So, I'd like to learn more about determining T/S parameters for a candidate driver design prior to actually building a prototype. Can anyone point me to reference materials?
I am trying to assess some novel driver concepts. I have access to and decades of expertise using heavy-duty engineering software (including advanced structural and electromagnetic FEA). What I lack is experience in driver testing, which is obviously the final arbiter of quality, but if I can iterate a design numerically, it might help save time and money.
Any papers or links would be appreciated. Thanks in advance . . .
I am trying to assess some novel driver concepts. I have access to and decades of expertise using heavy-duty engineering software (including advanced structural and electromagnetic FEA). What I lack is experience in driver testing, which is obviously the final arbiter of quality, but if I can iterate a design numerically, it might help save time and money.
Any papers or links would be appreciated. Thanks in advance . . .
Very good information, and I will dig in deep, but not related to my original question. I am asking about using analytical techniques to predict driver performance PRIOR to actually building and testing said driver.
Reverse SpeaD:
Reverse SpeaD - Loud Speaker Designer Software - Reverse Engineer Any Speaker
No idea of the current pricing, previously it was out of the reach of any non-commercial users.
Interesting. The devices I'm considering do not fall within any "conventional" electrodynamic speaker concept I am aware of. Canned software solutions are unlikely to suffice.
I'm really looking for journal articles or other references that discuss methods of analysis to determine T/S parameters. Calculations of compliance, motor force, etc. and how to fold those into T/S parameters. Again, finite element analysis (FEA) is not only possible but probably required.
I'm really looking for journal articles or other references that discuss methods of analysis to determine T/S parameters. Calculations of compliance, motor force, etc. and how to fold those into T/S parameters. Again, finite element analysis (FEA) is not only possible but probably required.
The T/S parameters are the design parameters for a particular simple lumped/0D model of a conventional cone driver. This model is given in many places like the original T/S papers. The model is basic and has been improved by, for example, Scan-Speak who give parameters for an improved model on their spec sheets. The T/S parameters are not applicable to different types of transducers like, for example, electrostatic ones which require a different model.
If you are unable to write down by inspection a lumped/0D model for your transducer (or it is distributed requiring a 1D, 2D or 3D model to be reasonable) and don't have access to an engineer able to do this for you then your best approach might be to perform detailed 3D simulations in order to quantitatively understand what is going on. Such simulations can be very informative in quantifying the relative importance of the various physical processes present but you would need to be able to recognise them or be prepared to learn to recognise them which, depending on your background, may be a challenge without someone experienced to talk to and learn from.
It is hard to answer with precision without knowing what questions you want to answer, the type of transducer you wish to model, the audience for the model, etc...
Understood, and thanks.
As far as the drivers are concerned, consider them standard electrodynamic transducers, but designed for specific application requirements that preclude off the shelf solutions. The goal would be to iterate through multiple concepts before cutting metal.
The tools to simulate structural and electromagnetic behavior are not only available to me but are things I've used daily for decades. It's what I do (among other things). What I don't do is design loudspeaker drivers . . . until now. Most of the work I've found in determining classic T/S parameters uses experimental techniques. I'm simply trying to find alternative resources that might shine a light on how others have iterated through detailed designs to avoid building and testing prototypes until design confidence is reasonably high.
I understand that T/S is simply a model, and a small-signal model at that. I'd like to start there and then move on to more complete models, such as the Scan-Speak variation you alluded to.
It wouldn't shock me if the type of information I'm looking for is largely kept under wraps by the major manufacturers. If that's the case, so be it. I'd still like to believe that there have been some attempts, perhaps in academia, that were published (JAES, dissertations, etc.).
As far as the drivers are concerned, consider them standard electrodynamic transducers, but designed for specific application requirements that preclude off the shelf solutions. The goal would be to iterate through multiple concepts before cutting metal.
The tools to simulate structural and electromagnetic behavior are not only available to me but are things I've used daily for decades. It's what I do (among other things). What I don't do is design loudspeaker drivers . . . until now. Most of the work I've found in determining classic T/S parameters uses experimental techniques. I'm simply trying to find alternative resources that might shine a light on how others have iterated through detailed designs to avoid building and testing prototypes until design confidence is reasonably high.
I understand that T/S is simply a model, and a small-signal model at that. I'd like to start there and then move on to more complete models, such as the Scan-Speak variation you alluded to.
It wouldn't shock me if the type of information I'm looking for is largely kept under wraps by the major manufacturers. If that's the case, so be it. I'd still like to believe that there have been some attempts, perhaps in academia, that were published (JAES, dissertations, etc.).
Scottmoose, did you reply but delete? My email shows a nice detailed reply but I'm not seeing it on this thread.
A lumped/0D model expressed as an equivalent circuit can be analysed by a wide range of software such as the many SPICE derivatives. If you want to use this approach then background for lumped transducer models can be found in standard acoustic text books (Kinsler&, Olson, Beranek, Morse&, etc...) from the 40s and 50s. I would be careful of revisions past about the 70s or modern equivalents because some of the loudspeaker examples are likely to be revised away.
If I was designing a novel transducer rather than a DIY speaker I doubt I would find much use for such lumped models because they cannot simulate or provide insight into the details of the transducer. The T/S parameters are primarily intended to be used by those designing speakers not transducers.
To design a transducer I would primarily develop 3D models to learn the quantitative balance of the physical processes involved and to test design ideas. This would need to be supported by some experiments to determine some real world values of various properties like damping which is tricky to model accurately but is likely to play a significant role in the quality of the transducer.
Any significant academic research into speaker drivers ceased many decades ago. The research and development performed by commercial speaker companies will be kept in house if it is considered a commercial advantage but the design and manufacture is sufficiently widespread, low tech and outsourced to third parties this is unlikely to apply. There is the odd exception like the MBL radialstrahler which would require significant R&D effort by a manufacturer to reproduce but given the very tiny demand this would be unlikely to be commercially worthwhile.
Last edited: