1. Please take 30 seconds to register your free account to remove most ads, post topics, make friends, earn reward points at our store, and more!  
    TalkBass.com has been uniting the low end since 1998.  Join us! :)

Somewhat Puzzling WinISD Results

Discussion in 'Amps and Cabs [BG]' started by mrjim123, Jan 15, 2012.

  1. mrjim123

    mrjim123 Supporting Member

    May 17, 2008
    Elkhart, IN

    I am going to rebuild an Avatar B210 Neo cabinet in order to reduce weight. I'll use the existing 10" drivers, the tweeter and crossover will be eliminated, and thinner plywood with dowels for bracing will be used, and the shelf will be replaced with round ports for (hopefully) further weight reduction.

    I obtained driver data from Eminence and modeled the design in WinISD Pro ALPHA, version #0.50alpha7. WinISD did not have the particular driver's data in its database, so I entered it manually. I followed the steps outlined in an online tutorial (
    http://www.hometheatershack.com/forums/diy-subwoofers-general-discussion/6330-winisd-pro-tutorial-download-detailed-guide-how-use-winisd-pro.html) - twice, just to be sure I didn't screw something up the first time. Note that the volume I entered, 2.9 cu. ft., is virtually the same as the existing Avatar box. I got the same WinISD results both times and am somewhat baffled by a few things. I'm not sure if I'm doing something wrong or simply misinterpreting the results. So, here is the data, followed by some questions.

    This is the manufacturer's driver data sheet:


    WinISD Driver Editor results:


    More WinISD modeling data (Filter/EQ data not shown - I left it blank):



    1. The Maximum SPL chart only goes to 104 on the vertical axis, but the calculated SPL is off the chart - above 104 at about 35 Hz. Is this typical? Is there any way to increase the range of the vertical axis?

    2. The Maximum Power chart is a flat line at 450W, but it dips to 140W at 70 Hz. Why is the dip so far above the tuning frequency? And isn't 140W quite low, given the driver's rated power handling? Does the data represent only one driver or both drivers in the box? Note that I didn't specifically find "Pe" on the driver data sheet so I entered 225, listed as "Pmax" on the data sheet.

    3. The Transfer magnitude chart shows a +3 dB hump at 80 Hz that doesn't fall below +1 dB until about 200 Hz. At about the tuning frequency (45 Hz) it is -3 dB. I understand that such a hump will cause the box to sound boomy, which surprises me a little because I would not describe the sound of the existing Avatar box this way at all - to me it sounds warm, with a nice fat bottom. Maybe this is just another way of describing "boomy". It seems that whatever I do to reduce the hump always results in a ridiculously low tuning frequency, 20 Hz, for example. Am I just too hung up on maintaining a tuning frequency of about 45 Hz?
  2. will33


    May 22, 2006
    1. Doubleclick (or right click?) on an open area of the graph and you can change spl range, bandwidh, etc. Set the vertical axis to show up to about 130db and your line will stay on the chart.

    2. The powerhandling dip at 70hz is normal. Driver excursion is at it's greatest just above the tuning frequency. It calms down at the tuning frequency as most of the output there is directly from the port. Then shoots up below the tuning frequency----"unloading". This is why we say power ratings are thermal and real world power handling might be half that.

    3. A 3db bump is livable. It'll help fill in the bottom of your 210. Get much higher than that and it can start causing boomy sound in already bad sounding rooms and start to lose a sense of pitch to your notes. Raising the tuning frequency makes a bigger bump, lowering it makes a smaller bump. These adjustments will also affect power handling in the lows.

    All in all, it sounds like pretty normal results and the new box should sound pretty much like the old one did.
  3. mrjim123

    mrjim123 Supporting Member

    May 17, 2008
    Elkhart, IN
    Thanks will33. I edited the post after you responded and added this question: "Does the data represent only one driver or both drivers in the box?"

    So, I'm probably OK powering two of these 8 ohm boxes with an amp rated at about 500W RMS @ 4 ohms if the amp's volume is turned up to, say, 4 out of 1 to 10?
  4. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    The effect is normal, but the large drop in power handling is undesirable.

    If I were designing a speaker, I would want to get rid of it by finding a different combination of driver, box, and tuning that would not have it.

    Essentially, the driver excursion increases at lower frequencies because the "volume velocity" that corresponds to that power *in the air* is larger. The tuned box presents a high "acoustic impedance" to the speaker, while the port moves the air.

    If the tuning frequency is to low, the box will not "take over" the air movement smoothly, and you get a dip between the two.

    OK, some power handling ratings are un-solvable, the speaker size simply cannot move sufficient air to go as low as you want at the power you want and reach to the box tuning needed. Then you may need to add more speaker area.

    Box dimensions and tuning CAN be adjusted to change the "Q" of the tuned circuit and widen the port effect. But the best combination may not fit with a given speaker well. And the "optimum" combination is a compromise, as with almost everything, because the ideal "Q" value is usually better.

    If you have a big excursion-limited power handling dip, the speaker will probably "fart" badly on at least one note.
  5. will33


    May 22, 2006
    All good points...but...the OP is simply rebuilding an existing speaker (Avatar 210) into a lighter, more manageable box. Same volume and tuning.

    @OP--when you loaded up the model, if you selected 2 drivers, your results will be for both drivers together. If you only selected 1 driver....you're modeling a 110.
  6. billfitzmaurice

    billfitzmaurice Commercial User

    Sep 15, 2004
    New Hampshire
    Owner, Bill Fitzmaurice Loudspeaker Design
    Yes, in the Options window.

    Because driver excursion is at a minimum at Fb.
    It's actually above average.
    One, unless you specified two.
    The driver Qes/Qts is too high to get rid of the midbass hump. Personally I wouldn't use a driver with Qts higher than .5 in a ported cab.
  7. will33


    May 22, 2006
    To your power question. WinISD assumes a constant sine wave, which an instrument signal is not. There is a few milliseconds power spike when you strike the string, then much less power as the note rings out.

    Add to that, the controls on most amps are not even throughout the dial. I've had some where the entire useable range was from off to half way, above that was just clipping. 4 out of 10 does not mean 40% power at all.

    The content of your signal also places different demands on the drivers. Boosting bass increases excursion and powerhandling goes down, reducing bass make powerhandling go up. You can simulate this in the eq section of WinISD. Add a wide Q, +4db boost around 60hz to simulate a healthy bump of the bass knob and watch what excursion does.

    What all this means is that in the real world, your cab will seem to handle a bit more power than the graph indicates and as with anything, listen for sounds of distortion/stress to know where the limit is and don't rely on the numbers on your amp. A 500watt @ 4ohms amp will be able to make your cab give you all it is capable of...no need for more power.
  8. Jerrold Tiers

    Jerrold Tiers

    Nov 14, 2003
    St Louis
    I realize he is duplicating a design...... sort-of.

    If it is a good model correctly showing the original (not a "110" by accident) maybe we know something about that speaker now.....

    Or, as is possible, the change from a shelf port to round ports is making a difference. I don't know if that program can model the difference.

    Shelf ports are generally "lossier" because they have more friction, from air movement along the "sides". There is more "side" to the shelf port than if round, which is the lowest length of "side" for the same area. That friction tends to lower and widen the box resonance, and could "fill in" some of the power dip. Possibly the original design is better than modeled partly on account of that.

    10" speakers with 450W handling won't be able to do that to a low freq regardless, without really large excursion....which you might not like even if it were possible, you have to make a reasonable allowance. It's all about the right set of compromises.
  9. will33


    May 22, 2006

    WinISD is indeed not that great for calculating shelf ports. His actual tuning will be a bit lower than the model predicts. With tube/pipe ports, I find it's very accurate.

    Some correction math and a link to a more accurate online shelf port calculator in this thread.

Share This Page