I'm not sure if the math is correct. If you guys could test the page out with subs you're familiar with and post your thoughts, it would be really helpful. As soon as I fix the sealed version, I'll start ported and bandpass versions also. It's written in PHP.
Sealed box design
Sealed box design
It's kind of neat.
The couple drivers I tried seemed to work out right. Do you plan on dipping into the ported designs?
The couple drivers I tried seemed to work out right. Do you plan on dipping into the ported designs?Yes, I plan on adding a ported and bandpass plotter also. To me it seems like the low end rolls off a bit too high, no?
Well, I compared it to Lspcad Pro, BoxPlot 3.0, two different spreadsheets, one using Bill McFadden's HP calculator program and the other using the Margolis-Small HP calculator program, a proprietary program of mine, and MJK's sealed Mathcad worksheet. All plot slightly different responses for a given Vb due to how much the formulas have been simplified or values rounded off, etc., but while yours calcs the same Qtc (at least once yours is rounded off to the first three decimal places anyway), all plot a considerably lower F3/Fb than yours does.
FWIW, listing Qtc to three decimal places is sufficient and I prefer to specify Qtc and let it calc the Vb.
HTH,
GM
FWIW, listing Qtc to three decimal places is sufficient and I prefer to specify Qtc and let it calc the Vb.
HTH,
GM
You can find the proper equations at:
www.diysubwoofers.org
No need to campare to programs...off the top of my head:
Qtc=Qts*sqrt(Vas/Vb+1)
Fc=Fs*sqrt(Vas/Vb+1)
Fn=F/Fs
dB=20*log10(Fn^2/(Fn^2+Fn/Qtc+1))
At Fc, dB should be 20*log(Qtc)
Yours seems to be off by about 3 in magnitude, not sure how you are doing the graph, etc... perhaps your equations are right and you just have a viewing transformation - scaling issue, easy to do when making raster plots.... +/- 6dB is too tight, perhaps do +6 and -24, also go down to 10Hz. THese models certainly aren't valid above 500Hz for woofers....
Hint, Sealed box is the same as vented, you just use Fb=0 or Qp=0, depending on the model used......
...oops, forgot a parentheses....
www.diysubwoofers.org
No need to campare to programs...off the top of my head:
Qtc=Qts*sqrt(Vas/Vb+1)
Fc=Fs*sqrt(Vas/Vb+1)
Fn=F/Fs
dB=20*log10(Fn^2/(Fn^2+Fn/Qtc+1))
At Fc, dB should be 20*log(Qtc)
Yours seems to be off by about 3 in magnitude, not sure how you are doing the graph, etc... perhaps your equations are right and you just have a viewing transformation - scaling issue, easy to do when making raster plots.... +/- 6dB is too tight, perhaps do +6 and -24, also go down to 10Hz. THese models certainly aren't valid above 500Hz for woofers....
Hint, Sealed box is the same as vented, you just use Fb=0 or Qp=0, depending on the model used......
...oops, forgot a parentheses....
Yes, I did have the proper equation. The problem was that PHP assumes that a logarithm with no specified value has a base of e instead of 10. Once I forced the 10 base, it seemed to work properly, try it out.
Thanks for the feedback, I am working on adding these options. Please try the program now that it is fixed and let me know if there's anything else you'd like to see added, maybe a user defined dB range etc.
Thanks for the feedback, I am working on adding these options. Please try the program now that it is fixed and let me know if there's anything else you'd like to see added, maybe a user defined dB range etc.
Solid Snake said:
Others have pointed out that the plots don't look to be correct. Other than that, you need to go to about -15dB on the low side, for folks who will use EQ.
Added:
-Colored 0dB and F3 lines
-Fixed Qtc to 3 decimal places
-Increased range: -24dB to +12dB
Working on Qtc input
-Colored 0dB and F3 lines
-Fixed Qtc to 3 decimal places
-Increased range: -24dB to +12dB
Working on Qtc input
Now that I'm home and have been able to compare it to BB-Pro, it appears there are no accommodations for Ql and box filling. If I set Ql = 1000 and no filling, your program matches very close to BB-Pro. Extension to 10Hz would be nice as well.
The low cutoff is now 10hz. I'm not really sure how to simulate any stuffing, I'm sure it's a very gray area.
- Status
- Not open for further replies.