Blog Forums Member Area Hello Everyone

4 voices
6 replies
  • Author
    Posts
    • #15339
      jrand
      Participant

      Hi!

      I am going to start building a pair of DINAS soon but although I had not even finished that build, I am already planning a second one, also inspired by Toids design (actually, I would love to just be able to copy his design, as then I would be on the safe side considering mistakes, but I am having fun learning how to do it myself and understanding what goes behind building DIY speakers).

      I am a physicist by trade and in my specific field bode plots (or frequency response plots) are quite common, so I already have a little experience with filter orders, resonances, transfer functions etc. which, I guess is handy…? Also, I hope that this can serve to help with the community somehow if my knowledge can answer someone’s question.

       

      Either way, I love DIY and as music has always been something incredibly present in my life, I guess it was only natural until I would find myself delving into the DIY world of speaker building. I do not believe that this will become a full-fledged hobby, but I can imagine that I will probably build at least 3 pairs of speakers in my life….but…also knowing myself I can also imagine that that number will be 5…7…maybe even 10 at some point plus 5 pairs re-engineered, 6 modified etc. Oh well.

      At the moment I have been experimenting with winISD after checking several videos and will probably post some questions soon as well.

      I wish you all great builds and I would also like to thank Nick for his contributions to this community!

    • #15340
      123toid
      Keymaster

      @jrand welcome! It is great to have you on board! Feel free to ask as many questions as you like. Sounds like you’re going to have quite the background already. 

       

      The website’s going to be changing a little over the next week, which should help with the social aspect. That way this community can continue to grow.

      I’m excited to hear that you’re starting with a dinas, those are great speakers! You’re going to really enjoy them! Out of curiosity did you go with the passive or active?


    • #15341
      ajc9988
      Participant
      Posted by: @jrand

      Hi!

      I am going to start building a pair of DINAS soon but although I had not even finished that build, I am already planning a second one, also inspired by Toids design (actually, I would love to just be able to copy his design, as then I would be on the safe side considering mistakes, but I am having fun learning how to do it myself and understanding what goes behind building DIY speakers).

      I am a physicist by trade and in my specific field bode plots (or frequency response plots) are quite common, so I already have a little experience with filter orders, resonances, transfer functions etc. which, I guess is handy…? Also, I hope that this can serve to help with the community somehow if my knowledge can answer someone’s question.

       

      Either way, I love DIY and as music has always been something incredibly present in my life, I guess it was only natural until I would find myself delving into the DIY world of speaker building. I do not believe that this will become a full-fledged hobby, but I can imagine that I will probably build at least 3 pairs of speakers in my life….but…also knowing myself I can also imagine that that number will be 5…7…maybe even 10 at some point plus 5 pairs re-engineered, 6 modified etc. Oh well.

      At the moment I have been experimenting with winISD after checking several videos and will probably post some questions soon as well.

      I wish you all great builds and I would also like to thank Nick for his contributions to this community!

      So what software do you use to do simulations? Are you good at Comsol? If so, can you give some advice on setups of the software for doing simulations? I want to pick your brain.

      Right now I’ve used Inventor with NASTRAN and have done modal analyses, but I want to step up the simulations to do even better.

      My method is to start with WinISD for figuring out the rough size, the length of ports, etc. I then move to an excel spreadsheet to figure out dimensions and finalize roughly the size, etc. I then take that to Inventor. Then the simulations. Etc.

      Since you are familiar, I assume, with the simulation softwares, anything you could give on advice would be wonderful.

      Also, you need a DATS or similar to get the T/S parameters of the drivers (spec sheets do not always match). And a microphone for measurements (specifically a calibrated mic, the umik-1 from miniDSP is a good, affordable model).

       

    • #15342
      elliottdesigns
      Participant

      @ajc9988 Wow, you jumped straight in there with that comsol question didn’t you? 🤣

      I’d be interested in that too by the way since I’ll be using that program at uni next year 😂.

      I was just going to say, simulation isn’t the be all and end all. Whilst accurate to the models, the models can never fully represent what is built in real life, no matter how complex the model, especially in acoustics and more specifically DIY. This is why we then build prototypes, test them and sometimes make adaptations, effectively you get the law of diminishing returns from software. Specifically, when DIYing a loudspeaker, it’s typically MDF, the MDF panels can be joined in many ways (changing how the deflection occurs), many different qualities of MDF exist, meaning the properties of the material vary(rigidity, damping), the tolerance of joins, etc, etc. I’m sure you see my point, they all move away from ideals used in modelling software. Therefore (in my opinion), there is something as too far in terms of trying to achieve simulation accuracy when doing DIY loudspeakers.

      …anyway, I digress (as per usual) 😂

      Totally agree in terms of DATS being super useful and the manufacturer values not always being the most accurate (particularly cheaper models as tolerance typically worsens as the driver gets cheaper due to manufacturing costs, from what I’ve seen online). Still saving up for one myself though 😅

      A piece of software I use called LTspice also simulates a some resonances from box geometry (well their affect on the frequency response anyways). I have a video planned for that, but it will come out after my tutorial on using it to build transmission line enclosures. So no idea on when that’ll be released 😂. (I find it interesting since it actually helps you make a sealed box with less resonance since you can design a sealed transmission line).

    • #15343
      elliottdesigns
      Participant

      @JRAnd Welcome to the forum!!!

      We’ve had a few new people here recently which has been amazing! As with Nick, I’m very interested to know whether your builds have been active or passive (or some of each)!

      It’s really nice seeing how wide a variety of professions people have here! Happy to see we now have a physicist, I’m sure there’s plenty you’ll be able to bring to discussions, and I’ll be looking forward to having them with you 🤞😁

      • #22291
        jrand
        Participant

        Wow, that was hyper direct @ajc9988! 😀

        So, a bit more background, my main field is ultrafast optics, experimental. I code extensively, nowadays, solely python. A few years back I did everything in C++. I do a lot of experimental control (so basically computer-scientific hardware communication, APIs etc.), also UIs, some simulations, but my bread and butter is data analysis. Even so, I do not consider myself a great coder, because that is still supposedly a small part of my supposed job, which is to build and use lasers.

        About Comsol, never used it as its not a typical use case for us. I sometimes have deal with “theoreticians” who use finite element to model electromagnetic modes in nano devices, but I have never touched software like that.

        I do CAD because most of our research involves custom solutions, so I am forced to design optics holders or some sort of hardware using stepper/DC motors (in my previous lab we had several custom devices that used a huge speaker which had its membrane cut out to drive a custom linear stage using the coil, with high precision). Unfortunately I learned to use Solidworks in my previous lab to a quite good degree, my current lab has a different software and at home I’m using fusion360, and the differences are enough in both that I see the speed at which I design things in CAD fall by like 50% or more.

        Now, I agree with @ElliottDesigns that models can help but will probably fail once the experimental data is at hand. For example, I noticed in WinISD that in certain conditions a change of 1mm on a slotted port’s width, can alter considerably the needed length and hence tuning of the port. So, if I install a slightly crooked port, ending with flaring (I am calling flaring in this case the effect of the total area/perimeter of the slot to be dependent on the position along its length), I guess the tuning and air velocity will make weird things, as your waveguide now accepts different kinds of modes. I am supposing it is like that, just from playing around with WinISD, which already tells me that building well sounding speakers is much much harder that it seems, even when following plans.

        Nonetheless, one thing that bugged me with winISD is that I can only plot one curve at a time, and my first instinct was to write my own software with an UI that would let me show several parameters at once. But I did not look at what kind of equations go in the back of winISD, if they are simple functions or there is some sort of PDE solving in the background for them. Either way, if I ever get to write such a software I will surely share it here as open source….which will be shameful as then everyone will be able to read my horrible coding :p

        Lastly, @123Toid, I am building the active version of the speakers, as currently I do not own any amplifier and I only found out about the passive after having already ordered all the parts for the active. I will probably get an aiyima tube-preamp to be able to change sources and have a BT connection.

        But, sincerely, these welcome messages felt incredibly warm, thanks guys for the curiosity and interest!!

      • #22315
        ajc9988
        Participant

        LOL. I’m an attorney, not currently practicing. So, when I want information, I usually am pretty direct (same with asking out women; throws some off, but others, even if rejecting me, appreciate the direct manner in which I ask as it does not waste any of our time and we know where we stand).

        And the equations are complex in that all of them make assumptions, and as such, there are numerous equations to determine, for example, proper port length and length correction for flare, etc. For example, Bass Box Pro comes close to, but provides a different length value for, the length of a port being calculated. The truth is, unless you have a good equation backing you up, and depending on the accuracy of the values of the variables input, you will not know if you are where you need to be on values.

        Because of this, if possible, be prepared to put the port in and be able to adjust the length when constructing. Also, cool trick to doing a vented box is to use a DATS to do an impedance sweep. This will give you the tuning value of the box. The tuning value is related to the length of the port (truly, the volume of the port, because the air is creating a weight that is loading the diaphragm/cone, so by adjusting the volume of air being moved in the port, you adjust the tuning of the box), so the DATS will help to hone in on the port length when testing the build. That is also why when you vary the factors otherwise, it varies the length of port needed in order to calculate what the equation estimates should give you the tuning frequency of the box when you use WinISD. What I usually do for finding the value for the box tuning is to setup the high pass filter in order to prevent x-max higher than the natural x-max of the speaker (the hump to the right being as high as I allow the hump on the left) in order to prevent bottoming out the driver. In other words, I put what is sometimes called an infrasonic filter, which is a high pass filter for crossovers, in order to keep linearity and prevent destructive behaviors of the driver. I then, after setting that value, return to the box tuning and adjust those values. If you are using insulation, you can also try to estimate absorption losses to compensate for that effect on the box. That is found in the advanced tab at the bottom of the Box tab in WinISD. Lined walls are sometimes considered around a value of 20, whereas fully stuffed is considered 5. It defaults to 100. Once again, a model is only as good as what values you put in.

        After I do the recursive practice on the box tuning and the filter values, then I add in the vent size and dealing with air velocity. With that, I try to figure out a way that while building, I can pull the last bit of the vent over to the belt sander and take some off to tune the box during the construction phase. Then, after you hit close enough to the value, glue or epoxy it all up and in place and move to the next phase.

        If you do make that software, let me know. I would definitely be interested. I remember in high school (I no longer have this book), I had a book on how to calculate the box by long hand. It had integral formulas, but at that point, I was in like high school calc I or trig (forget which for when I got the book versus my final math class in high school, where I did a bit beyond calc I, but in college, definitely failed calc II (old joke: you know what they say about the pre-reqs of calc II, you need to have taken calc II. LOL. ). At that point I went to philosophy and then law, but…

        But happy you are here and ready to learn. There are all sorts of books on cabinet building if you care to go through the equations. From Loudspeaker Design Cookbook to Bullock on Boxes. One review of Bullock on Boxes says the following:

        “I must say I liked this book. I have read a lot of papers and books on this topic and I think this is one of the better ones. I am a mechanical engineer, and while I can feel for some people who don’t think this is a beginner’s book, in reality, this book is not that bad from a technical standpoint. There is no real theory presented, simply equations to calculate various performance parameters.

        Pros:
        – More discussion I have seen in any book regarding alignments. I really have no desire to expand into filter theory, so this was very much appreciated.
        – Examples. The author gives some good examples.
        – The author actually states references! How very proper for a technical publication.
        – For the most part well laid out and easy to read.

        A Pro and a Con:
        – There are a fair amount of mistakes. I can’t quite tell, but they may mostly be due to printing errors and not on the part of the author. However, the author seems to go out of his way to correct them. Make sure you read the entire book, including the correspondence section at the end to make sure to catch all of the errata.

        Cons:
        – The author does, like ALL technical books, at some point, make large jumps in discussions to arrive at a point. It is unfortunate, but I can’t think of any book that doesn’t do this.
        – The author does a poor job in putting into words what he is discussing. For example, when talking about vent tubes and the correction factor, k. I had to read that small paragraph about 15 times to understand what was being said. A simple picture would have been much better (not to mention that there is NO reference to where these magic k numbers actually come from… a pet peeve!).
        – The author does make a basic mistake when it comes to vent tubing. The author states that he uses PVC pipe for his vent tubes. If he indeed does this the dimensions for his vents will always be wrong. When dealing with pipe, the nominal dimensions of the pipe are not the actual sizes. To get the actual sizes of what you are using, you need to measure the pipe or use a pipe chart. For example, a 1″ pipe has NO dimension equal to 1″. It’s OD will be 1.315″ and the ID will be dependent on the schedule size (wall thickness) of the pipe you buy. If you use pipe, make sure to make adjustments to your calculations as necessary.

        All in all I like this book and go back to it frequently. Well worth the money.”
        https://www.amazon.com/Bullock-Boxes-Robert-M-III/dp/096241915X

        So I hope that clarifies a bit about the equations used, even if it further muddies the waters. Welcome to the forum!

You must be logged in to reply to this topic.

©2022 KLEO Template a premium and multipurpose theme from Seventh Queen

CONTACT US

We're not around right now. But you can send us an email and we'll get back to you, asap.

Sending

Log in with your credentials

or    

Forgot your details?

Create Account