Announcement

Collapse
No announcement yet.

Building Around a Hivi Q1R Tweeter

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

  • #31
    Attach the .zma and .frd files used please.
    http://techtalk.parts-express.com/fo...khanspires-but
    http://techtalk.parts-express.com/fo...pico-neo-build
    http://techtalk.parts-express.com/fo...ensation-build

    Comment


    • #32
      I would not think that you'd need a lot of BSC added to what you have there. The 5dB rise into the bass will be plenty. You do need to kill the peak at 3.2k, or that dip will persist in the tweeter response. You are currently -5dB at 3kHz for the tweeter, so I'd call that a 3kHz xover currently. LR xovers sum at -6dB point. You also need to include driver phases, or you cannot tell if the alignment works. Impedance phase shows whether or not it will be a happy or unhappy load for the amplifier.

      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


      • #33
        Alright I went into it again to show the phase and ended up messing around again. I put a notch on the 3.2k peak and tried to tweak the L pad on the tweeter to match the woofer, but now my crossover point is a bit high for any feasible center to center mounting distance for the drivers (even for cutting the tweeter faceplate, which I'm not sold on yet!). I'm still trying to comprehend everything to be mindful of on the phase, all I've gleaned so far is that you want the phase of both drivers to approach the valley and then peak of crossover at close to the same point, which doesn't look too close in my case. This latest try is fairly flat for what I've been having so far, not sure how flat I need to push for. The perfectionist in me is fighting the frugality of not wanting to load the crossover full of parts. The reverse null also looks a bit weird, so that tells me I still have a ways to go. I attached the .frd and .zma files used.

        Click image for larger version

Name:	System Null.PNG
Views:	102
Size:	361.9 KB
ID:	1428435 Click image for larger version

Name:	System Response.PNG
Views:	99
Size:	360.8 KB
ID:	1428436 Click image for larger version

Name:	Woofer Response.PNG
Views:	96
Size:	308.4 KB
ID:	1428437 Click image for larger version

Name:	Tweeter Response.PNG
Views:	96
Size:	325.5 KB
ID:	1428438

        Comment


        • #34
          Limited to 5 attachments per post! Here are the .frd and .zma files.
          Hi-Vi Tweeter BlendedMiniPhase V2.frd Swans 2-2 Tweeter.zma Woofer box min phase.zma Woofer box min phase.frd

          Comment


          • #35
            Looking much better! I still want to see the impedance phase on the system impedance. As long as your CTC is under 6", you are likely okay.

            What I'd try;
            - increase cap or coil on woofer low pass to see if the knee lowers further. That bump is not huge but might sound forward. If that doesn't help, try reducing 3 ohm resistor.
            - use a laminate core on the woofer to reduce dcr level and not pad tweeter so much.
            - if impedance phase dips hard, move series resistor out front of high pass.
            - increase parallel tweeter resistor to about 15-20 ohms. Having it at 1.9 ohms will drain a lot of current and you can increase series resistor to compensate.

            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


            • #36
              If this is really your first crack at this, you are doing much better than I was when I first started! You have a great coach in Wolf. He has taught me a lot - giving me pointers many years ago. His blogs were/are quite helpful, as was following the many threads he was involved with. That said, I still have much to learn. My only advice would be to set your acoustic targets and see how close you can get to approaching the basic shape of the slopes.
              Co-conspirator in the development of the "CR Gnarly Fidelity Reduction Unit" - Registered Trademark, Patent Pending.

              Comment


              • #37
                Maybe something like this.

                Click image for larger version

Name:	partial filter.jpg
Views:	97
Size:	250.6 KB
ID:	1428454

                Comment


                • #38
                  Alright I tried to implement Wolf's suggestions and I came up with some mixed results. Increasing the cap helped push the 3.2k problem down further. I also changed the coil out for a P core with a DCR of 0.066 and reduced the inline resistance on the cap. The reverse null looks better, which as I understand it is a bit of an indicator towards better imaging. I moved the series resistor in front of the tweeter crossover to help with the impedance, but the impedance still dips a bit low for my liking. I raised the parallel resistor on the tweeter to 15 ohms but I couldn't get it flat again so now it's rising as it goes towards 20k. The woofer is looking better, but the tweeter could be better. The weird plateau before it rises again around 3k seems like it would cause some issues getting close to the woofer response. Tom_s this is indeed my first time around with designing, that's why I was skeptical of any initial success I had! For some reason I can't get the targets to overlay, I enter numbers and select a shape but nothing comes up, I think I'm just missing the button to turn them on or something. Rpb, I do like the shape of your example, I was just trying to keep the number of (expensive!) coils in the crossover. I'll definitely keep that in mind though, I'm not opposed to one more component, just trying to see if there's a way to accomplish something close with fewer components. But at the end of the day I won't turn down a good solution! I really appreciate all the suggestions, the shapes of these graphs have been better than I would've come up with on my own after this much time!

                  Click image for larger version

Name:	Reverse Null.PNG
Views:	89
Size:	331.1 KB
ID:	1428492 Click image for larger version

Name:	Woofer Response.PNG
Views:	83
Size:	298.8 KB
ID:	1428493 Click image for larger version

Name:	System Response.PNG
Views:	83
Size:	350.5 KB
ID:	1428494 Click image for larger version

Name:	Tweeter Response.PNG
Views:	79
Size:	325.3 KB
ID:	1428495

                  Comment


                  • #39
                    The target button is on the right side, down low, next to the 20k marker. You can turn it on and off as needed. Look at the second picture in the post before this one. (Post #38)

                    Comment


                    • #40
                      Not quite what Wolf suggested, but I see no issue with a smaller parallel cap. (Ignore the woofer shown, it's not yours.)

                      Click image for larger version  Name:	tweeter filter.jpg Views:	0 Size:	263.8 KB ID:	1428507

                      Comment


                      • #41
                        Alright one more try before bed! I added the other resistor on the tweeter rpd had in his last example, kept the parallel resistor at 20 like Wolf suggested, but I kind of tried to keep both drivers up higher rather than try to fight all of it down to 75. I was able to get rid of the notch filter and put the woofer to a 3rd order crossover, so two fewer components. I figured out the targets too! Not to say I followed them explicitly, however. My question about this current iteration is: there's the big dip between 80 and 700 Hz, if there's an easy way to describe it, what would that sound like? I don't think this is how I want to leave it, but I'm curious what something like this current graph would sound like.

                        Click image for larger version

Name:	System Response.PNG
Views:	93
Size:	356.1 KB
ID:	1428536 Click image for larger version

Name:	Reverse Null.PNG
Views:	92
Size:	360.7 KB
ID:	1428533 Click image for larger version

Name:	Woofer Response.PNG
Views:	89
Size:	307.7 KB
ID:	1428534 Click image for larger version

Name:	Tweeter Response.PNG
Views:	81
Size:	321.0 KB
ID:	1428535

                        Comment


                        • #42
                          I like to round of the part values' and see how picky the x-over is when a part is bigger, or smaller than the ideal value. Sometimes 20% bigger, or smaller makes almost no difference. Sometimes it does, and the filter might be difficult to actually get right.

                          Comment


                          • #43
                            I would keep the notch filter you had earlier in woofer section - maybe get even more aggressive with it while still watching your impedance. That will help pull 700-1K down and likely allow you to omit L3 and still have good phase tracking. I agree about normalizing values - make L2 a 1mH - a solid core 1mH is much cheaper than a .9mH air core with a similar DCR. Oh, and make sure you have correct DCR entered. Set both targets to the same level. Your tweeter target is 80dB - woofer 75dB. I would also try deleting one of the series resistors in the tweeter section - usually no need for both.
                            Co-conspirator in the development of the "CR Gnarly Fidelity Reduction Unit" - Registered Trademark, Patent Pending.

                            Comment


                            • #44
                              Almost forgot - what you currently have might sound fine in the right room. Every room is different. If you have lots of soft surfaces like heavy drapes & over-stuffed furniture, they might be be fine. My room is pretty bright, so they would likely sound too forward for me. It's a delicate balancing act - I'm still not completely satisfied with any of my creations. But I am getting closer with every build.
                              Co-conspirator in the development of the "CR Gnarly Fidelity Reduction Unit" - Registered Trademark, Patent Pending.

                              Comment


                              • #45
                                You can see that your peak (1k-2k) is coming from your woofer filter (it's right there in its xfr fn).
                                Not a good idea (unless, maybe if you're filling in a "depression" - which is not the case here, other wise your FR plot would be flatter).
                                (forgot which way it goes, but) Try making the series component larger / and the shunt proportionately smaller; OR the other way around. One way makes that shoulder "peakier" (higher "Q"), and other makes it shallower (lower "Q" - which is what you want).

                                Comment

                                Working...
                                X