Should WinISD be used for figuring out if one sub is better than another?

I use Unibox rather than WinISD as I find it easier.

However, the general answer is that using a good modelling program, correctly input data and then some idea of what you want from the design, then yes, in terms of frequency response and theoretical SPL you can tell which will perform better.

For LF, if I've entered the data correctly, I find Unibox gives close to measured for FR. Some (mainly commercial) designs I've modelled myself design looking for a flat FR using an estimated room gain or some other design tradeoff to give acceptable performance within a budget.

What these sort of programs can't tell you is which drivers have lower distortion or the distortion spectra or the power compression. For wideband speakers there are a host of other design considerations as well such as directivity, diffraction, baffle step etc.

I seldom give much credence to what others report in listening reports as I find they rarely agree with what I hear.

WinISD is the first tool I go to, after a brief glance at the T/S parameters. I always used to build sealed subwoofers (and full range speakers), because the math for vented units was daunting, and it was difficult to easily see the effect of different box sizes and vent frequencies. Since I started using WinISD, 90% of my designs have been vented, and they all came out just how I planned.

IF the T/S parameters are accurate, you can absolutely use WinISD to compare the merits of one driver against another. Beyond that, it gives you a huge insight as to the behavior of any given driver/box combination, because you can instantly see the effects of changing box size or tuning, on any number of important performance characteristics.

Sure, this type of software has been around for a while, but it's either been very expensive, or extremely limited compared to WinISD, in my opinion.

3dB advantage, across an entire range of 10-40Hz, is fairly significant. That means the more efficient system is giving you the same performance with 100 watts input as the other is using 200 watts input. That said, if you did as you suggest, and adjusted them to both have as flat a response as possible, you won't see a 3dB different from 10-40Hz, unless one of them isn't a subwoofer! You can see different -3dB points, and so below those, there will be disparities. Now, I'm talking strictly about the top graph in WinISD Pro, Transfer Function Magnitude. When you use different graphs such as SPL or Maximum power, you may see greater differences.

A higher output speaker won't necessarily have more distortion than a lower output speaker. If both are operating in the linear range of their motors, distortion should be similar. Now that doesn't take into account various technologies such as shorting rings around the pole piece, or more advanced distortion reduction motors, such as the XBL^2 from Adire Audio. A high output XBL^2 motor can and certainly will have considerably less distortion than a run-of-the mill woofer using conventional construction and only 1/3 the linear throw.

All else equal, distortion does go up with increasing excursion. But when it comes to woofers these days, things are rarely ever equal.

What winISD does not tell you is how the driver performs with increasing drive level. Parameters can shift 20% under high drive levels, inductance can vary wildly with coil position and amperage through the coil, suspension can be increasing stiffness / decreasing ( IE: poorly designed ), eddy currents can cause coil offset, shifting the magnetic center in the gap up and down.

Small signal analysis is just that. What really matters is what happens when you start throwing lots of power at something, and how it 'behaves'.

Cone area matters here, as for the same 'X' output level, the larger the surface area, the less excursion is needed.

WinISD does not tell you if a driver is "good" (for the most part). It tells you if a driver is suitable for your application given the T/S parameters.

Hi Eddie, responding to your PM in a minute too. First, do you have a determined size you need to stick within for the enclosure? That will give a lot more for me to help you compare your options. Other things you need to compare are not just the normalized response which just shows what the shape of the curve will be, but look at what happens with a given input power. Our woofers are significantly more efficient and have lower inductance, so they have a lot more output higher in frequency. That pushes down the relative levels at the lower frequencies. Let me know what enclosure sizes you're looking at and i can get you some curves.

John

Set 'compatiblility mode' on the properties link of your WinISD shortcut to "Win 98/Win ME".


Graphs like SPL, excursion, vent air speed, etc. won't show any results until you enter "system input power" on the project's "Signal" tab.

-Brent

Just want to make sure I'm looking at things here. Which one of these dimensions is the height? How much room on the face of the enclosure do you have for mounting the drivers? The AV12's require about 7" mounting depth and the AV15's require about 8". You'll want a good inch or so behind them to allow air to flow out the pole vent easily.

John

You must not be adding ports for each sub or something?

Length up to 168 inches??? Cool, you might just be interested in building a transmission line subwoofer, if you have that length available.

If I model a single driver in a given box (lets say sealed for arguments sake) and I model two of the same driver in a box 2x the size, I get the same response with 6dB more output capability.

Double the displaced volume of air =+ 6dB if the power to each drive unit is the same. The response should not change.

For ported boxes, the above remains true but you need to adjust the port length and area to suit the new Vb.

If you double the number of drivers, make sure to double the box volume and double the number of ports. That will keep tuning the same. Or you an double the cross sectional area of of the port in the simulation if you don't want to add more of them. Make sure you're doing the cross sectional area though, not just doubling diameter.

John