Announcement

Collapse
No announcement yet.

Thinking about a Budget Dayton Project with 1stOrder XO

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

  • Thinking about a Budget Dayton Project with 1stOrder XO

    I wasn't sure if this would be appropriate in DrJay's thread discussing "An Amazingly Simple 2way with Some Good Reviews", because the original post sounded like it might be focused on a particular set of drivers. Figured this would be safer than stepping on any toes, and a fun little challenge.
    Mostly just a lot of rambling, thinking out loud. Feel free to point out any places to improve or things you think could be cool if done a different way. I noticed PE is uploading FRD/ZMA files for more brands, so that could open up some options I'm not used to seeing for playing with simulations.

    For now though, I started to narrow down a short list of Dayton drivers with relatively flat or decending frequency-response around the XO point (guessing XO will end up around 3000-5000hz) and naturally smooth enough FR overall to avoid the temptation of any additional sculpting. This immediately cut-out my beloved TD20 which seems to favor a steeper XO or a notch to tame its 1700hz peak, but the ND13/16 tweeters all appear to roll-off more smoothly with only a capacitor. For woofers, it looks like several of the RS-series and a few DC-series could be a decent fit using only an inductor. The DMA90 also looks like a really good fit FR-wise for a simple LP if the build doesn't need to get too loud or reach super low.

    I don't like the idea of spending a lot more for a woofer/tweeter that fits a single-part XO build if simply buying another cap+resistor+inductor and a much cheaper set of otherwise similar-quality drivers can do the same job at a lower overall cost. With a nicely designed 2ndOrder XO I think the TD20f and DS135 can probably make a really nice yet affordable speaker. With that in mind, I decided to set a rough budget aiming to land in a similar price range as the TD20f and DS135 plus the cost of the additional cap, coil and resistor those drivers would probably need to get a nicely flat response....giving around $5-10 of extra driver budget in exchange for the savings in skipping some of those cheap XO parts. Those ND13/16 tweeters are in a pretty comfortable spot, but many of the RS-series woofers are beyond my arbitrary imagined budget (though I think the rs150...rs150p? looks like a really nice option for a bit more money). The cheaper RS100/125 are probably too small to keep up with something like a DS135, but the 8ohm versions of the DC160 appear to keep up very well while only asking for 1-2liters more room to reach similar bass-depth.

    Before checking BSC, it kind of looked like the DC160-8 could potentially sneak by without any XO parts AT ALL. The lack of BSC also let the ND16-6 appear to work with only a tiny cap and no resistor to level-match:
    Click image for larger version  Name:	DC160-8 and ND16FA-6 no BSC.png Views:	4 Size:	26.9 KB ID:	1469941


    Unfortunately, the BaffleStep REALLY wants to boost the ~1500hz peak along with the expected bass fall-off unless the SIMed baffle was heavily altered (huge roundover or a lot of extra space added). So the SIM response including BaffleStep really benefited from a 3.5mH inductor on the woofer as well as an 8ohm resistor taming the tweeter...still only 3parts including the tweeter's 1.5uF capacitor.
    Click image for larger version  Name:	DC160-8 and ND16FA-6 graph.png Views:	4 Size:	26.6 KB ID:	1469942 Click image for larger version  Name:	DC160-8 and ND16FA-6 XO.png Views:	0 Size:	4.1 KB ID:	1469944
    The 0.39ohm value on the woofer is just the natural DCR of the inductor...not an actual R.
    These are just sims using the PE FRD/ZMA files, and there isn't any Zaxis offset although it does appear to be somewhat forgiving around +/-26mm of offset.


    I still feel like shooting for single-part HP and LP is mostly a fun, but dumb, but fun exercise. :P
    Change my mind.
    Attached Files
    Last edited by LOUT; 05-21-2021, 01:52 PM. Reason: Added horizontal directivity graph, removed more of the glitched phase data from the ND16 and changed from 3mH to 3.5mH on the Lowpass
    My first 2way build

  • #2
    Oh right. The cabinet model (nearly 0.5ft cu/13liters, port tuned around ~42hz) shows it should be able to reach ~99db before BSC losses with F3 around 40-45hz.


    Will that ND16fa-6 tweeter probably explode with that XO-point and roll-off?
    Or sound like poo?
    Or does that look fairly reasonable?

    I think the tweeter should only be seeing 4-5watts (way up high) and only around 0.5watts around the crossover.
    My first 2way build

    Comment


    • #3
      Not sure I'll change your mind, but that big oopsy around 1.5kHz is right where it will sound the most obvious. Voices especially might sound a little bizarre.
      Francis

      Comment


      • #4
        I believe a 12db/octave would work much better. Linkwitz / Riley crossovers work very well for two ways and only
        require two more components. I'd shoot for 2300 to 2600 crossover point.

        Comment


        • #5
          Originally posted by fpitas View Post
          Not sure I'll change your mind, but that big oopsy around 1.5kHz is right where it will sound the most obvious. Voices especially might sound a little bizarre.
          It is a bit unfortunate that it sims at a pretty respectable +/-2.5db (yay) but then has it all happen at once in a stark spot right where it's most noticeable (boo).
          I switched out the 3mH coil for a 3.5mH coil which seems to have helped a little.

          Originally posted by Regore View Post
          I believe a 12db/octave would work much better. Linkwitz / Riley crossovers work very well for two ways and only
          require two more components. I'd shoot for 2300 to 2600 crossover point.
          I totally agree. With only $5-10 more in cheap XO parts/cabinet you get SO many more viable driver combinations and can do a lot more with them.
          I suppose one positive aspect of the more limited XO above is that it probably wouldn't look as scary to someone who's very new to layouts/schematics, but a good printed layout is often enough to get around that fear.
          My first 2way build

          Comment


          • #6
            I'm thinking something like the Peerless BC25SC08-04 might be a better choice for a tweeter. At 98.8 dB, it will "keep up" with almost any woofer, it will realistically cross at 2600 Hz at full power, handle 100 watts and is available and not too pricey. The DC160-8 doesn't shine in a sealed enclosure, but in a ported 15 L 40Hz is easy. I haven't run it in Win ISD, but the HiVi M6N gets lots of good reviews. Factoid: Better quality (read more expensive) drivers integrate with simpler XOs usually. Johnny Richards has made a name proving this wrong, bless his heart! But the inexpensive drivers he used have gotten scarce. Best of luck, OK!

            Comment


            • #7
              From my very limited knowledge, it seems like the DC160 usually works best when crossed around or just below 2,000 Hz: as in the Classix II, Dayton BR-1 kit and Dennis Murphy's 'Murphyblaster' mods for the BR-1. I think they're all second order with Zobel.

              Of those designs, I've only built the Classix II and really like them. BC25TG tweeter, with a fourth order crossover so rather more complex and costly than your goal with this project.

              If (a big if) the 2000Hz crossover point is a reasonable assumption, a different tweeter may be needed?

              Depending on whether you want a 5.5", 6" or 6.5" woofer, you might also consider the 5.5" Peerless 830656 or SB Acoustics SB13PFC, both well behaved, smooth roll off and cross-able at 2500-3,000 Hz. The 6.5" versions of those drivers (830657 and SB16) are happier at about 2,000 due to their break up at around 4,000 (I think) so they need a small (0.22) and cheap cap on the woofer inductor. I've used all of those except the SB13, but it's well reviewed and widely used.

              Geoff
              Last edited by Geoff Millar; 05-20-2021, 08:56 PM.

              Comment


              • #8
                I think if you want a strict design goal of first order xos that could be a fun challenge. But I think you will get a better sounding speaker and a lot more drivers to choose from if you go with 2nd order xo's.

                Jeff Bagby's Quarks have a 4 element crossover (2nd order on tweeter + a resistor and 1st order woofer).
                Any reason specifically to limit yourself to Dayton drivers vs simply drivers available from from PE?

                Comment


                • #9
                  The limit of only 2-3 crossover parts was just for the fun/challenge after reading Drjay's "Amazing 2way" thread.
                  I don't personally think there's much/any practical benefit of restricting a build to 2-3XO parts when a couple more dollars worth of parts makes it so much easier to get flatter responses from a much larger pool of drivers.

                  In this particular situation, the very affordable drivers and slightly simplified+cheaper XO might (if the end result actually sounds good) be a cool starter project for someone trying to dip their toes into some DIY speaker building for the first time...but it's mostly just a fun limitation to try working around.

                  Sticking to Dayton drivers is mostly for the ease of having FRD/ZMA files to play around with and their decent blend of lowish price plus trustworthy specs. I noticed PE has FRD/ZMA files available for some other brand/s as well, but I'm not sure if that's common or a new thing. I hope PE expands their testing and file-sharing because it's really convenient and there are several other good price/performance brands that I've barely paid any attention to (I hear a lot of good things about some Tangband and Peerless stuff, for example).
                  My first 2way build

                  Comment


                  • #10
                    I think the SIMed off-axis response looks pretty good despite the high XO point on a 6.5" woofer. There's a slight dip between 4-6khz, but it's only around -2db compared to the surrounding response at a given angle which should be pretty subtle.
                    Click image for larger version

Name:	DC160-8 and ND16FA-6 directivity.png
Views:	219
Size:	19.5 KB
ID:	1469960
                    My first 2way build

                    Comment


                    • #11
                      Originally posted by LOUT View Post
                      I don't like the idea of spending a lot more for a woofer/tweeter that fits a single-part XO build if simply buying another cap+resistor+inductor and a much cheaper set of otherwise similar-quality drivers can do the same job at a lower overall cost.
                      The problem with this idea is that the lower cost drivers may be made to look good on paper, but they probably won't sound nearly as nice as higher quality drivers with a similar response. Especially with first order crossovers.

                      First order systems do not usually look as good on paper, but to my ear they almost always sound better. What do you want, a straight line of pixels on paper or good sound?

                      Drivers with flatter impedance curves are usually more expensive and are also much easier to use with a first order crossover.

                      I'm not saying that low cost drivers cannot sound good with first order crossovers. They can, but it is much more of a challenge. Challenges can be great fun too and you can learn a lot that way!

                      I will say that I was not able to successfully use the Dayton 16mm tweeter with a first order crossover. It just didn't sound right without a inductor in parallel, and I crossed it pretty high. First order with a damping resistor in parallel might work, I don't recall if I tried that. I have always thought that the DC28 tweeter could cross over low, but I have never purchased one to try.
                      Last edited by Billet; 05-22-2021, 02:27 PM.

                      Comment

                      Working...
                      X