DIY Synergy/Unity spreadsheet

Discussion in 'Waveguides and Horns' started by BillWaslo, Apr 29, 2012.

  1. 2k would be near the tweeter/mid crossover, so could be an issue with either (or their crossovers?). How big a bump? Maybe the mid port holes (try filling some of them)?
     
  2. Hi Bill,
    The holes are 5/8" and you can see how I cut them at few post back. The dip like a deep valley down to the bottom and gradually come back up to around 4k. Could it be easier to make change on the crossover?
     
  3. Maybe post a picture?
    A deep valley usually means something is cancelling. Do you have the relative polarities of the midrange and tweeter correct? (Note: depending on the crossover design you use, they aren't necessarily same polarity! - also watch out for mislabeled drivers, I have some here that are marked wrong....
     
  4. Hi Bill,
    I'll try post the pictures asap.
    The crossover I'm using is the one on V.5 of which I direct copied the high and the mid.
     
  5. Hi Bill,
    Here come the two graphs I took at 2.48K and 390hz, both are with 1/12. It seems to me that they're right at both ends of the mid drivers.


    IMG_20160604_153714.jpg IMG_20160604_153725.jpg
     
  6. Try inverting the polarity of the midranges
     
  7. BTW, no need to take a screen shot with your phone camera -- just click on the graph form, do Alt-PrtSc on the keyboard, then open PaintBrush and paste the image. You can then save it to a file, much better graphics than a camera shot.

    It might be a good idea to disconnect the drivers sets (woof, mid, tweet) except one and get plots of each set by itself to see what is going on.
     
  8. Hi Bill,
    Thanks for the guidance. I'll try everything recommended. Let you know.
     
  9. Hi Bill,
    The last couple weeks were the funniest and hardest weeks for me. I had been busily working with the crossovers and together with the speaker modification. Please check the graphs below and advise if there is room I could make improve.
    I checked the phase and there is nothing wrong with the design. I have the crossover point changed a bit, they work out fine with a little bit EQ. I'll post some pictures of the speakers later.
    Albert
     

    Attached Files:

  10. Hi Bill,
    As you can find on the graphs that there are still some serious dip at 2300 and the roll off above 10k. I did try serious EQ on both areas but didn't do too good. At the 2300, could it be the problem that I cut the hole too deep? yours only cut down 1mm as I could remember. The 10k thing I don't have a clue what I have done wrong.
     
  11. alberti, I don't see a 'serious dip' in either graph -- those are 3dB per division plots, right? Those both look flatter than mine do (at least before I add EQ). Look at some REAL plots of commercial speakers (or measure a few), see how very few are truly flatter than yours.

    At the high frequency end, same comment. Two things I note are: the graph horizontal is "log frequency" and the vertical is 3dB/division. Where the graph hits the -3dB line, the frequency is about 18kHz? Unless you go with a beryllium tweeter (or eq a compression tweeter so that its breakup peaks look flattish), or a non-constant directivity horn (that peaks the treble, but only on-axis), you're not likely to do better than that.

    And unless you're under 5 years old (and your dad is reading this to you?) your hearing is very unlikely to be usably sensitive anywhere near 18kHz!
     
  12. Bill,
    Yes, they are all -3db/division.
    Yes, they sound absolutely fantastic if compare to all speakers I built before. My wife gave the same comment as she sat by my side for over an hour on their trial run. The speakers sound more mature the longer I play music on them even though they are heavily EQed. Don't know how to praise them, they sound really good. I'd encourage anyone who have minor skill on wood go make a pair and you won't be regret.BTW the subwoofer at the bottom helps from 70 down.
    Thanks again.
    Albert
     

    Attached Files:

  13. I keep on modify this horn to spec. as close as possible and end up I have better response at high frequencies. I even have better response on that bump at 2.3k by fulling the hole with bondo and cut the hole to 9/16". I only need very minimum EQ at this range now. I have the color changed to gold too. A very happy camper I'm.
     

    Attached Files:

  14. Looking good, albertLi! I like the gold color, too.
     
  15. Big thanks to you Bill.
    Let me know if you have any new horns coming up. I'll certainly make them.
     
  16. Thanks Albert,

    Well, now that you mention it...
    I'm in the (i hope) finishing stages of a new design "SmallSyns", a miniature Synergy type horn with good directivity, using one of Erich's SEOS15 waveguides as a starting guide. Theres a thread started over at DIYAudio (multi-way section). When finished, I plan to put big pressure on Erich to offer it as a kit.
    http://www.diyaudio.com/forums/multi-way/292379-small-syns.html
     
  17. Bill,
    It's good to know
     
  18. Hi Bill, do you by any chance still have the images you posted...the links are all dead? I am getting ready to build a 2 way synergy horn, and am trying to get all the info I can to get it correct.

    tia,
    Ron
     
  19. Ron, can you describe which pix you need (so I don't have to scan all through this to figure out which)? A lot of my pix are on my server at http://libinst.com/SynergyDIY/ , though not organzed in any particular order and with several projects mixed in. If you don't find what you're after there, I probably have some other on a hard drive here.

    Bill
     
  20. Hi Bill
    Thanks for this wonderfull thread !
    This winter will be the one for me to start a project.
    I managed to simulate the Mid´s, using your horn data and the Visaton FRS5X 8 ohms.
    It looks allmost excactly like the Sim with your Gento´s (now obsolete). So far, so good.
    This speaker could be a good candidate for the Mid´s.

    Response.jpg

    Input data.jpg

    But, to avoid too much sawdust, I have a few questions, regarding your simulation versus your actual build :

    1: Your (and everybody else it seems) S1 is 2,53 cm2.
    This is not 1" dia (5,06 cm2), nor the inscribed square of that (3,2 cm2).
    Why 2,53 cm2 for S1 ?

    2: You simulate a horn with throat S1 = 2,53 cm2.
    Then build a horn w. S1 = 1,79 cm square throat (0,707" square).
    Finally this is drilled to 1" (2,54 cm) dia. cylindrical.
    This is twice the simulated area, which I believe could upset the correlation of simulation vs. build.

    3: Is all of the above because you somehow include the CD driver´s inherent flare distance and initial area ?
    Cell C29 of your spreadsheet (distance of port to the throat) looks suspiciously short, considering that a quarter wavelength of e.g. 1,2 khz (X-over frequency) is about 7,2 cm.

    4: Your simulated port length, Lpt, is 1,02 cm, but your text and pics. tell you grind it down to 1/8".
    Wont that upset the correlation of simulation vs. build as well ?
     
  21. Hi Barniboy,

    Keep in mind that there are 2 purposes to the spreadsheet --
    1) set up values for the midrange simulation
    2) calculate board dimensions and angles using flat stock.

    Since the walls are made with flat material, there's no choice for #2 but to calculate to a square throat. But the tweeter exit is circular, so we design for a square throat with diagonal dimension matching the tweeter exit, then drill it out with a 1" drill. That also provides a decent gradual transition from round tweeter exit to rectsngular horn.

    But why simulate with the smaller area square throat? That's because this is a MIDRANGE simulation and the midrange doesn't care much about that throat area (always think about the wavelengths involved for the midrange) -- only about the distance to the tweeter diaphragm and the expansion rate where the mid ports enter, which to get right has to match the horn designed for a square. The tweeter behaviour would care, but I'm not modeling that!

    Why the ports drilled closer than modeled? That's because the actual critical distance for throat reflection is to the tweeter's diaphram, not it's exit. So for the reflection notch to model reasonably, you have to model with the aperture distance longer than you actually drill it.

    The cutting of the port frustrums is done in a 'tuning step... the simulation in Horn Response is just that, a ballpark approximation. You'd need much more powerful and ghastly slow software, and more than a small handful of number values to get precision of any degree!

    Hope that helps
    Bill
     
  22. Thanks, Bill
    I finally managed to 3D-model most of the Cosyne horn in order to see how things go together, prior to my build.
    I have but a question about S2 placement.
    According to your build of the Cosyne I, the distance from throat to mid S2 should be 1.175" (or 29,84mm) as per below pic.
    But, from my illustrations further below, this distance is a bit short compared to simulations, when you use a plane cross-section.
    However, conical horns supposedly propagates sfaerical waves, so in such respect the port positions are probably quite correct (wavefront curvature guesswork, though).
    I remember reading that you had concerns regarding these ports position when finishing Cosyne II.
    Was this the issue ?
    Would you recommend me to use the 1.175" position, or should I move the ports those 6mm (38 - 32) out ?

    Port dims.jpg

    I guess you mean this distance on the outside (blue line below):

    Distance S2 outside.jpg

    From the inside, this is looking like this (blue line below):

    Distance S2 inside.jpg

    The cross-section at the center of the ports on the inside is like this (magenta rectangle below):
    This area is 45 cm2, and thus not the 55 cm2 simulated.

    S2 cross section at L =1,175 inch.jpg

    The distance from throat at this point is 32mm (blue line below), and thus not the 38mm simulated.

    Distance L1.jpg
     

    Attached Files:

  23. Can the 4 ohm version of the Aura N6-255 woofer be used fo this project?
     

Share This Page