Made a Omni-directional speaker. Amazingly they sound good, especially as they where quick to make and cost me near on nothing.
However, upon embarking on testing them, I got sidelined trying to work out the TS parameters.
Following for example sound whsite dot net, I got stumped on measuring the Vas...
HELP!
Video link of it in action - www.facebook.com/SparrowLegsSpeakerDesign/videos/1146497162115909
However, upon embarking on testing them, I got sidelined trying to work out the TS parameters.
Following for example sound whsite dot net, I got stumped on measuring the Vas...
HELP!
Video link of it in action - www.facebook.com/SparrowLegsSpeakerDesign/videos/1146497162115909
Attachments
I'm not sure you can translate them to your "melon."
I have a copy of Dick Small's thesis, and his assumptions and math are based on piston transducers. In particular, the transformations from electrical to mechanical can involve the cone area with an implicit assumption the cone is moving as a piston which the melon does not as it is bending if I imagine correctly.
I suppose in a general sense there must be an equivalent stiffness/resistance to motion, and there must be some kind of moving mass, and there is some kind of magnetic force product. So probably you could calculate Thiele-Small parameters, but I have doubts that it would be meaningful. Or useful-what do you want the parameters for, since it is a free-air speaker?
What does the impedance curve of that thing look like? (and if you have one, how was it obtained? Because a dirty little T-S secret is that the parameters will vary depending on the measurement method and drive level).
I have a copy of Dick Small's thesis, and his assumptions and math are based on piston transducers. In particular, the transformations from electrical to mechanical can involve the cone area with an implicit assumption the cone is moving as a piston which the melon does not as it is bending if I imagine correctly.
I suppose in a general sense there must be an equivalent stiffness/resistance to motion, and there must be some kind of moving mass, and there is some kind of magnetic force product. So probably you could calculate Thiele-Small parameters, but I have doubts that it would be meaningful. Or useful-what do you want the parameters for, since it is a free-air speaker?
What does the impedance curve of that thing look like? (and if you have one, how was it obtained? Because a dirty little T-S secret is that the parameters will vary depending on the measurement method and drive level).
You are right about the "Melon" bending/flexing. Also, yes your right about it being an air free speaker, but at some point the gaps between the petals will be sealed.
At the moment, I'm in the process of perfecting the build process, though this will take time due to other commitments.
Regarding the impendance curve, I'll be going the most basic root, not using any sophisticated equipment... Pink noise and a 10 volt generator...
Patients is needed on my part.
At the moment, I'm in the process of perfecting the build process, though this will take time due to other commitments.
Regarding the impendance curve, I'll be going the most basic root, not using any sophisticated equipment... Pink noise and a 10 volt generator...
Patients is needed on my part.
It doesn't have T/S parameters since it isn't a cone driver. It has various parameters, of course, that characterize it.
Since you have a speaker, what you are able to measure are the parameters of speaker performance (FR, distortion, dispersion....), not the speaker model components like the T/S cone driver model.
B.
Since you have a speaker, what you are able to measure are the parameters of speaker performance (FR, distortion, dispersion....), not the speaker model components like the T/S cone driver model.
B.
You mean just to get an average impedance?Regarding the impendance curve, I'll be going the most basic root, not using any sophisticated equipment... Pink noise and a 10 volt generator...
I would think you could fine tone generation apps to get a more precise curve. You can't measure the parameters without a pretty fine impedance curve. The usual method used to be
generator --> large resistor --> speaker
to simulate a constant current source, but that is not really how most speakers are driven.
generator --> 0.1 ohm resistor --> speaker
is more realistic, but you need something sensitive enough to measure the voltage across the resistor well.
Though I can make a speaker, I'm still learning the electrical side. (God help us!!!
) Just for clarification, "0.1 ohm resistor" at 1w for 1m??Also I need to take into consideration that the transducers are wired up in parallel?
Though I can make a speaker, I'm still learning the electrical side. (God help us!!!
Also I need to take into consideration that the transducers are wired up in parallel?
for impedance measurements look at the rew site there is a guide how to.
I think you are overlooking how impedance measurements looks to some guys with no tools and not sure what a resistor is or where to buy one. You and I can improvise with paper clips and light bulbs (as resistors), but not all readers.
I've been encouraging use of REW and the laptop mic. No kidding. Gets the job started.
B.
Cheers Bentoronto,
Couple of things, I have a copy of REW, which I don't seem to have a problem with as yet. Bentoronto is right, I don't have much in the way of audio tools. Amps, computers etc may help but not perfect!
I know what a resistor is and its function, due to my location - French countryside, I cant just go to "Radioshack or Maplins" to collect the parts. Thus I recycle old components from existing electrical goods.
Turning theory into practice is my problem. Though the REW Impedance Measurement diagram looks simple, I'm struggling understand (for example), where a power amp comes into play? (And I don't wish to kill my laptop sound card!
)With a HUGE caveat that distortion is terribly hard to measure accurately, unless you are absolutely sure that your ENTIRE measuring chain has lower distortion (including noise floor) than the device your are going to measure. That can be hard to know for sure. I certainly would not assume a laptop microphone would have low enough distortion for that, nor the audio circuits. Now, 0.1% is -66 dB if I trust an online calculator and my memory. That would make you think all should be fine. But the distortion and noise in a laptop chain could have noise peaks (60 Hz and related hum maybe) which could exceed that. OR maybe I'm being over-cautious based on misusing a MLSSA system and then realizing you could not use its FFT function like that.
Frequency response and dispersion, sure. But does REW have a way to truly calibrate a laptop mic? Their own site sells or at least links microphones. I'm curious, because I haven't used it.
As for impedance, most cheap setups use a big resistor in series because it is easier, meaning the drive level is a very small constant current. It is better to use a small resistor in series (like transistor amps), but it means you need sensitive measuring equipment.
--> So let me restate my point as "different methods and different volume levels will give you different results."
To just see where the resonance frequency is and what is the minimum impedance, fine. But to calculate Thiele-Small parameters the different methods can yield significantly different numbers, and the small-resistor method is superior.
Right. You never know, until you try it.
REW shows just what's in the air as you sweep the fundamental. Clear as day and immensely informative.
You can see the mud level by doing nothing but sweeping with no output from the speaker.
Doesn't sound too hard.
(See next post)
B.
A delightful post.
Sticking with my point of view that it is easy to make rough but valuable measurements, impedance doesn't have to be correct absolutely since you mostly just want to know where it happens in the freq compass.
So, you put any size resistor in series (10 to 300 Ohms) and see where the speaker impedance rises across the speaker. Since you understand theory, you probably can calculate the absolute values anyway.* What I do is: replace the speaker with resistors of, say, 5, 20, 60 Ohms and run REW. That will provide precise curves (rather, straight lines) for each of those resistors and can be overlayed against the speaker curve. Piece of cake.
When REW boots, it always starts by harassing you to set the sound level to match your sound level meter. In practice, you will be shifting volume controls hither and yon and whatever truth to initial settings, that truth no longer holds as soon as you crank up the volume on your amp. And knowing the absolute level is rarely important. Even when you are establishing loudness peaks or distortion at 100dB or whatever in your testing, simply reporting, "very loud" is accurate enough for sensible critics.
B.
*but when doing electrical measurements, you must care track of grounds and not get the wires reversed.
Last edited:
Just a little update, thought I'd share a link showing my build working. Still needs a little polishing, but they are getting there! - www.facebook.com/SparrowLegsSpeakerDesign/videos/1182096815222610
@Sparrowlegs -
Here is a guide that is pretty clear and simple for using LIMP and a couple components.
Easy ZMA and T/S Parameters Using The LIMP Module in ARTA - An Illustrated Guide - Techtalk Speaker Building, Audio, Video Discussion Forum
Here is a guide that is pretty clear and simple for using LIMP and a couple components.
Easy ZMA and T/S Parameters Using The LIMP Module in ARTA - An Illustrated Guide - Techtalk Speaker Building, Audio, Video Discussion Forum
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.