Announcement

Collapse
No announcement yet.

Adjust woofer/mid balance in XO \n 3 way

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

  • Adjust woofer/mid balance in XO \n 3 way

    How do you guys approach woofer/mid tonal balance XO adjustments in a 3 way (10" - 6.5" - 1")? Given 400hz and 2khz XO points. It's a familiar (tarkus) design here, parts are all soldered correctly. Woofer tuning verified, impedance sweep with DATS looks good too.

    Voices lack warmth and body, especially male vocals. Quick measurements at 1m, gated to 4-5 ms confirm bass and lower mids (100 - 300 hz) region is slightly depressed by 3-4 dB relative to the upper mids and tweeter

    I'm looking to do more accurate measurements using Bagby's quasi anechoic down to 10hz method soon, to understand the 20Hz - 20 kHZ more accurately.

    Should i be reducing the roll off on the 10" woofer, or adjusting the high pass on the mids? When i played the mids without XO, they actually sound pretty good. But in box + XO, they are really soft in output relative to the tweeter. Messed around with the sealed stuffing as well, now i'm at about 3 ounces acoustistuff with 1" fibreglass lining only at the back.

    Also, in the pictures i attached, did i map out the right circuit values to the PCD schematic? Am confused about C2 (65 uF) cap in the circuit. Was trying to simulate and see the effects

  • #2
    The 65 uF (C2) in Paul's design would be C16 in PCD, not C15. Paul uses a first order electrical high pass filter (the single 65 uF cap) and second order electrical low pass filter.
    Craig

    The lowest possible F3 box alignment is not always the best alignment.

    Designing and building speaker projects are like playing with adult Lego Blocks for me.

    Comment


    • #3
      I think adjusting the high-pass on the mids should be a little safer for avoiding dropping the impedance too low compared to reducing the roll-off on the 10" woofer, but I'm not sure how the impedance looks as-is so that's just a guess. The frequency response looks like it should sound pretty nice as far as I can tell though I don't know much about the differences or possible inaccuracies between measuring and listening and sims...I don't see the 3-4db depression between 100-300 that you mention in the images. Is that left-side image the software sim, are you able to post the measured response you mentioned (or did you already and I'm just not seeing it)?

      Did your mid woofer's sealed area end up with different internal volume compared to the build recommendation, or could the 3-4db dip be a room response characteristic (which I think is common when measuring under 500hz from distances more than a couple inches). ?

      Comment


      • #4
        Check the wiring for polarity errors.

        Switch polarity on the mid, and remeasure. It should have nulls at the x-over points. Don't forget to switch it back.

        Comment


        • #5
          Originally posted by rpb View Post
          Check the wiring for polarity errors.

          Switch polarity on the mid, and remeasure. It should have nulls at the x-over points. Don't forget to switch it back.
          Yeah. Before you start redesigning, it's worth a few minutes to see if it's a wiring error etc.
          Francis

          Comment


          • #6
            Originally posted by LOUT View Post
            I think adjusting the high-pass on the mids should be a little safer for avoiding dropping the impedance too low compared to reducing the roll-off on the 10" woofer, but I'm not sure how the impedance looks as-is so that's just a guess. The frequency response looks like it should sound pretty nice as far as I can tell though I don't know much about the differences or possible inaccuracies between measuring and listening and sims...I don't see the 3-4db depression between 100-300 that you mention in the images. Is that left-side image the software sim, are you able to post the measured response you mentioned (or did you already and I'm just not seeing it)?

            Did your mid woofer's sealed area end up with different internal volume compared to the build recommendation, or could the 3-4db dip be a room response characteristic (which I think is common when measuring under 500hz from distances more than a couple inches). ?
            I haven't posted up the measured response yet. I'm not sure if that is the software sim or actual measurements of the original design. Yea i'm wondering if it's a room response characteristic as well.

            Sealed area is spot on 13 liters as called for in design

            Thanks guys. I've verified and tested the tweeter and mid polarities already with a battery. They are indeed wired out of phase, according to the original design.
            I've also tested circuit continuity at every joint of the crossover.. so it's not that either

            Comment


            • #7
              Originally posted by kristriper View Post
              I haven't posted up the measured response yet. I'm not sure if that is the software sim or actual measurements of the original design. Yea i'm wondering if it's a room response characteristic as well.

              Sealed area is spot on 13 liters as called for in design

              Thanks guys. I've verified and tested the tweeter and mid polarities already with a battery. They are indeed wired out of phase, according to the original design.
              I've also tested circuit continuity at every joint of the crossover.. so it's not that either
              Measure the response with a mid reversed. It would be interesting to see some measurements if you are set up to take accurate ones.

              Comment


              • #8
                It LOOKs like you could lift the bottom end of the mid by about +2dB (provided everything else checks out - which it seems like you have) by droppin the 4.7n series resistor to 3ohms (OR parallel the 4.7n w/an 8n - for fun). To double the effect, drop it down to 1n (OR parallel the 4.7 w/a 1.5 - but if you go this route, you might need more power handling on that resistor, like maybe 20w?).

                Comment


                • #9
                  Here's the values entered into PCD. The resistors that are .3 ohms are estimates for the coil dcr. The 12 ohm resistor may be the wrong value. I couldn't read it for the line drawn through it. The woofer files are one of my HDS drivers.

                  Click image for larger version

Name:	PCD FILTER.jpg
Views:	143
Size:	255.0 KB
ID:	1435219

                  Comment


                  • #10
                    Originally posted by Chris Roemer View Post
                    It LOOKs like you could lift the bottom end of the mid by about +2dB (provided everything else checks out - which it seems like you have) by droppin the 4.7n series resistor to 3ohms (OR parallel the 4.7n w/an 8n - for fun). To double the effect, drop it down to 1n (OR parallel the 4.7 w/a 1.5 - but if you go this route, you might need more power handling on that resistor, like maybe 20w?).
                    Thanks so much for the detailed help, Chris!

                    Originally posted by rpb View Post
                    Here's the values entered into PCD. The resistors that are .3 ohms are estimates for the coil dcr. The 12 ohm resistor may be the wrong value. I couldn't read it for the line drawn through it. The woofer files are one of my HDS drivers.

                    Click image for larger version  Name:	PCD FILTER.jpg Views:	0 Size:	255.0 KB ID:	1435219
                    Awesome, thank you rpb! 12 ohm is the right value

                    Are there any other phase/XO issues i should be aware of when making these changes?

                    Comment


                    • #11
                      If you make the high-pass 2nd order, the lower end of the mid response will come up quite a bit. It might not add well with the woofer though. Try adding a 2mH coil to see what I'm talking about.

                      Comment


                      • #12
                        Changes in capacitance and inductance will rotate phase, simple resistance changes (basically) will not.

                        Comment


                        • #13
                          Just a thought.....floor bounce null?
                          https://mehlau.net/audio/floorbounce/

                          I estimate at a 3m listening distance using the above on tweeter axis - the floor bounce (null) occurs at 196Hz due to the woofer height from the floor.

                          Now, here's a quick and dirty test.... flip the woofer box upside down. This would move the floor bounce cancellation up in frequency to approx 600Hz - not much above your XO point to the mid, but not in the middle of the midbass.

                          Of course this will floor load the woofer and you'll get a boost at the low end, which may or may not be to your liking. At least you can re-measure to see if floor bounce is having an effect.

                          Comment


                          • #14
                            Here's where active crossovers shine, especially if the mids and bass drivers have a wide response overlap like 100-1000hz. A person can choose a crossover point anywhere in that range and measure smooth flat response but get much different sound as the point is varied. When setting up my system using the AE TD15 and 18Sound 6ND430 I swept the crossover point while changing between male and female vocal recordings to find the sweet spot at just over 400hz. It took an afternoon but not a single calculation or measurement was used. Biamping is well worth the cost and extra equipment. In the past I spent months tweaking a crossover with many measurements.

                            Comment

                            Working...
                            X