Announcement

Collapse
No announcement yet.

Help pls.....crossover or filter advice

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

  • Help pls.....crossover or filter advice

    first of all, sorry for the back to back posts, i was wanting to to edit the previous title but i couldnt....

    I wont beat about the bush, i'm very green when it comes to desiging filters or crossovers and may have picked a tricky one to start and when i say green i would say non-existant.

    I'm wanting to do what the folks call a "fast crossover", where ill use a woofer upto 200hz and then going with a full range from 200hz to 20000hz. The more i'm thinking about this arrangement, i'm wondering if maybe the traditional cross over you would do with a tweeter doesnt really apply here as technically a woofer and a full range would be capable of playing the frequencies a bit better than a tweeter would and not actually have a cross over point like the traditional method would have? Not sure if all im really doing is providing a LPF and HPF instead as this is really no different to a sub and FR idea.

    So far i have one driver sitting about 10dbs lower than another running parallel on the SPL graph and another not even showing up, so definately in need of more research and training on my behalf.

    I tried watching a couple of youtube videos by kirby meets audio, which are good for me as i can play along side watching it, but i feel like im making it harder on myself with what i'm trying to accomplish.

    So the advice i need is,

    * Am i approaching this correctly for a Fast crossover
    * What is the best and easiest sim software i should be using for being an absolute rookie.
    * Does anyone want to tell me what cross over parts i should be using so i can try and sim it first? ie inductor and resistor on one driver and blah blah on another.
    * When does the enclosure sizing come into this simulation?

    I will be reading through a suggestion on the Paul Camody website re simulated measurements later today when i get a chance, so please be patient with me as i really dont know what i'm doing and keen to learn and willing to be hand held along this first journey if required.

  • #2
    FAST is a full-range speaker with a subwoofer. Typically this would use an active crossover either in a plate amplifier or receiver. That crossover will send the high-passed signal to the speakers and the low-passed signal to the subwoofer. The crossover will result in the lows and highs rolling off such that the combined SPL of the drivers around the crossover point are relatively flat.

    A two-way passive speaker is just going to use passive components to create a crossover that does the same thing, regardless of whether you are using a full-range driver or tweeter. Except now, rather than simply turning a dial on the plate amp or setting some inputs in the receiver to select the crossover point, you have to model the speaker and design an appropriate crossover using capacitors, inductors, resistors, etc..

    I don't know the easiest software. Maybe WinPCD?

    You can determine the enclosure size before any crossover work. BUT you need the T/S parameters. You can get these from the mfg specs but these may not be accurate for you particular drivers, so it is usually recommended to measure them yourself. That requires building some sort of jig (there is a sticky on the forum landing page) or buy something like DATS. You might be able to find measurements at other sites or other builders. Some manufactures are known to be more reliable than others. WinISD is a good program for modelling enclosures, although I think Vituix includes this functionality.

    Comment


    • a4eaudio
      a4eaudio commented
      Editing a comment
      Also: post the exact drivers you will be using. People may be willing to assist with the trace programs, extracting minimum phase, baffle step, etc.

  • #3
    Originally posted by a4eaudio View Post
    FAST is a full-range speaker with a subwoofer. Typically this would use an active crossover either in a plate amplifier or receiver. That crossover will send the high-passed signal to the speakers and the low-passed signal to the subwoofer. The crossover will result in the lows and highs rolling off such that the combined SPL of the drivers around the crossover point are relatively flat.

    A two-way passive speaker is just going to use passive components to create a crossover that does the same thing, regardless of whether you are using a full-range driver or tweeter. Except now, rather than simply turning a dial on the plate amp or setting some inputs in the receiver to select the crossover point, you have to model the speaker and design an appropriate crossover using capacitors, inductors, resistors, etc..

    I don't know the easiest software. Maybe WinPCD?

    You can determine the enclosure size before any crossover work. BUT you need the T/S parameters. You can get these from the mfg specs but these may not be accurate for you particular drivers, so it is usually recommended to measure them yourself. That requires building some sort of jig (there is a sticky on the forum landing page) or buy something like DATS. You might be able to find measurements at other sites or other builders. Some manufactures are known to be more reliable than others. WinISD is a good program for modelling enclosures, although I think Vituix includes this functionality.
    thanks a4eaudio, i only really use WinISD for the enclosures and pretty happy with being able to use that, but for me i've never ventured out from my simple designs.

    The two drivers i'm trying to incorporate into a build are;

    Peerless 830985 4ohm
    https://www.parts-express.com/Peerle...oofer-264-1050

    Visaton W 130x - 2 x 4 ohm
    http://www.loudspeakerdatabase.com/VISATON/W130X-2x4

    i bought the w130's for a potential boom box project as the manufactorer actually refers to it as a subwoofer, but seems to be more a general woofer. I would really be just looking to run the woofer at 4ohms, unless there is a reason to run the 2x4ohms part.

    so for me i have the t/s parameters for both drivers, which are just from the manufacturer spec sheets, i havent got around to building that jig yet. I've done the box sizing for what i want, which is around the 0.01cuft sealed for the peerless @ F0 approx 150hz and the W130x of 0.07cuft and a F0 mid 70's and f3 in the low 60's late 50's. I will be refining this further if it even works from the filter/x over perspective, but in the way of a vent was looking at 70mm wide, 10mm high and a length around 320mm. I'm considering this an active speaker as well, so most likely be looking to have a class D or T board in one, but thinking i may run out of space fast internally as the Dayton Boards are quite large and i dont know if the DTA2 class T board will have enough juice at 2x15w as that what i have currently in my desktop setup. I know i could just pair the w130x with a tweeter and attempt this a lot easier, but i have the Peerless drivers as well and they apparently have a good off axis response and im wanting something tall and narrow at my desk and too be honest a bit fixated on that W5 loudspeaker look with the side firing woofer.

    Comment


    • #4
      I've created and attached FRD and ZMA files for you from the charts provided for these drivers. I've run the frequency through Blender to extract minimum phase as well.

      The peerless is around 85db and the Visaton woofer ~78db.

      I recommend VituixCAD for crossover simulation.

      I thought I would take a look at the two drivers and what it might take to blend them as desired with a passive network. Here is quick starting point.
      Click image for larger version

Name:	Crossover.jpg
Views:	237
Size:	345.3 KB
ID:	1471786

      There will be some baffle step for the full range but without the box dimensions I'm not sure where this will occur.

      I'm also not exactly sure how to deal with the dual voice coil on the woofer so I have just assumed you can wire up a single voice coil as a 4ohm driver.

      The resistor on the full range brings its level down to match the woofer and the added notch takes care of that huge spike in output at about 13K. Nothing we can do about the huge dip after that though.

      The resistance of the coils will make a big difference here so as you run the simulation don't forget to check the real DC resistance of available coils and factor that in. Most simulators will only account for minimal resistance by default.
      Attached Files

      Comment


      • 3rutu5
        3rutu5 commented
        Editing a comment
        Thanks mate, I'm going to have a better read and see if I can replicate something similar. I really need to work through this stuff for it to sink in. I might be picking a fight with my 3d printer and trying to cram a 5 inch in it, with a 2.5inch and those components not to mention a vent but at least if I can get the software looking right I could experiment with a smaller woofer or two

      • Kornbread
        Kornbread commented
        Editing a comment
        wrong post edit

    • #5
      Why VituixCAD?
      I'll half-a$$ things up using the factory specs/measured specs and box sim/diffraction sim added into the equation to get a feel for whether the pairing might play nicely together, knowing the crossover will be later finalized, if from here, things look ok. Then, I'll go ahead with building the box/mounting drivers. Next, I'll get to the taking .zma and .frd measurements for the final'ish crossover sim. Then on to the does it sound right to me part of finalizing the crossover. Not as skilled as some on here who can sim it nuts on from start to finish, but it works for me.

      So, yeah, my advice at this point would be trying to figure out if the drivers might play nice together, not finalizing a crossover. Build it. Measure it. Sim the crossover. Tweak. Done. Easy Peasy. braaaaaaaaaaap braaaaap braap
      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


      • 3rutu5
        3rutu5 commented
        Editing a comment
        Why? Just because there was a couple of YouTube videos stepping through the process for my dumba$$

        My driver selection may be flawed a little as I bought them and looked at the winisd transfer graph for box sizing for each and not the others, guess that was the first mistake of many in my design life

      • Kornbread
        Kornbread commented
        Editing a comment
        Actually, I was questioning DeZZar, but that was one of the responses I was looking for; learning curve, advanced features, etc..

      • 3rutu5
        3rutu5 commented
        Editing a comment
        All good kornbread, I was feeling a little overwhelmed and just needed to put my hand up for help on this one. I had a few looks at some of the parts above and think that I'll be up around a passive aggressive kit in cost with some od those pieces.....but....I still want to have a go with the sim

    • #6
      Originally posted by 3rutu5 View Post
      first of all, sorry for the back to back posts, i was wanting to to edit the previous title but i couldnt....

      I wont beat about the bush, i'm very green when it comes to desiging filters or crossovers and may have picked a tricky one to start and when i say green i would say non-existant.

      I'm wanting to do what the folks call a "fast crossover", where ill use a woofer upto 200hz and then going with a full range from 200hz to 20000hz. The more i'm thinking about this arrangement, i'm wondering if maybe the traditional cross over you would do with a tweeter doesnt really apply here as technically a woofer and a full range would be capable of playing the frequencies a bit better than a tweeter would and not actually have a cross over point like the traditional method would have? Not sure if all im really doing is providing a LPF and HPF instead as this is really no different to a sub and FR idea.

      So far i have one driver sitting about 10dbs lower than another running parallel on the SPL graph and another not even showing up, so definately in need of more research and training on my behalf.

      I tried watching a couple of youtube videos by kirby meets audio, which are good for me as i can play along side watching it, but i feel like im making it harder on myself with what i'm trying to accomplish.

      So the advice i need is,

      * Am i approaching this correctly for a Fast crossover
      * What is the best and easiest sim software i should be using for being an absolute rookie.
      * Does anyone want to tell me what cross over parts i should be using so i can try and sim it first? ie inductor and resistor on one driver and blah blah on another.
      * When does the enclosure sizing come into this simulation?

      I will be reading through a suggestion on the Paul Camody website re simulated measurements later today when i get a chance, so please be patient with me as i really dont know what i'm doing and keen to learn and willing to be hand held along this first journey if required.
      Here's how I would approach something like this design. First off, it's really still a normal 2-way, but the xo point is much lower. It's kind of like a 3-way, but deleting the tweeter to save money. The first thing I looked at was the Peerless frequency response. It looks pretty smooth with no real glaring problems that I can see. The spl is about 83dB up near 10k, and a little more at 1k. Looks workable. The woofer looks smooth, and has no crazy peaks or dips to deal with. Should be workable. I can't tell how loud it plays at 2.83v, but I'm assuming it's not too loud for the mid. Crossing the two drivers should be rather easy. I think it would be easier at 500hz, than 200hz, but I'd probably prefer under 300hz. It may take more parts to do so. If you run a sim with the Peerless mid files entered as the tweeter, and the woofer files entered as the woofer, then you would simply work with it like any 2-way sim. Enter a target response, and try to match it. Make the spl of the targets different, so that the woofer is louder.

      On the Peerless, I'd start with a simple cap, and see how it looks. Maybe go 2nd order to improve the roll-off. Depending on baffle size, the mid will start dropping below 800hz without help from the filter. With a first order filter, the mid-range might level out making 1k the same spl as 10k, which would be my goal.

      On the woofer, see if you can get a LR2 200hz response without the deeper bass humping more than a few dB. Low xo points cause this worse with some drivers than others.

      When the two add in the sim, you want the bass to be louder than the mids and treble. Somewhere between 3dB and 6dB louder, with a gradual transition from one to the other.

      Comment


      • 3rutu5
        3rutu5 commented
        Editing a comment
        Thanks mate, gives me a.few things to look at, I think because I've bought the drivers for different projects and now trying to make them work I might be tieing one hand behind my back. But let's give this simming a go

      • rpb
        rpb commented
        Editing a comment
        I mix and match drivers all the time. Many I bought long ago. Trial and error x-overs are my preference, but sometimes take longer. Sims are fun too, and useful when you get stuck, or have to order xo parts.

    • #7
      Originally posted by Kornbread View Post
      Why VituixCAD?
      Well, its very good software with plenty of features, easy to use, doesn't take much to get up and running and the author is active and constantly updating it. It comes with excellent guides and support and provides all the functionality you need to completely engineer a crossover (passive or active) either from real measurements or manufacturers plots.

      It allows for 8 separate simulations/versions of your crossover in a single project so you can keep track of iterations. It can consider driver layout (three axis), has the ability to handle 360degrees worth of horizontal and vertical measurements per driver, and has the ability to optimize any part of your circuit with a goal-seek type algorithm that works through hundreds of iterations in a few seconds.

      It has a bunch of handy pre-defined circuits in a library that you can use to quickly get the ball rolling on a design, you can leave 'locked' with their own formulas for ongoing fine-tuning or take over the values yourself and even allows you to define your circuits to add to the library.

      This is of course just in addition to a very decent user interface with a drag/drop snap-grid style interface for the circuit design.

      Its other embedded functions are brilliant as well. For instance it's trace function is one of the best I've seen. It also has diffraction/baffle step simulation and handy calculators for things like your time interval for gated measurements given the dimensions of your setup, wavelengths, nearfield measurement calculations, box volume and time alignment to acoustic centers.

      It can merge near and far field measurements, and it can combine measurements with a plethora or options.

      It's graphs are easy to navigate in and out of and set the limits and view and you can see 6 charts on screen at once covering the SPL, Power and Directivity Index, Directivity Polar plot (which includes multiple chart options), Group Delay & Phase, Filter Response and Impedance. You can export any one of these or all six at once. You can even add your own overlays to these charts - for instance, you might overlay the results of a previous speaker build within the SPL chart of the current project so you can easily see differences or perhaps target a similar voicing.

      It even has a section covering room response/reflections that I've never really played with.

      Honestly, considering the engineer offers it for free (which is astonishing), covers so much functionality and responds to emails - I don't know why anyone would bother with any other crossover tool. Just my opinion (and I've paid large sums for other professional tools).

      Comment


    • #8
      I suk at learning new software.
      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


      • #9
        Originally posted by Kornbread View Post
        I suk at learning new software.
        Lol. I get it. We lead busy lives. And sometimes when we open new software we just want the knowledge to download into our heads as well matrix style. But realistically, it took more effort to write the above post for you than it would to download the software and read through my basic get started guide that I linked.

        Comment


        • #10
          Originally posted by DeZZar View Post
          Lol. I get it. We lead busy lives. And sometimes when we open new software we just want the knowledge to download into our heads as well matrix style. But realistically, it took more effort to write the above post for you than it would to download the software and read through my basic get started guide that I linked.
          Haha I'm the same, I'm a civil designer by trade so have a few CAD/Design packages under my belt, just afraid after all this time I just don't retain it as good as the last few decades....don't mean I won't try

          Comment

          Working...
          X