BassBox 6 Pro Analysis (RMS or PEAK?)

[SubDevo]

Ok everyone here we go...

My question to harristech:

Hello,I have a pre-purchase question. I was wondering if graphs of excursion and port velocity, reflect RMS or Peak values. In BBP help on cone excursion page, it states peak excursion, but results seems to indicate RMS when compared to other softwares. (I have read conflicting responses as well, on the web). Please clear this up for me.

The response from harristech:

Both the Cone Displacement graph and the Vent Air Velocity graph are based on the peak diaphragm displacement assuming a constant input voltage that the user can control. Perhaps the confusion is due to the fact that both of these graphs default to using the max power rating (Pe) of the driver. We normally recommend that designers take a conservative approach and enter an RMS steady-state maximum power value for the Pe parameter. If they do, then the default voltage used for the Cone Displacement and Vent Air Velocity graphs will be based on the peak excursion at the rated RMS max power of the driver. But the user can enter their own input power or input voltage for these graphs. Internally, the program uses input voltage so an input power level will be converted to voltage based on the nominal impedance of the driver.
Best regards,

Harris Tech Support
[e-mail]

Ok, I don't accept this. So I did some testing on my own...

Software used for testing:
UniBox
Subwoofer Simulator
LSD (Loudspeaker System Design - my own software - unreleased)
WinISD Pro
BassBox 6 Pro

These are the steps taken to help ensure these comparisons are valid.
All softwares used in this test were set to the following parameters:

Speed of Sound = 343.68 m/s
Air Density = 1.2009 kg/m^3
Closed Box (least amount of variables)
Unlined, All Losses set to 10,000 (infinite/lossless)
Qtc = .707
0 ohms Series Resistance
Le NOT Modeled
500 watts input
   NOTE: Subwoofer Simulator only accepts Voltage.
   Equation used: Vin = sqr(Pe*Re) = 40.447 Volts


These are measured parameters, not manufacturer specs. I own this Driver.
Driver: DAYTON Titanic MKIII (TIT320C-4)
Fs=26.433 Hz
Vas=57.443 L
Qes=0.5448636
Qms=6.6803
Qts=0.5037743
Re=3.2719 ohm
Pe=500W (RMS)
Le=2.5102 mH (NOT USED IN TESTS)
Xmax=18.7 mm
Diam=23.857 cm
Sd= 447.0145 cm^2
Bl=13.3664
Mms=179.1391 grams
Rms=4.453698 kg/s
Cms=202.3751 uM/N

Results: All values at 10Hz


Click link below to view an html page of more images from test.
lsdp.100webspace.net/test5.html
(The last image on this page is an animated gif, with the PEAK excursion graphs scaled and overlayed. )

Observations:
You need to double the power input for BB6P to bring it inline with the other simulations.
And even then, excursion is 1.85mm greater than PEAK avg @ 500w. 1.31mm greater than RMS avg @ 250w. IMO, something strange is going on in BB6P.
I can't account for these small differences. (rounding errors?)

EDIT:
I figured out the small differences are mainly attributable to the way BB6P converts Power to equivalent Voltage. BB6P uses Znom instead of Re. So Vin = sqr(Pe*Znom).
BB6P computed Znom =   3.92628 ohms (1.2*Re) for this driver.
In SubSim @ 250watts this becomes 31.33V. SubSim excursion is then 18.732mm. BB6P excursion is 18.3mm. The .432 difference is then probably due to rounding.

BB6P only shows excursion limiting with 1000w input, leading me to believe that BB6P is assuming input power is PEAK not RMS. With this assumption, BB6P graphs are inline with PEAK values (and correct excursion limiting), ONLY if input power is PEAK (2*Pe_RMS).

Conclusion:
BB6P assumes input power is PEAK (2*Pe_RMS) then excursion graphs are realistic(PEAK), and results are similar to other simulations. If WinISD is changed to RMS, you will note that Xmax line changes and the project is still excursion limited. This is correct. WinISD assumes input power is RMS. Now, this excursion graph is similar to BB6P, but in BB6P it does not show excursion limiting!
The other 4 simulation softwares assume input power is RMS and scales graphs correctly as PEAK.

This really bothers me.
If harristech specifies that the graphs are PEAK, then why are they so different? I believe BB6P is giving a user a false impression that a driver will not exceed xmax in their simulations. If using BB6P, you MUST enter PEAK input power (2*Pe_RMS) in order for the results to be relevant.

Hopefully, you will be able to understand my post! ( I did repeat myself a few times)
If anything is confusing or unclear, feel free to ask!
I welcome any and all comments regarding this test.

Regards,
Michael

[SubDevo]

Well guys,
I'm impressed! harristech actually read this post and emailed me a response!

We're sorry for the delay in responding---we're in the middle of upgrading to new security software across our network and it has slowed our response time until the upgrade is complete. We read your post (if your user name on Linearteam's forum is "SubDevo") as well as your link to your images. The first thing we observed was that you did not specify the "Input Voltage" directly in the BassBox Pro simulation and this produced an "apples to oranges" comparison. As we explained earlier, our program actually uses voltage for the Cone Displacement graph simulation so it must convert the power to volts and it does this based on the driver's net nominal impedance (Z). What you should have done was click on the "Input Power" label to the left of the input box in the upper right of the graph and the input box would have changed to "Input Voltage". Then you could have entered the voltage as you did for your other software and we believe your questions would have been answered. By the way, you can force BassBox Pro to always accept "Input Voltage" with the Preferences window of the program. With an input voltage of 40.45 V, BassBox Pro will display a max cone displacement of 16.4 mm at 10 Hz. From this it seems clear that BassBox Pro is displaying an RMS excursion and we'll have to correct the typo in our manuals which states that it is a peak excursion. As for the peak excursion, repeat the test with an input voltage of 57.21 V (which is 40.45 / 0.707) and BassBox Pro will display a max cone displacement of 23.2 mm at 10 Hz. This is reasonably closed to the results produced by the other programs and we'd have to examine the underlying assumptions of the equations of each program to identify why they vary by as much as 1 mm. When you entered an "Input Power" of 500 watts, BassBox Pro used Z = 3.926 ohms to calculate the input voltage to be 44.31 V ( 500 / sqr( 500 / 3.926 ) and this is why you got an excursion close to 18 mm. Note: We had to examine your images to discover that you used a Z of 3.926 ohms because you did not list its value in your post. Normally a value of 4 ohms would have been used for this driver. We hope that this answers your question. We apologize for contributing to the earlier confusion.
Best regards,

Harris Tech Support
[e-mail]

Note this line...
"From this it seems clear that BassBox Pro is displaying an RMS excursion and we'll have to correct the typo in our manuals which states that it is a peak excursion."

I did overlook the fact, that in BB6P, you can change to voltage input.
Even though this doesn't change the results of this test.
For the sake of fairness, I changed Znom to equal Re in BB6P. This makes the internal voltage the same as the other test programs.
With Znom = Re, excursion is:
24mm @ 1000w (exactly equals UniBox Peak)
16.9mm @ 500w (exactly equals UniBox RMS)
12mm @ 250w
Thus it is still verified that BB6P results are RMS.

To the HarrisTech team, if you read this, I truly appreciate the personal and quick support you have provided. This is rare in this day and age!

Regards,
Micheal

[bubba]

Hi Michael, That was a good job catching that glitch in BBP. Anyhow, I have Harristech's X-over 3 Pro. When I had purchased it some time back, I thought it was going to give me the results of designing good passive crossovers as Christian Ouellet from Solen uses this program for the designs on his website. He told me that there were a couple of errors with the program, but he would tell me where they were and what to do about to get the results as he was getting. Back then, I wasn't aware that this program is not capable of achieving good results, but rather may be used to configure pro-audio speakers and I also say that with some doubt. I had recently advanced my capabilities for designing loudspeakers by purchasing the SoundEasy v15 program. I am at the point of getting my microphone calibrated and ready to proceed into the rest of the journey with SE. Anyhow, I just thought I would compliment you on the BBP glitch.

BUBBA

[SubDevo]

Hi Bubba,
Thanks for the kudos!
I've always known BB6P was incorrect, but it finally got to me! Had to settle it once and for all! lol

You sound pretty set in the software dept... soundeasy? hmm what need do you have for my software or WinISD? lol
Soundeasy, from what I hear, has a pretty steep learning curve to it... you have some reading to do!

Take care,
Michael