Announcement

Collapse
No announcement yet.

Question regarding crossovers

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

  • #16
    I might be misunderstanding something, but it sounds like you liked the fullrange at first, then kind of turned a full 180degrees away.
    Originally posted by Stash View Post
    I installed the full-ranges in place of the 1 inch soft domes, using the mounting pods of the tweeters (see pics). The results aren't bad at all. The full-ranges seem to mesh much better with the 6.5 woofers. I made some small adjustments via my EQ. I will listen to them for several days and decide if I need to try the dome tweeters again.
    Originally posted by Stash View Post
    Do not ever use a mini full-range as a tweeter. Wow! Even more terrible than what I had previously lol. What a horrible mistake.
    What happened?
    My first 2way build

    Comment


    • #17
      With a 2nd or 3rd order XO, is there a reliable way to get decent results AND avoid dropping impedance too low when the driver's impedance and frequency slopes are unknown or loosely known? I'm assuming these aren't being measured in the most traditional of ways (might be assuming wrong?).
      My first 2way build

      Comment


      • #18
        I spent a few hours late this evening messing with my car audio. I have come to some conclusions.

        I re-installed the 1 inch silk dome tweeters. I installed them to aim upwards, towards the ceiling. The ceiling is cushioned, so it may absorb or at least cause a little less high freq reflections. I am also trying to lessen beaming.

        I kept the 2.2uf cap crossover. For now, it works.

        I set my EQ at flat across the spectrum, except the 16kHz band. I boosted 16kHz the full +9dB. I have read articles claiming our brains focus on the peaks of sound. So, I am trying to 'mask' the harshness by boosting the extreme treble.
        it seems to work to some extent. And I think the tweeters need that boost, anyway

        I now realize some of the harshness, or whatever, is actually coming from the 6.5 inch woofers mounted low, in the door panels.
        As I mentioned, above, by boosting the 16kHz band, my tweeters now play louder than the woofers. The tweeters seem to be masking some of the mentioned harshness. I'll continue to listen as-is for a couple weeks. See how I feel about it

        I think I've been overdriving the built in 22w/ch amp in my HU.
        I turned down the radio section gain by adjusting the volume offset to -3dB, which seems to help a bit.
        I also had BT and DISC sections offset at +6dB. I reset both of those sources to 0dB. Streaming music now sounds somewhat smoother as well.

        I think part of the problem is the built-in HU amp. I looked up the data sheet for its amp IC. Distortion at 20kHz is pretty high (0.2%, regardless of output, soaring much higher after 17 or 18 watts/ch).

        My dome tweeters claim a response up to 23kHz. I already suspected the quoted rating is jibberish. I have the same pair of tweeters in my home office AV system. When measured, on-axis, the tweeter's response falls after about 10kHz.
        By comparison, when I measure the on-axis response of my horn tweeters in my MACH ONES, the tweeters output a basically flat response up to and most likely beyond 20kHz. I can hear the difference in the detail and nuance.

        Conslcusion: I should probably buy a discreet amp, some day. Maybe around the next Black Friday. My HU has line outs. The line outs are supposed to exhibit much lower distortion, measuring around 0.01% to 0.06%, which would provide much cleaner audio.

        Along with an amp, I'd look to upgrade to a better pair of dome tweeters.

        building better crossovers will also wait. I just don't feel like removing all the door panels again and rewiring the drivers. It's too hot outside lol.

        Right now, what I have, stays as-is.
        attached is my new mounting of my tweeters.

        Comment


        • #19
          Originally posted by LOUT View Post
          With a 2nd or 3rd order XO, is there a reliable way to get decent results AND avoid dropping impedance too low when the driver's impedance and frequency slopes are unknown or loosely known? I'm assuming these aren't being measured in the most traditional of ways (might be assuming wrong?).
          Here is an example third order Butterworth high pass @2.5K on a real measured SBA29 tweeter. The impact on impedance is pretty low - however, NOTE that this assumes no ESR on the capacitors. When ESR is factored in the effect on impedance is zero to positive.

          Original measurements in red - crossover effect in blue.

          3.6 to 3.2 ohms net effect. However, any inclusion of an L-Pad or factoring ESR for the capacitors essentially increases the impedance above the original measurement.

          Click image for larger version

Name:	XO-schema_3rd.png
Views:	144
Size:	6.9 KB
ID:	1471921
          Click image for larger version

Name:	XO-schema_3rd_fr.png
Views:	106
Size:	35.6 KB
ID:	1471922 Click image for larger version

Name:	XO-schema_3rd_imp.png
Views:	107
Size:	21.1 KB
ID:	1471923

          Comment


          • #20
            Originally posted by Stash View Post
            ....I kept the 2.2uf cap crossover. For now, it works.
            Mate I have no idea what you are up to here...what sources you are educating yourself with or what theory it is that's driving your troubleshooting process....

            This is the effect of a single 2.2uf cap on a measured 4ohm SBA29 textile dome tweeter....
            Click image for larger version

Name:	2.2CapOnSba.png
Views:	124
Size:	35.8 KB
ID:	1471925

            This is the "crossover" transfer function of that single 2.2uf cap. And it pretty much looks like this even if the tweeter was a perfectly flat 4ohm its whole range...
            Click image for larger version

Name:	2.2Transfer.png
Views:	105
Size:	13.3 KB
ID:	1471926

            It is the single most useless crossover I've ever seen and essentially takes a 4ohm tweeter out of most of its operating range.


            Originally posted by Stash View Post
            ....I now realize some of the harshness, or whatever, is actually coming from the 6.5 inch woofers mounted low, in the door panels.
            Of course it is. Look at the above frequency response....what do you have filling essentially the entire gap in response somewhere between 1K and 15K? I doubt the 6.5inch has the ability to meet a tweeter crossed that high without significant breakup, beaming and distortion (translate: outright hideous screeching)...and if its sitting way down in your door you have to consider that there is also an enormous sonic hole in the response making it to your ears where you are subjected to essentially just bass/mid-bass and the very top end of the treble....this will make it SOUND like its harsh because there is nothing filling the gap and rounding out the response.

            Literally, if I dropped the entire midrange from my system and just turned it up with the upper treble and bass playing it would sound, subjectively, harsh! As soon as I returned the midrange, it would sound perfectly balanced.


            Originally posted by Stash View Post
            ....My dome tweeters claim a response up to 23kHz. I already suspected the quoted rating is jibberish. I have the same pair of tweeters in my home office AV system. When measured, on-axis, the tweeter's response falls after about 10kHz
            Newborns and dogs might be able to hear that, but generally when we find treble harsh/sharp etc we are referring to somewhere around 5-8K range.

            If you have the capacity to take measurements, drag that stuff out to your car and take some measurements. I bet there is a giant hole in the response! It will significantly improve if you apply a proper crossover to that tweeter, make it play lower, and pad its volume down to align with the fact it needs to meet up with the output of a woofer down in the door for your listening position. You measurements will tell you exactly what that relative difference is at your listening position.

            Worst case scenario your response probably looks like this with a low sensitivity woofer (made worse by position in door) potentially out of phase with a tweeter of much higher sensitivity with a terrible capacitor on it. If its this bad, this system would pretty much sound like sitting next to a water jet cutter....
            Click image for larger version

Name:	OutOfPhaseMismatch.png
Views:	105
Size:	30.4 KB
ID:	1471927


            Take some measurements and stop guessing.


            Comment


            • #21
              Haha..I like all the graphs. Yes, there is a midrange hole. And yes, with the tweeter crossed over so high, I can hear the woofer distorting/breaking-up at higher frequencies. I've taken some overall measurements in the car. That became apparent. However, lowering the tweeter with a simple larger cap didn't work. Tweeter becomes distorted, as you guys probably guessed. But, that's all I have on-hand at the moment.

              I did not want to use such a small 2.2uf cap. I only did so to reduce distortion from the tweeters, to make them listenable.
              With this cap, and only the 16kHz band boosted by +9dB,
              I have some balance in the treble end, while reducing distortion. At least for now.

              I already began making a parts list for 12 and 18dB crossovers, for the tweeters. I totally agree with all of you, that I need a real crossover network. I had another pair of 12dB tweeter crossovers with an attenuation switch. Regretfully, I took them apart. My mistake.

              One point I'd like to make. I again messed with my system this morning. The other definite problem is the amp. The HU internal amp, like most amps, plays fine and very clean at half power. However, like most amps, pushed beyond about 65%-70%, noticeable distortion becomes obvious, especially in high frequencies.
              So, here is the problem with my HU. It has a 0-35 increment digital Volume Control. 1/2 power is about 17-18 setting. Distortion is apparent around 26-27 setting. That only leaves a few clicks between moderately quiet to fairly loud and distorted. It sometimes seems as if just one click changes the sound from decent to distorted. Frustrating.
              I would like to see HU's with wider range volume controls, like my home receiver. It features 0-97, and exhibits distortion around 67-70 setting. But, my receiver features 1/2 dB steps. So there is plenty of adjustment options between the half-way point and top usesble end. It's too bad HU's aren't designed with volume ranges of at least 50-60 increments. That's how I see it anyway.

              I will take up your suggestions. Build tweeter crossovers, first. See how the car sounds afterwards. Then I'll post back here. I'll probably order parts from PE this afternoon.

              If I decide on an 18dB crossover, how do I wire it? In phase or out of phase? Or do I experiment to see which sounds better?

              If I do 12dB, I'll wire in phase.

              Comment


              • #22
                You guys had me laughing with your comments.
                I went out to my car. Ripped out the 2.2uf caps. Replaced them with the 6.8uf.
                Tweeters def playing much more treble. Had to severely tone down 4 & 6kHz with EQ. Still boosted 16kHz.
                Also used tweeter crossover on amp to tone down -3dB from 10kHz on up.

                Finally, I reduced Volume Offset to -3dB for radio, BT and DISC. Gives me a little more range of adjustment before hitting distortion levels.

                The original crossovers which came with these dome tweeters used the 6.8uf caps and an air-coil inductor (guessing around 0.23mH) for a claimed 12dB crossover at 3.5kHz.
                I installed the two above remaining crossovers in my home office speakers, using the same dome tweeters. When I took rough response measurements, I noticed the crossover point to be about 4kHz, which is pretty consistent to the claimed point.

                Now....I still have at least 1 of the coils. If I can find another coil, roughly 0.23mH or thereabout, I could reconstruct the exact tweeter 12dB crossover for the same effect, in my car. I also still have the damping power resistors as well, to reduce tweeter output by -2dB.

                I will order either one coil, if I can find it with free ship, or, a pair from PE, of about the same value. Then I can rebuild the 2nd order crossover at roughly 4kHz.
                If I choose to build woofer crossovers, I can do so later.
                Does this.sound like a plan?
                suggestions? Comments? Sarcasm?



                Comment


                • #23
                  Originally posted by Stash View Post
                  I went out to my car. Ripped out the 2.2uf caps. Replaced them with the 6.8uf.
                  Tweeters def playing much more treble. Had to severely tone down 4 & 6kHz with EQ. Still boosted 16kHz.
                  Also used tweeter crossover on amp to tone down -3dB from 10kHz on up.
                  The 2.2 uF caps were effectively doing what you did with EQ, I would have left them in. One of the benefits of a first order filter is that it is easy to make adjustments while listening to see what sounds best.

                  I use first order crossovers whenever possible, and I would rarely use anything over 2.2uF for a 4 ohm tweeter. My car speakers do have a 4.7 uF on the tweeter and nothing on the woofer. They sound great to me, but not all drivers will work well with this arrangement. A woofer with a clean roll off at the high end is required.

                  When going from a second order to first order, it is usually necessary to reduce the capacitor by half or more. The impedance peak on the tweeters low end can cause power handling issues, the smaller cap helps with this. The higher impedance of the low value cap reduces the need for a series resistor in many cases. If the tweeter is still too loud, I prefer an Lpad resistor arrangement to add some damping while keeping the impedance fairly constant.

                  For tweeters in a new design, I usually start with a 1.0 to 1.5 uF cap. First order systems usually rely on a large overlap between the drivers to bring the mids up. The rising response of the tweeter also helps with power response. Normally in a power response plot, the high end is falling. Maybe this is why you are boosting at 16kHz?

                  Comment

                  Working...
                  X