No announcement yet.

First speaker build, TriTrix.... already deviating. Could use some crossover help.

  • Filter
  • Time
  • Show
Clear All
new posts

  • First speaker build, TriTrix.... already deviating. Could use some crossover help.

    Greetings All,

    Quick history: new to the forum (been lurking for a while), but have been repairing/building speakers as a hobby for a while. I have a real interest in audio equipment, as it feeds directly into my love of music. I mistakenly decided to try and upgrade a pair of my grandfather's Audio by Fisher speakers with new components. I was looking for more accurate reproduction and clarity than the "vintage" speakers could. Due to incompatible crossovers, the result was lackluster. It was painfully obvious in retrospect that I should have considered the drastic differences in response compared to the stock 1970's speakers (not to mention crossover points, enclosure dynamics, and just about everything else). I decided to harvest the new speakers I had purchased for a complete build. The TriTrix design seems to fit most of the size, performance, and aesthetics I was looking for.

    Here are the differences:

    1) DSA135-8's vs. DC130BS-4's (similar response, with a resistance difference)
    2) DC28FT-8 vs. DC28FS-8
    3) Custom or Dayton 2-way crossover to account for the woofer resistance differences

    From what little I know about crossovers, it looks like the stock TriTrix uses a 2nd order with some resistance compensation or attenuation? Maybe?

    I'm familiar enough with fabrication techniques to fit the drivers and build the cabinets. I need assistance with the crossover. I'm considering using the Dayton premade 2-ways at anywhere from 2.5K Hz to 4.5K Hz. It's either that, or I try to build a custom crossover. I've used a few of the on-line calculators to see what I'm in for. However, my lack of experience and knowledge with regards to design will likely lead down a similar path, where I'm not pleased with the results after considerable effort.

    If anyone could help me decide on an existing crossover, or to suggest changes to the specified crossover, I'd really appreciate it. I'll be forced indoors over the winter, and I'd love to have this project on the road to completion before the snow flies.

    Thanks in advance for the help.


  • #2

    You will get replies from people who know much, much more than I do, but here are a few comments:

    are you building the MTM transmission lines, the MTM vented cabinet or the TM version? I've built the MTM vented version and am very happy with the performance.

    Mr Campbell designed the cabinets and crossover for all variants with the specified drivers in mind: if you search this forum, and the PE Tritrix Q and As, you'll find that the DC130 is the only driver recommended for this project.

    For example, people have asked about substituting other woofers in the transmission line cabinet, and the answer is a definite "no". I know nothing about TL enclosures, but the cabinet was designed around the DC130's parameters; unless the DSA had identical performance and specs it would not work as well as the DC130.

    I don't know as I haven't used it but would expect the DC28T to have very similar performance to the standard DC28F

    off the shelf crossovers won't work properly with (almost) any project. For example, feed the data from the spec sheets into a crossover calculator and compare what you get with Curt's brilliantly simple crossover

    why do you want to use a different woofer? Many, many people have built the Tritrix over the years and they're all very happy: check out the reviews! If you use the transmission line cabinet - I couldn't due to space issues - its bass is regarded as quite sufficient for normal listening.

    Having said all that, there are several if not many members here who could help you out with a custom crossover for those drivers

    Good luck



    • #3

      Thanks for the reply. Short answers:

      I was planning on the MTM TL design. The cabinets will be smaller than the current "monkey coffins" I'm using. The only real reason to use the different woofers is to utilize the speakers I already have in my possession, not to mention a few of the specs seem to be superior, while maintaining many similarities. However, as a true novice, I may be misinterpreting the spec sheets, or underestimating the importance of certain parameters with respect to the cabinet design.

      Thanks for the information regarding the pre-built crossovers. It seems it's down to buying new woofers, or braving the wilds of crossover design...



      • #4
        NB the DC130s in the Tritrix designs are the 4 ohm version, wired in series: the DSA135 is an 8 ohm speaker. This obviously has implications for crossover design, if you connect two DSAs in parallel this will be 4 ohms load.

        Looking at the woofer specs and understanding nothing at all about TL design, the calculated ideal vented cabinet volumes are quite different: 0.2 cu ft for the DSA compared with 0.36 for the DC130. I would guess that this alone would raise some questions about how the DSAs would behave in the TL cabinet.

        Maybe use the DSAs for another project and build the Tritrix as designed?



        • #5
          So, YOUR design (which isn't really a "TriTrix" at all) requires a 4-ohm stable amp. due to the paralleled 8-ohm woofers.
          The modeled impedance dips down to 3 ohms around 200Hz (which is normal for a nominal 4-ohm design).
          The T/S parms indicate that a vented box will be nearly 1/2 the (interior) size as the TriTrix design.
          I'd go with 0.5cf with a 2"id port tube that's 5" long. This yields an Fb and F3 in the low 50s.
          Keep the baffle width and driver spacing the same (as the TriTrix), and decrease the volume by making the cabs shorter and/or shallower.

          The HP (high pass filter for tweeter) needs a 0.5ohm series resistor and a 10uF series cap, with a 0.50mH shunt coil (to gnd) across the tweeter. If Dayton has a #20, use it. We need to "un"-pad the tweeter a bit to keep up w/the paralleled woofers.
          The LP for the paralleled woofers uses a 1.0mH series coil, and a 50uF (npe - a 47uF would also be fine) shunt cap. A cheap coil (#20) with a higher DCR will give a little less BSC than a low DCR (iron core) coil. Cross point is roughly the same. The DC woofers in the TriTrix only have about half the Xmax as the DSAs, as a result, the DSAs should be able to take 2x to 4x the power before excursion limits are met. Could be interesting.


          • #6
            Mr Roemer, that's amazing!



            • #7

              Sincere gratitude for looking into this with such detail. With my limited knowledge of enclosure / TL design, I would not have guessed that the few speaker parameter differences would drive such a drastic change in the cabinet volume. Can you describe which T/S's drive such a difference? I'm genuinely curious.

              I can try to sketch up a cabinet that might decrease the volume as you say. I'm a little concerned about the changes in geometry. Would it be effective to "raise the floor" of the cabinet, while maintaining the appropriate baffle angle? I assume I'd seal the back of the cabinet, then add the port at the rear? Most of the TL speaker cross-sections I've seen open to the front or back.

              Mr. Millar might have the best suggestion: buy the original drivers and enjoy the speakers! The sensitivity of the design to the changes suggested seems to increase the risk of "sub-optimal" performance at best.

              Thanks again for the excellent information.


              • #8
                Jason, may I suggest you read the Speaker Building Bible sticky at the top of the forum? Mr. Paul Carmody's comments on "I wanna build this design but with these drivers" are a wonderful tutorial.


                • #9
                  Be prepared to learn alot about passive crossovers!!
                  Some amazing fullrange drivers are out there.
                  A good first project would be full range speaker.
                  Less problems to solve.
                  Very little parts. They can be powered and thus active from any audio source. 2 channel to 7 channel!! No problem!! Every finished speaker should be tested after for response curves, phase etc
                  Not sure about free software anymore.
                  I myself was hoping to use Speakers Workshop that did some measurements like nearfield, gated impedance peak etc.. I liked that software!!
                  Left in the dust I guess. Looks like Sound Easy now!! One day soon.
                  You can save on some great designs that have been released for a Everything figured out already.
                  I think MTM are overated!! Lol
                  Ive had many.


                  • #10
                    JTA, the "big 3" T/S parms are Qts, Vas, and Fs. The 1st 2 drive the cab's volume requirements, and "how low it can go" is driven by Qts (also) and Fs.
                    Volume varies proportionately with changes in Vas. Bottom end reach (F3) varies directly with Fs.
                    As Qts rises, so does target volume (but not proportionately), BUT . . . F3 goes down (also, not proportionately).
                    A GREAT (paper) book to self-educate yourself is "SpeakerBuilding 201" by Ray Alden.

                    At a Qts right around 0.40, you can usually get near optimum results by building a vented box with a volume (Vb) close to the driver's Vas value, and it should reach down to about the driver's Fs limit. A good (FREE) program for simulating all this is WinISD avail. at .

                    If both woofers had a Qts of 0.40, then a single DC130 should reach down to 59Hz in 0.4cf. For a pair of drivers, you need to double the box volume (so . . . 0.80 cu.ft.). A pair of the DSA135s should reach 52Hz in 0.56cf.

                    BUT: the DC130s have a "Q" (Qts) a tad higher at 0.42, which means they can go a bit lower (maybe 57Hz) in a tad larger box (about 0.90cf).;
                    and the DSAs have a Q a tad lower (0.38), raising their F3 to about 54Hz, and dropping the Vb to about 0.47cf (which is why I said about half the volume).

                    On a box design "difficulty scale" of 1 to 10, designing a closed box (which is typically about 1/2 the vol. of a vented box - at a loss on the bottom end of approx. 1 octave - or nearly so) is around "2", a vented box - maybe around 4-5, and a TL about 8-9. There are those who think there's no better bass than a TL (which do seem to need to be larger than a vented cab), but many think a vented box comes very close in bottom end reach, at a bit of a savings in volume requirements (and are much easier to design). WinISD can do closed and vented boxes, but not TLs.

                    A GREAT place to put a rear vent is directly behind a tweeter (since they're not "open back" and there's no sound coming off the back of the dome to "leak" out the port tube).

                    You COULD build a "floorstander" the size of the TriTrix MTM TL enclosure, and just raise the floor (as you've said) to decrease the upper (working) cab volume. The bottom ccan be empty, or filled with some "weight" (like sand, or some use kitty litter) for added stability. The false bottom is also a good place to put the XO (crossover).

                    Just position the drivers on a baffle (the same width) using the same driver layout (inc. horiz. offsets), with them also the same distance down from the top of the cab.

                    With these DSA woofers having double the Xmax, I'm pretty confident that you'll end up with a very similar "sound" (to the TriTrix) with greatly enhanced power handling (provided the tweeter doesn't get over driven).


                    • #11
                      Mr. Roemer,

                      Thanks again for the insights and assistance. I will take some time this weekend to draft up some thoughts, based on your suggestions.

                      Thanks again for everyone's inputs! Given the feedback on design and construction complexity, I think I'm going to stick with the original design for the first try. If I like it, then I'm going to pursue the modified version. In the meantime, I can work on getting the cabinet design in the ballpark, while trying to figure out appropriate crossover topology as suggested by Chris. Or... start with a simpler design altogether!

                      Also, I began reading the stickys as suggested... excellent info. As a person that appreciates a great stereo, I've recognized the mountain of information and experience that's required to design and test great sounding speakers. Now looking into the gritty details, it's even worse!