Announcement

Collapse
No announcement yet.

XMAX and THD question

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

  • #16
    Originally posted by fpitas View Post

    Oh, horns. Does anybody even use those anymore?
    Only those of us who like high sensitivity and inaudible THD.
    www.billfitzmaurice.com
    www.billfitzmaurice.info/forum

    Comment


    • #17
      Originally posted by billfitzmaurice View Post
      Only those of us who like high sensitivity and inaudible THD.
      Do horns and/or excursion-control also help against intermodulation distortion?

      Intermod dist and doppler distortion aren't the same, nor closely related are they? I thought IMD was somewhat like harmonic distortion except that it wasn't a fixed harmonic, but more of a moving target (not sure if the cause is similar or quite different between IMD and HD though).
      While doppler distortion is just the larger cone movement adding a doppler effect to the shorter wavelengths...I read that doppler dist isn't normally worth worrying about (too suble) both in an earlier thread and in a distortion topic I was reading a couple days ago.
      My first 2way build

      Comment


      • #18
        I haven't seen any comparisons of IM in horns versus direct radiators. Doppler tends to be a problem with longer excursions, which is exactly what modern subs use to overcome their low sensitivity.
        www.billfitzmaurice.com
        www.billfitzmaurice.info/forum

        Comment


        • #19
          Klippel has shown us the 3 major sources of distortion in woofers. BL (magnet curve), Cms (suspension curve), and Lvc (VC inductance curve). All are nonlinear with excursion, all cause essentially a "signal compression" type of distortion. At higher drive levels, (asking for higher excursion regardless of frequency), the woofer does not provide what was asked for. OTOH, this is also what keeps the cone under control, ie not trying to jump out of the gap.

          As the frequency is halved, the excursion quadruples, and so the compression distortion rises quickly. This is why you see distortion rising with falling frequency, and with an upward shape to the curve.

          The secret to a low distortion design is to start with a low distortion woofer (ie good Klippel tests), and then don't take it down too low, or drive it too hard. So we're back to Hoffman's law. High output, low extension, and low distortion requires a big woofer in a big box. It simply doesn't work as hard (aka travel as far). All of us on this board know this.

          There are offsets to the BL, Cm, and Rvc curves, too, and those are important, and quickly shown in the Klippel curves. Offsets lead to distorted sine waves.

          A good example of getting good performance is to choose a woofer large enough to achieve your goals, and then match it up with dual passives. As the frequency goes down, the woofer excursion reaches minimum at tune frequency, and essentially hands over to the passives, where their excursion is max. But they don't have motors, and so the BL and Rvc distortions are all gone, leaving only the Cms distortion. And the cone area is also twice, meaning less excursion required. In my speaker like this, the woofer can be pushed pretty hard, and the passives don't seem to throw as far. Of course, Fb is 18Hz, and I have no music that reaches down that far, so then it is up to video tracks.

          Another tip is to put the passives on opposite sides, use spikes, actively power the woofer, and have enough weight in the speaker. This all contributes to holding the basket still, which also lowers compression distortion!

          Finally, don't be tempted to "see how low I can cross this driver". Instead, consider how high you can cross the driver. Lower distortion, higher power, louder output. Hooray!
          --

          Philip Bamberg
          BambergAudio.com

          Comment


          • Steve Lee
            Steve Lee commented
            Editing a comment
            YES!

            Thanks, man.



            Big speakers . . .Mmmmm . . . .

          • PEB
            PEB commented
            Editing a comment
            I find it an enjoyable challenge to try and get big sound from a compact speaker. By big, I mean wide bandwidth and plays loud but composed.

        • #20
          Even though I'll probably keep asking slightly different versions of these same questions over and over, it's mostly because reading the answers given in slightly different wording tends to wiggle those answers into my brain in different ways (hopefully resulting in one or more of the answers "sticking"). AKA, don't feel discouraged from chiming in and please don't feel like I'm ignoring your answers when I end up basically asking the same question in a different way...I'm definitely reading (and slowly absorbing) the information.
          Thanks to all of you folks taking the time to try and educate me and anyone else who might be reading from a similar place of ignorance as I am.


          Found a neat thing I didn't really notice before on a fun site I've messed with earlier. https://www.audiocheck.net/testtones_thd.php
          For anyone (understandably) unwilling to click a link from an online stranger, this page has different buttons which play a 1Khz sine wave with different levels of included harmonic distortion ranging from 0%-5% (.001/.005/.01/.05/.1/.5/1/5%). It mentions an obvious but helpful warning that;
          -"Beware, your equipment's own THD will corrupt this listening test! You will not be able to discriminate THD figures lower than what your own audio equipment gives you access to."
          And that got be wondering, how accurate would it be to use this as a rough by-ear test for driver THD at a given frequency (there's are multiple frequency options below 1khz)?
          For example, on my current system I can't hear any difference between the 0% all the way up to the 0.1%, but at 0.5% I can start to hear the increased distortion a little and by 1% the distortion becomes obvious to me.

          Is this a fair indicator that my overall system THD (at this frequency and level, at least) is in the ballpark of 0.1%-0.5%,
          OR
          is it more likely that regardless of my system's distortion, the 0.5% and higher THD artificially added to the signal should become obvious even on top of a system with similarly high distortion of its own?

          Using a simple sine-wave test like this and a hypothetical system with 1% THD, is it easy for most listeners to hear an increase to 1.1%THD or 1.5%THD or 5%? At what point does the system's own distortion tend to mask the artificially added distortion, and can this be considered at all useful or is it just a neat toy to test your own ears.
          My first 2way build

          Comment


          • #21
            Originally posted by LOUT View Post
            Do horns and/or excursion-control also help against intermodulation distortion?

            Intermod dist and doppler distortion aren't the same, nor closely related are they? I thought IMD was somewhat like harmonic distortion except that it wasn't a fixed harmonic, but more of a moving target (not sure if the cause is similar or quite different between IMD and HD though).
            While doppler distortion is just the larger cone movement adding a doppler effect to the shorter wavelengths...I read that doppler dist isn't normally worth worrying about (too suble) both in an earlier thread and in a distortion topic I was reading a couple days ago.
            Doppler distortion is by its nature a form of IMD, one frequency is modulating another. "Adding a Doppler effect to shorter wavelengths" means creating new frequencies as sidebands.

            Whether it's worth worrying about depends on a lot of things. A lot of tests indicate that many people are pretty insensitive to distortion in general. YMMV and all that.
            Last edited by fpitas; 04-13-2021, 10:37 AM.
            Francis

            Comment


            • #22
              Originally posted by LOUT View Post
              Even though I'll probably keep asking slightly different versions of these same questions over and over, it's mostly because reading the answers given in slightly different wording tends to wiggle those answers into my brain in different ways (hopefully resulting in one or more of the answers "sticking"). AKA, don't feel discouraged from chiming in and please don't feel like I'm ignoring your answers when I end up basically asking the same question in a different way...I'm definitely reading (and slowly absorbing) the information.
              Thanks to all of you folks taking the time to try and educate me and anyone else who might be reading from a similar place of ignorance as I am.


              Found a neat thing I didn't really notice before on a fun site I've messed with earlier. https://www.audiocheck.net/testtones_thd.php
              For anyone (understandably) unwilling to click a link from an online stranger, this page has different buttons which play a 1Khz sine wave with different levels of included harmonic distortion ranging from 0%-5% (.001/.005/.01/.05/.1/.5/1/5%). It mentions an obvious but helpful warning that;
              -"Beware, your equipment's own THD will corrupt this listening test! You will not be able to discriminate THD figures lower than what your own audio equipment gives you access to."
              And that got be wondering, how accurate would it be to use this as a rough by-ear test for driver THD at a given frequency (there's are multiple frequency options below 1khz)?
              For example, on my current system I can't hear any difference between the 0% all the way up to the 0.1%, but at 0.5% I can start to hear the increased distortion a little and by 1% the distortion becomes obvious to me.

              Is this a fair indicator that my overall system THD (at this frequency and level, at least) is in the ballpark of 0.1%-0.5%,
              OR
              is it more likely that regardless of my system's distortion, the 0.5% and higher THD artificially added to the signal should become obvious even on top of a system with similarly high distortion of its own?

              Using a simple sine-wave test like this and a hypothetical system with 1% THD, is it easy for most listeners to hear an increase to 1.1%THD or 1.5%THD or 5%? At what point does the system's own distortion tend to mask the artificially added distortion, and can this be considered at all useful or is it just a neat toy to test your own ears.
              I like how others build speakers and have build some myself(amateur stuff and some with help of members on this forum) but i always have those OCD? questions like this one of yours. If you look on hificompas you can search speakers and look at the frequention/distortion charts with different inputs(every next input is double the power. So when i look at max power and the distortion in the bass region, its mostly about 10% ( the few speakers i looked with low fs). If you half the power, distortion seems to half to and the further you go the less it will go down(distortion). So with my amateur eye and toughts i would say if you half the max power, distortion will half to! I read on the internet that the human ear is the most sensitive at 1000hz and less sensitive at bass. So bass distortion of 10% can sound 'normal' but 1% distortion at 1000hz can sound to much, just like you mentioned!

              Comment

              Working...
              X