Loudspeaker design using Open Office Calc (or Excel)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Member
Joined 2008
Paid Member
After researching the free speaker design software available, I am trying to make some OB modeling in Calc/Excel. It is very easy to digitize FR and impedance plots shown in speaker datasheets. It is quite easy to compute amplitude response of a given type crossover (have not tried phase yet - should not be much of a problem, too). There are baffle influence simulators, whose output can be easily imported to Calc/Excel too.

There is only one big problem - I do not know how to "graphically" add FRs in Calc/Excel. The problem is, each data range has different number of points, so I cannot add easily y-coordinates for a certain x-coordinate, as the x-coordinates are different in each data range. Is there an easy way of doing it?

What is the formula behind XLBaffle - it should not be difficult to implement in Calc/Excel - where could I find the theory?

Such a spreadsheet would be very nice for crossover design from the real measured response shown in speaker datasheets - I guess I could get much closer to the final result than with textbook formulas.
 
Member
Joined 2008
Paid Member
I forgot to mention a few things - so that the post is also useful for someone else:)

I capture the plots to the clipboard directly from the screen with MW Snap
http://www.mirekw.com/winfreeware/mwsnap.html

Then I paste from the clopboard to Engauge Digitizer for FR and impedance plot digitizing - it is as easy as it gets (and free:) ) Do not forget to use log scale on X - http://digitizer.sourceforge.net/

I get exported plot as a series of numbers - that is easily imported to Calc/Excel - two columns separated by comma.

Crossovers are easily computed from circuit theory using ideal components - needs a little math, though.

It is easy to include any kind of compensation circuit - LR transform, dipole compensation, etc.

I yet have to find some theory for baffle diffraction and ultimately, some LF room influence should not be that difficult to compute either. It should be also easy to add simple sealed and bass reflex box and driver modeling with TS parameters.

Most of these things can be done by Basta! - but I'd like to have a little more flexibility event at the expense of more work. And what is most important - learn the math behind traditional loudspeaker design.

I will be glad to share my spreadsheet when I come to something useful.
Right now it is just killing time during a business trip
:D
 
Member
Joined 2008
Paid Member
Thanks for the suggestion - I will check it out. The main purpose is of course not reinventing, but learning and understanding:)

update: Through unibox I have found what I was looking for - FRD's Frequency response combiner:) Now I can combine frequency plots till eternity:D
 
Hi,

I use winISD pro freeware professional software. Not fot exel or openoffice.

To design:

closedbox
bassreflex
closedbandpass
ventedbandpass

winisd-pro
 

Attachments

  • untitled.jpg
    untitled.jpg
    83.5 KB · Views: 349
Member
Joined 2008
Paid Member
WinISD is OK, I use it from time to time - but is does not allow to use real measured data - that is why I needed something else.

I have first results - SPL Tracer and SPL viewer do not work on my PC. Tracer behaves strangely after selecting "trace" and does not trace anything, viewer crashes with an error message while trying to load a response to view.

FRC kind of works:) I am just playing with it.
 
Member
Joined 2008
Paid Member
This FR Combiner makes me crazy! I am trying to expand the FR, but it always moves beginning to 2 Hz and resamples - and that totally messes up the FR curve - what am I doing wrong? I just select "extend top value" and "extend low slope" - and when I press resample, it does not extend anything, it just moves as I described above.
 
Member
Joined 2008
Paid Member
Even though my programming skills are a little limited, it makes me thinking about making some small program for just resampling the curves. Any extension would have to be done by hand - actually I may try to extend the curves I have by hand in Calc - should be very easy.
 
Member
Joined 2008
Paid Member
Frequency combiner just does not work in Microsoft Office Excel 2003 SP3 (Czech language version). I tried to extend the FR plots manually - but it still does neither resample nor add two FR plots correctly. I guess its google time:)

Edit: No suitable freeware found yet. If anyone knows some nice plot resampling software, please let me know. Command line interface is not an obstacle:)
 
Member
Joined 2008
Paid Member
Yes, that looks promising. One thing is, that I do not like to buy "a rabbit in a sack" as we say here. The A B C dipole is actually very close to what I am trying to achieve. I will have a look at the demo.

And I have finally found the thing - Octave (a free MATLAB lookalike). That should provide me with all data processing capacity I will ever need:) There is a similar software - SciLab - which has a component really close to SIMULINK - then a complete model could be made using blocks. Nice.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.