advertisement


Speaker design software

C

Cesare

pfm Member
Having recently got a pair of Tannoy Gold 15 inch drivers (actually HPDs with the cones replaced with gold style ones) the next step is building a set of cabinets for them.

What i've realised, looking at the various online resources for the original Tannoy cabinet designs is that the traditional ones pre-date the Thiele/Small theory papers, and hence are really designed and tuned by ear, or by outmoded analysis techniques.

So, i'm somewhat torn between building an original design (which would be fun, i'd probably build a pair of Ardens, that sort of size and design) or tweaking the design to something original based on more modern analysis. Of course in an ideal world i'd try both, but i'm not sure the missus would appreciate two sets of 100+ litre cabinets in the front room :)

So that brings me to what software to use to get some sensible predictions for a given design? Is there a package of choice that people are using for such analysis?

I was wanting to at least get some predictions of the behaviour of the designs I can find online. I'm guessing that some of the later studio tannoy designs might also be worth analysing to see how they behave (in theory).
 
I quite like Unibox, an excel based spreadsheet, lots of box design options and easy to use.

WinISD is basic but easy to use while VituixCAD is far more powerful but much steeper learning curve.

For the best results, best you measure your drivers (thiele small parameters) you actually have, with a DATS system (or DIY/REW etc) to get an accurate box design.
 
Yea Unibox is good also try the Visaton free program it does the baffle as well.
 
Hoopsontoast said:
I quite like Unibox, an excel based spreadsheet, lots of box design options and easy to use.

WinISD is basic but easy to use while VituixCAD is far more powerful but much steeper learning curve.

For the best results, best you measure your drivers (thiele small parameters) you actually have, with a DATS system (or DIY/REW etc) to get an accurate box design.
Click to expand...

VirtuixCAD looks like an interesting choice, and one that I hadn't come across. I'll download and have a play.

My experience in cabinet design is limited to building some bandpass subwoofer designs in the early 90s (is bullock on boxes still in print?), but generally i've been more a consumer of pre-built stuff than a maker. I think though it will be interesting to plumb existing Tannoy designs into such software and get a better understanding of the tuning that was applied and whether it would work for me. The big problem with these larger tannoy designs is that the volume goes up quickly, so it's I think worth understanding the benefits from the chunk of the front room they would occupy vs more manageable cabinets.
 
Cesare said:
What i've realised, looking at the various online resources for the original Tannoy cabinet designs is that the traditional ones pre-date the Thiele/Small theory papers, and hence are really designed and tuned by ear, or by outmoded analysis techniques.
Click to expand...

The Thiele/Small equations are a design method for non-technical engineers and technicians. The equations themselves were in use by technical engineers for decades prior. If you look at an acoustic text book from the 50s (and possibly 40s) you will see the same equivalent circuits to model speakers written down by inspection. I would expect one or two people in Tannoy were perfectly capable of writing down and analysing these equivalent circuits prior to the Thiele/Small papers.

Cesare said:
So, i'm somewhat torn between building an original design (which would be fun, i'd probably build a pair of Ardens, that sort of size and design) or tweaking the design to something original based on more modern analysis. Of course in an ideal world i'd try both, but i'm not sure the missus would appreciate two sets of 100+ litre cabinets in the front room :)

So that brings me to what software to use to get some sensible predictions for a given design? Is there a package of choice that people are using for such analysis?
Click to expand...

I use 3D FEM to simulate the vibration of the cabinet and BEM to simulate the sound radiation from the cabinet. I started with open source packages such as code_aster, acousto, salome, paraview but have recently started to develop bits of my own software in order to more efficiently simulate viscoelastic damping. Not wholly required but it is a hobby interest. In order to simulate the sound radiation, motion and forces from the drivers I wrote down a lumped model by inspection which will be close to the Thiele-Small model.

Open source solid modelling, FEM and BEM software is freely available if you are interested. I can provide one or two pointers. Commercial equivalents would be preferable if you can access them via work but they are usually too expensive to consider purchasing to support a hobby interest.

Cesare said:
I was wanting to at least get some predictions of the behaviour of the designs I can find online. I'm guessing that some of the later studio tannoy designs might also be worth analysing to see how they behave (in theory).
Click to expand...

In order to do this you will need the values of the driver parameters used by the Thiele/Small design equations. Are you able to measure them?
 
There are essentially two approaches to using CAD software. The basic approach simply relies on published or measured T-S parameters, and is really useful only for modelling box and port dimensions to optimise bass loading. The resulting filters for the drivers won't take into account the drivers' native frequency responses.

The more accurate method is to obtain measurements of the drivers (mounted on the intended baffle). The two essential measurements are frequency response and impedance. These measurements, if correctly obtained, will be instrumental in the design of filters to integrate the drivers seamlessly. I used LspCAD, which comes bundled with JustMLS to measure FR and Z.
 
Another vote for VirtuixCAD. It's is highly powerful and free. It takes an iterative approach so you can start with some manufacturers data for TS parameters and also trace the manufacturer measurement etc. As the design progresses you can add in you own data and measurements.
 
You must log in or register to reply here.


advertisement


Back
Top