How accurate is WINisd?

jakelm

I just recently bought a Dayton 12"DVC from a friend in our forums. I found the T\s specs for the sub and punched in the numbers in WINisd to build a new box. From what the graph shows me, a ported box about 3.0 squft would be the best for this sub. But the graph also shows a 6-7dbs drop around 28hz. How accurate is this graph? Is the slope down, that dramatic?

Jake

tryrrthg

I would think it's pretty accurate, otherwise people wouldn't use it.

I use WinISD Pro for finding a ballpark alignment that I am happy with. Then I use Unibox (Excel program) to fine tune it. From what I read at the DIY forums most people think Unibox is a better tool.

jakelm

Thanks for the help. l'll try Unibox and see if I get the same results. I was just kind of hoping the sub and box would perform alittle better.


Jake

tryrrthg

jakelm wrote:

Thanks for the help. l'll try Unibox and see if I get the same results. I was just kind of hoping the sub and box would perform alittle better.
Jake

You don't have to use the alignment that the programs automatically give you. You can tailor it to how you want it to look by changing the box volume and port size and length. Just keep an eye on cone exursion and port velocity. It's all about tweaking!

tryrrthg

In the thread that madmax sold these, Mazeroth says they're pretty much identical to the Adire Shiva drivers. if that is the case just follow one of the attached plans for the Shiva.

I'm having trouble getting both files attached. the other file (for vented boxes) can be found here:

http://www.polkaudio.com/forums/showpost.php?p=370250&postcount=3

jakelm

I've done all the tweeking that can be done with WINisd, its the slope after 30hz that gets me. It shows with no matter what port or box size I use, after 30hz to 20hz there is a huge db loss. Now I dont expect a flat responce down to 20hz but not that much. Here is a pic, I'm at 4.00 cuft box, a 4" dia port 19" in length at 19.3 tuning. What do yall think?

jakelm

Here is another with 4.2 cuft with tuning at 21.7hz a 4" dia port 14" long. Does this look about right?

tryrrthg

Looks fine to me. it's only down 5db's at tuning point, unless your room is huge, room gain will more than compensate for that loss in SPL.

if you want a bit more low end, increase the box size to 5 cubic feet and a 20hz tune. This alignment will be more like the EBS alignment in the Shiva plans.

jakelm

Ouch!! 5 cuft? Thats a biggy. My room is 16w x 22L x 9H. I plan on building 2 subs, the other sub will be the same driver just not the same size. The second box will only be about 3 cuft (WAF)

Jake

jakelm

Ok here is 5 cuft with 20hz tuning. But I must say 5cu ft might be too big for the area where the sub will be placed (in the corner)

jakelm

Oh by the way, it was either the Dayton or the AscenadntAudio Assasin sub. I'm glad I went with the Dayton. Punching in the parameters of the Assasin sub the responce was much worse.

Jake

tryrrthg

Your room is similar in size to mine. Have you considered putting them both in the same box and building a sealed sub? I built a sealed Atlas 15 and room gain more than made up for my low end response (see link below).

http://www.angelfire.com/alt/tryrrthg/atlas15/index.album?i=24

The lesson here is your 4 cubic ft box will be fine. Room gain is your friend and will make up for the loss in SPL. Plot your driver as a sealed sub and see how fast it rolls off, that's pretty much what my design looked like on paper, but in the real world it looks like the graph in my link.

jakelm

I did plot is in a sealed design. Much worse. The problem with 2 in one box is that I have a very bad null in one of the other seating positions. I compisate that null with 2 different boxes.

Thanks for you opinion, I kind of figured room responce would be better than what the graph shows. But I just needed to hear it from another Polkie.:D

I am going to drive this with my Hafler DH-200, I dont know if this will make a difference, I hope not.

Thanks,

Jake

bknauss

I never had very good results with WinISD when I was using it a lot several years ago. I would compare the results with other modeling programs, and while the other programs were fairly close, WinISD was usually pretty far off. Had good results with Unibox though.

jcaut

I use WinISD beta for "quick-n-dirty" box calculations, then if I get more serious, I plug the info into Unibox. They're usually pretty close, but occasionally I'll find a substantial variation- I don't know why. Unibox requires that you input more information so perhaps some of those "assumptions" that WinISD makes have a greater effect on some alignments than others. I always verify the box tuning by running impedence curves, and while WinISD is close, Unibox is always very close.

The best thing about WinISD is that you can click on the "box" tab and then move the cursor up and down on the picture of the box to "draw" the curve you like. That, and being able to click on the units to convert to/from metric. Neat.