted by Drjay View Post
          Thanks a lot guys. I am away from home until Jan 3 but I will try these changes when I get back. What I'm most concerned by is the dip between 3.5K and 5K. Is there any fairly easy way to correct this?

          Thanks!
          Jay
          To my eyes, the wide peak between 800hz, and 2k looks to be a problem. (Not enough BSC).

          Comment


          • #6
            Happy new year Gents. After 5 hours of trying the suggestions offered above, plus many changes in the starting crossover, I have come up with what I think is a passable (for a beginner) crossover that I plan to actually use. It is not radically diferent than the original posted above, but to me looks a bit better (the green is the original the red the revised), It also sounds very good although only longer term listening to a completed pair will tell with greater certainty,
            For those of you with real world experience, would a few more parts flatten the frequency response enough to be worthwhile and if so what would you focus on and how would you address it? As it is now it seems pretty well tonally balenced on familiar recordings and very detailed, but a bit on the analytical side. I am not sure if that is a function of the drivers or the crossover or both.
            Anyway, the help offered is much appreciated and further suggestions for changes that could let these very nice drivers show their best would be welcome. The installed crossover boards will be readily removable for changes and I'm not thinking the attached crossover is the last one these speakers will see.
            Best,
            Jay

            Comment


            • #7
              I think you are xovering way too low, especially without using a parallel resistor or lcr.
              Later,
              Wolf
              "Wolf, you shall now be known as "King of the Zip ties." -Pete00t
              "Wolf and speakers equivalent to Picasso and 'Blue'" -dantheman
              "He is a true ambassador for this forum and speaker DIY in general." -Ed Froste
              "We're all in this together, so keep your stick on the ice!" - Red Green aka Steve Smith

              *InDIYana event website*

              Photobucket pages:
              http://photobucket.com/Wolf-Speakers_and_more

              My blog/writeups/thoughts here at PE:
              http://techtalk.parts-express.com/blog.php?u=4102

              Comment


              • #8
                Wolf, could you perhaps suggest, in beginner understandable terms, changes to the circuit which would help it work better with these drivers? I thought I was crossing over at around 2K which I believed was OK with the Peerless TX SC 90, but I don't know how to select parts values and placement to actually set a crossover frequency. I have been hopeful that by starting with a crossover that was based on successful designs using these relatively easy to work with drivers in similar speakers and tweaking component values I could come up with something pretty serviceable. I also thought that, judging from the frequency response and actual sound, that what I came up with was fairly good. Is the low crossover point a problem because the tweeter will distort?
                If I am "too far out of the ball park" skill and knowledge-wise to be helped without excessive time input on your part, please ignore this post. I appreciate the help you have given in the past and don't want to take advantage of you in any way.
                Thanks!
                Jay

                Comment


                • #9
                  The SC90 prefers being xovered above 2.6kHz, due to HD and nasal sound. I would not use a cap larger than 6.8uF, nor a coil larger than 0.4mH (smaller preferred). I knew the 10uF would be too large. I would also throw a 30 ohm across it to prevent the nasality.

                  Later,
                  Wolf
                  "Wolf, you shall now be known as "King of the Zip ties." -Pete00t
                  "Wolf and speakers equivalent to Picasso and 'Blue'" -dantheman
                  "He is a true ambassador for this forum and speaker DIY in general." -Ed Froste
                  "We're all in this together, so keep your stick on the ice!" - Red Green aka Steve Smith

                  *InDIYana event website*

                  Photobucket pages:
                  http://photobucket.com/Wolf-Speakers_and_more

                  My blog/writeups/thoughts here at PE:
                  http://techtalk.parts-express.com/blog.php?u=4102

                  Comment


                  • #10
                    J, have you read any books on speakerbuilding - like "SpeakerBuilding 201"?


                    I'm guessing your tweeter doesn't use ferrofluid. Generally ones that do don't have as severe of an impedance peak at Fs.
                    The grey plot is the Z-curve for your tweeter. Runs around 3ohms w/a 20ohm peak at Fs. That's SEVEN times higher. NO "textbook" XO filter can work w/that (in a way you'd expect it to).
                    The red plots show your 15n(ohm) series resistor on the raw tweeter. You're attenuating it about -12dB. You haven't changed the Z-profile at all, just added 15n to it across the board.
                    The purple plot is the driver w/an L-pad on it. 5n SeriesResistance and 2n PR (parallel).
                    This ALSO pulls the tweeter down by about -12dB, but it also doesn't show the SPL rise near Fs that your simple resistance does.
                    AND, look at the resultant Z-plot in purple. It's nearly a flat line (between 6-7n). You can basically throw a textbook filter on that and it will do what you want it to do.
                    Resistors cost about $1. Why not use 2 instead of 1 and get the results you want?

                    Click image for larger version  Name:	L-pad.jpg Views:	0 Size:	94.7 KB ID:	1428475
                    Last edited by Chris Roemer; 01-06-2020, 10:45 AM. Reason: complaint

                    Comment


                    • #11
                      Thanks for the additional input Chris. As I mentioned above in post 6, I DID try the suggestions you and others proposed. I must have hooked up the L pad incorrectly because it made the FR plot wonky and I thought that that was due to its particular interaction with the other components so I gave up on it.
                      I have a lot of difficulty with translating crossover schematics to actual crossovers, and as you may remember, had quite a bit of tepridation when you encouraged me to build my first 2 way several years ago. I have made some progress, but this is the first crossover I have attempted that did not start with a schematic from a proven design.
                      For me to add an L pad to what I had, I removed the 15 ohm resistor at the beginning of the tweeter circuit and put the suggested 4 ohm R in series between the positive terminal of the tweeter and the .40 mH inductor. I then connected the end of that resistor closest to the tweeter to one end of a 2 ohm resistor and the other end of the 2 ohm to ground. The resulting frequency response looked much more uneven than what I already had so I went on to try other things. Did I make some kind of mistake in the actual wiring of the L pad or was the poor frequency response result because of some other error I made when switching the alligator clips around on the other components?
                      When I'm tinkering with the xover I only change one thing at a time, but sometimes I have had odd results as a result of bad connections. I was aware that an L pad is generally the way to go on tweeter attenuation ( I have read SB101 and 2 other books on design} but used the in line resistor approach because it is what JR did in his tweeter circuit for the Tashas and because it is intuitively understandable to me which an L pad is not yet.
                      So, thanks again. I will try the L pad crossover again with the knowledge that if it doesnt provide good results I am somehow hooking it up incorrectly. I also am going to try the changes Ben suggested and will post results in the next couple of days.

                      Best,
                      Jay

                      Comment


                      • #12
                        I am not a proponent of using a 2 ohm resistor in parallel with a tweeter. This means the resistor will be shunting more power than the tweeter gets. My rule of thumb is to use a resistor that is half to full magnitude of the tweeter's Zmax, if suppressing the Fs is of a large concern and an LCR is not economical. So, Zmax = 25 ohms, 12.5-20 ohms resistor value.

                        Later,
                        Wolf
                        "Wolf, you shall now be known as "King of the Zip ties." -Pete00t
                        "Wolf and speakers equivalent to Picasso and 'Blue'" -dantheman
                        "He is a true ambassador for this forum and speaker DIY in general." -Ed Froste
                        "We're all in this together, so keep your stick on the ice!" - Red Green aka Steve Smith

                        *InDIYana event website*

                        Photobucket pages:
                        http://photobucket.com/Wolf-Speakers_and_more

                        My blog/writeups/thoughts here at PE:
                        http://techtalk.parts-express.com/blog.php?u=4102

                        Comment


                        • #13
                          Drjay, besides the advice you received here, it might be worth your time to look at this project for some crossover ideas:

                          https://www.parts-express.com/projec...copper-project

                          Comment


                          • #14
                            Hi Gents, after several more hours of tinkering and trying to utilize the suggestions offered, I have come up with a crossover that is pretty similar looking and sounding to the one posted above but which does use an Lpad and tries to cross the tweeter higher. I did try a 6.8 cap on the tweeter, but it put an even deeper dip in the 3 to 5K region than what is shown here. I did get a good null when polarity was not reversed and it looks like the crossover occurs at around 3K now. I even tried a third order on the tweeter but I couldn't get a good looking FR with it.
                            The speaker sounds really good overall, but I'm sure it has a way to go to get the most out of the drivers. One interesting note, the speaker has audible chiffing on very low frequency acoustic music even with a 1.375" port. It puts out a lot of clean low bass but needs an even bigger port .
                            Is this about as good as I can realistically get it with my current skill set or are there other things that are worth trying before I physically wire two of them up and put them in the cabinets?
                            Thanks!
                            ​​​​​​​Jay

                            Comment


                            • #15
                              I would measure each driver separately, and look at the roll offs. You have a 3rd order filter on the woofer, and a 2nd on the tweeter. That's unusual. What, if anything would prevent you from measuring with HOLM?

                              Comment

                              Working...
                              X