About speaker design - a kind question

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello !
I wonder if it is possible to design, build and optimize a speaker working only on one unit. :rolleyes:
I mean, instead of working on a LR pair just work with mono signals (test signals and music tracks of course) on a single speaker.
Then when the result is convincing just make a copy of it and you get a stereo pair ;)
The more that i think about it the more I feel the rightness of the approach.
If a speaker is good with a mono signal a pair with a stereo signal will be very good
What is stereo if not two monos ?
Anyway if you say that is a silly idea i trust you completely. :eek:

Then just one other question
The TS parameters can be measured ?
Thanks and kind regards,
gino :D
 
Last edited:
No, not silly question at all.... Actually I design all of my speakers in "mono", and when I am satisfied with the result, I just "copy" it for a pair. The only problem might be to find music material that is not "too" stereo, so both channels sound close to each other, otherwise some instruments might sound too much attenuated, because they are recorded on the other channel. For your second question, here is a little calculator for measuring T-S parameters. What you need is an accurate sinevawe generator with adjustable output, an AC RMS multimeter, and a 10 ohm resistor for calibration. If you need further info how to use it, send me a PM. Well... as it turned out .exe files cannot be uploaded, so if you want it, I can send it to your E-mail.
 
No, not silly question at all....

Thank you ! :happy2:
i am regaining trust in myself
I tell you why this idea. Yesterday i listened from a little center alone music
I was surprised by the quality
I am pretty sure that two of those center would make a nice stereo pair indeed.

Actually I design all of my speakers in "mono", and when I am satisfied with the result, I just "copy" it for a pair.
The only problem might be to find music material that is not "too" stereo, so both channels sound close to each other, otherwise some instruments might sound too much attenuated, because they are recorded on the other channel.

Thanks again. I am going to try for curiosity to get mono from stereo recordings that i have and like so much and in this way to create a library of mono music to use for some experiments with just one speaker.
I believe in the rightness of this approach.
As i believe that a symmetric speaker would be better
Symmetry is a quality of beauty ... i mean no L and R speakers.


For your second question, here is a little calculator for measuring T-S parameters. What you need is an accurate sinevawe generator with adjustable output, an AC RMS multimeter, and a 10 ohm resistor for calibration. If you need further info how to use it, send me a PM. Well... as it turned out .exe files cannot be uploaded, so if you want it, I can send it to your E-mail.

Your reply answers completely my need. TS parameters can be measured.
For now it is a little complicated for me ... but i am studying.
Thank you very much again
legkedvesebb Üdvözlettel

gino :)
 
Last edited:
"How do you get to perform at symphony hall?" Practice...Practice...Practice
"How do you build great speakers?" Simulate...Simulate...Simulate
"How do you confirm a great speaker design?" Measure...Measure...Measure
"How do you finalize a great speaker design?" Get experts to Listen...Listen...Listen

Some practical considerations.

1) Great stereo speakers require more extensive controlled directivity design considerations, as well as more room placement considerations for imaging, than a mono speaker. A great mono speaker design may not be a great stereo speaker.

2) It is not unusual for a speaker manufacturing process to drift, or even change, over time. The best chance of getting closely matched speaker pairs is to purchase them at the same time and add a note to your purchase order like "I would appreciate any small effort you can make to select speakers pairs with similar date codes or serial numbers."

3) Setting up the construction machines(saws, CNC, drills, etc..) takes brain-power, trial-and-error, and jigs. It is best to fabricate parts for two speakers at the same time.

4) Take advantage of well reviewed designs from generous experts. ADOPT....or....ADAPT to your goals.
 
Many exceptional professional designers design via crossover simulation programs and in box driver testing. A single unit verifies the adequacy of the crossover network and for tuning to voice the speaker. They may not even build/test a pair of speakers.

A stereo pair does little to aid the designer in design verification. The designer is more interested in meeting his goals for frequency/phase responses plus off axis performance vs. imaging. Once he is satisfied, two identical units will have the imaging and soundstage.
 
Designing in mono will not tell you anything about how well the speaker creates animage/soundstage. So even after getting the other bits right in mono, you may still not be done.

dave

If your speaker does not create a credible image/soundstage in mono (1pc.), it will NEVER do it in stereo. When my friends hear my "mono" designs in the process, they usually comment on the depth(!) of the soundstage it creates, and none of them is a newbie in the field.
 
Excuse me, but I don't really get this at all!

A loudspeaker is a loudspeaker. All speakers are designed and measured as single units. If the layout is asymmetrical they are usually manufactured as symmetrical LR pairs.

But, if one wants to use a speaker strictly monophonic, then wide and constant radiation pattern must be superior. If well designed it will use the room as advantage in creating balanced power/room response and reflections.

If one wants to use speakers as stereo pais or in a multichannel system, narrow radiation pattern is preferred at least for front channels. Narrow pattern makes stereo imaging better because if reduces reflections (ratio between direct and radiated sound energy and delay of reflections are important parameters), simplistically said.

Naturally we have many opinions about the suitability and betterness of different systm approaches. Many listening tests and modern acoustic analysis/simulation support the general description I said above. Read Geddes, Olive, Toole et. al.

Before the era of stereophonics hifi speakers usually had several transducers pointing at different directions, also upwards. Often they were placed right next to the wall. They were called cabinet radios or systems. Most of them were integrated with radio and a record player.

History - About - Wharfedale Hi-Fi
Sound Systems: Mono vs. Stereo
hifimuseum.de - Sie sind im Bereich : Hifi Historie und Geschichte
CarlssonPlanet.com • History of the Carlsson loudpeakers
 
Last edited:
Many exceptional professional designers design via crossover simulation programs and in box driver testing.
A single unit verifies the adequacy of the crossover network and for tuning to voice the speaker.
They may not even build/test a pair of speakers.
A stereo pair does little to aid the designer in design verification.
The designer is more interested in meeting his goals for frequency/phase responses plus off axis performance vs. imaging. Once he is satisfied, two identical units will have the imaging and soundstage

This is extremely interesting. Thank you for the information.
I understand that no driver is perfectly equal to another ... but it is always the case, we have tolerances.
The better the manufacturing the smaller the deviations.
Thanks a lot
Kind regards,
gino
 
If your speaker does not create a credible image/soundstage in mono (1pc.), it will NEVER do it in stereo.
When my friends hear my "mono" designs in the process, they usually comment on the depth(!) of the soundstage it creates, and none of them is a newbie in the field

Hello ! you have mentioned my magic word ... depth !
When i read that i get quite excited
Tell me please, which mono tracks do you use to assess the "depth" ability of a speaker ?
For me depth IS the thing really. Ok ... the room is also important but the speaker must be able to reproduce it clearly.
May i ask you some pictures of your designs ? i would be interested to have a look
Thank you very much indeed.
This is maybe the most interesting topic for me presently
To find mono tracks to assess the depth ability.
Kindest regards,
gino
 
I think your testing method is too subjective, but the idea testing a single loudspeaker is correct.

There are a number of good and scholarly articles on testing loudspeakers, none use listening as a sole means to "test" the loudspeaker design.

You need a minimum amount of equipment to begin, which allows you to get the T/S parameters for the drivers you are using so you can begin the design, a mechanism to generate a signal to drive the loudspeaker, and a test microphone with software to plot the resulting data.

Typically, loudspeakers are tested as single units and the tests are designed to verify that they comply with the original design goals. Tests confirm impedance, frequency response, distortion, directivity over the frequency range, transient response, and relative efficiency.

Testing itself is an art because you must set up a teat environment that is as free from room interactions as possible. This is not an easy task and there are tricks to minimize room interactions or simply test the loudspeaker outside in a quite place. You need to elevate the loudspeaker above the ground and any surrounding obstacles to avoid interactions.

Once this testing is complete you can have a good idea how a pair of speakers will perform, but the room you use them in may confound your efforts. However, that is another subject. Your testing and design goals should be to make the loudspeaker perform the best it can in an ideal setting, then treat the room as required to get as close as an ideal environment as you can.

While this may seem like a lot more than what you had originally imagined, you should consider that the results are going to be directly proportional to the effort applied.

If you expect good or better results you must be prepared to do the due diligence with the design and testing of your loudspeaker. Dumb luck is not a strategy.

I strongly urge you to buy The Loudspeaker Design Cookbook by Vance Dickason and read it before you do anything else. It will answer all of your questions and help you complete a very worthy set of loudspeakers that you will be both happy and proud of.

Hello !
I wonder if it is possible to design, build and optimize a speaker working only on one unit. :rolleyes:
I mean, instead of working on a LR pair just work with mono signals (test signals and music tracks of course) on a single speaker.
Then when the result is convincing just make a copy of it and you get a stereo pair ;)
The more that i think about it the more I feel the rightness of the approach.
If a speaker is good with a mono signal a pair with a stereo signal will be very good
What is stereo if not two monos ?
Anyway if you say that is a silly idea i trust you completely. :eek:

Then just one other question
The TS parameters can be measured ?
Thanks and kind regards,
gino :D
 
Last edited:
I think your testing method is too subjective, but the idea testing a single loudspeaker is correct.

Thank you very much indeed for the valuable confirmation.
In general i prefer testing with instruments of course.
All major manufacturers have really expensive and state of the art equipment for testing. So they have a reason to exist.
Mine is more a feeling that something based on science.
I truly believe in two identical speakers for L and R duties.
And testing just one it seems to me easier all considered.
At the point that the idea of using some selected and specifically made mono tracks using only one speaker to judge the performance of a pair will be my next move.
As i said above if a single speaker is able to reproduce very well a mono track two of them should be perfect for stereo.
Maybe is wrong but i want to try anyway.
Moreover i think that the tests in anechoic room are carried out on just one speaker, for instance.


There are a number of good and scholarly articles on testing loudspeakers, none use listening as a sole means to "test" the loudspeaker design. You need a minimum amount of equipment to begin, which allows you to get the T/S parameters for the drivers you are using so you can begin the design, a mechanism to generate a signal to drive the loudspeaker, and a test microphone with software to plot the resulting data.

Actually i asked about T&S parameters because this is indeed the starting point.
The ability to measure them is out of my current ability completely, but i understand that this is a fundamental first step.
If i am not wrong with T&S + box size the computer is able to give the simulated freq response. This is just great.

Typically, loudspeakers are tested as single units and the tests are designed to verify that they comply with the original design goals.
Tests confirm impedance, frequency response, distortion, directivity over the frequency range, transient response, and relative efficiency.
Testing itself is an art because you must set up a teat environment that is as free from room interactions as possible.
This is not an easy task and there are tricks to minimize room interactions or simply test the loudspeaker outside in a quite place.
You need to elevate the loudspeaker above the ground and any surrounding obstacles to avoid interactions
This sounds completely beyond my possibilities :(
I trust you completely anyway.
I think that the test in anechoic room try to minimize indeed the room interactions you mention.
Of course the listening room is not anechoic.
But i think that testing in the actual listening room should be also important.
Once this testing is complete you can have a good idea how a pair of speakers will perform, but the room you use them in may confound your efforts. However, that is another subject.
Your testing and design goals should be to make the loudspeaker perform the best it can in an ideal setting, then treat the room as required to get as close as an ideal environment as you can.

Why not testing in the real room ?
I was impressed by some video on youtube about active equalization
Very impressive. The eq sound was so much better even through youtube.
Very much better.

While this may seem like a lot more than what you had originally imagined, you should consider that the results are going to be directly proportional to the effort applied.
If you expect good or better results you must be prepared to do the due diligence with the design and testing of your loudspeaker.
Dumb luck is not a strategy.
I strongly urge you to buy The Loudspeaker Design Cookbook by Vance Dickason and read it before you do anything else.
It will answer all of your questions and help you complete a very worthy set of loudspeakers that you will be both happy and proud of

Thank you very much for the helpful advice
Actually i would like to start modifying some commercial speakers.
I have already noticed, for instance, that replacing parts in the x-over with other better quality improve the sound noticeably.
But i would very much need a set up for measuring the TS parameters.
This is a milestone. No serious project can start without knowing these parameters of the drivers used.
I still have to check but i am pretty sure that the drivers that i would like to replace are unnamed.
This is very unfortunate because i am sure of their very low quality.
Thanks again for the kind and valuable advice.
Kind regards,
beppe
 
I highly encourage you to reconsider. There are free measuring software available like HolmImpulse that does a fantastic job of generating swept sine waves and recording the resulting response. A simple multimeter is another valuable tool.

Microphones are not necessarily expensive and you can always DYI an electric mic for cheap.

You will need a sound card to get signals out and into your computer, but they are not expensive either.

Useful and reliable testing can be done on a shoestring. The challenge is not the money, it is learning the skill. Even that is very doable.

Also, testing in the room environment is not a good idea because the results vary widely depending on positions of the loudspeaker and the positions of the listener.

Just read the suggested book and you will see it is easier than you thing to do reasonable testing.

Using an EQ to correct for room issues always fails. You may get a single spot in the room to sound better, but you can't fix the room.

Lastly, if testing is out of the question, then I would highly recommend building a kit speaker that already has all the engineering worked out. This way you simply follow a blueprint of the design and plug and play.

There is nothing wrong with this approach and it is highly recommended for the first-time builder as it saves a lot of heartaches and waster time and money.

DIY speaker building need not be a financially expensive endeavor. You can do an amazing job with a shoestring budget, but you should count on spending a lot of time.

Thank you very much indeed for the valuable confirmation.
In general i prefer testing with instruments of course.
All major manufacturers have really expensive and state of the art equipment for testing. So they have a reason to exist.
Mine is more a feeling that something based on science.
I truly believe in two identical speakers for L and R duties.
And testing just one it seems to me easier all considered.
At the point that the idea of using some selected and specifically made mono tracks using only one speaker to judge the performance of a pair will be my next move.
As i said above if a single speaker is able to reproduce very well a mono track two of them should be perfect for stereo.
Maybe is wrong but i want to try anyway.
Moreover i think that the tests in anechoic room are carried out on just one speaker, for instance.




Actually i asked about T&S parameters because this is indeed the starting point.
The ability to measure them is out of my current ability completely, but i understand that this is a fundamental first step.
If i am not wrong with T&S + box size the computer is able to give the simulated freq response. This is just great.


This sounds completely beyond my possibilities :(
I trust you completely anyway.
I think that the test in anechoic room try to minimize indeed the room interactions you mention.
Of course the listening room is not anechoic.
But i think that testing in the actual listening room should be also important.


Why not testing in the real room ?
I was impressed by some video on youtube about active equalization
Very impressive. The eq sound was so much better even through youtube.
Very much better.



Thank you very much for the helpful advice
Actually i would like to start modifying some commercial speakers.
I have already noticed, for instance, that replacing parts in the x-over with other better quality improve the sound noticeably.
But i would very much need a set up for measuring the TS parameters.
This is a milestone. No serious project can start without knowing these parameters of the drivers used.
I still have to check but i am pretty sure that the drivers that i would like to replace are unnamed.
This is very unfortunate because i am sure of their very low quality.
Thanks again for the kind and valuable advice.
Kind regards,
beppe
 
You wrote, "This sounds completely beyond my possibilities..."

At one time reading was beyond your possibilities. However, you were able to learn how to read and write and you are certainly able to learn how to design, build, and test loudspeakers.

You are simply about to embark on learning a new skill. All you need to do is simply try. There are some real experts here that are very willing to help you at any step of the way for free. What an amazing opportunity that is!
 
Mic's good enough for crossover work are really easy. See Linkwitz. Most of the software is freeware. Measure, measure, measure. Someone said that already.

Let me add, measuring gets you to find what is wrong that you are hearing easier and quicker than the old way: Lots of folks milling around, lots of listening, and lots of swapping parts. Some decent speakers have been made that way. I venture no great speakers were. Hearing is truth. I do final voicing in stereo, with my subs, in the room where I am going to listen... I have never tuned a speaker out of an imaging problem, but I have tuned rooms out of imaging problems.

Everything you need to know about measurement is in Joe D'Appolitos book, Measuring loudspeakers. Understand what you are doing, so when you use WooferTester, LIMP, HOLM or whatever, you understand what you are measuring. You must learn to interpret your measurements. You don't get simple numbers.

TO really appreciate the tools, you can always measure T/S parameters with a signal generator ( PC) a resistor and a voltmeter. It takes hours and then yo plot it on graph paper. I believe it is worth doing once to really understand what the computer is doing for you. I won't give up my trig calculators, but I know how to do it by hand too.

Read all the documentation on HOLM, ARTA, LIMP, TrueAudio RTA and even read the manual for SoundEasy. Read sights like Zaph and True Audio.
 
You wrote, "This sounds completely beyond my possibilities..."
At one time reading was beyond your possibilities. However, you were able to learn how to read and write and you are certainly able to learn how to design, build, and test loudspeakers.
You are simply about to embark on learning a new skill. All you need to do is simply try.
There are some real experts here that are very willing to help you at any step of the way for free. What an amazing opportunity that is!

Actually i can read a little now
and this amazes me sometimes :D
Joke aside i am sure that the learning path starts with reading and then measuring drivers alone.
It is important to assess the properties of the main elements of a speaker
Then great design tools exist.
A friend uses a simulator for bass response.
I will try. I promise.
And thank you again for the kind and valuable support.
Kind regards,
gino
 
Mic's good enough for crossover work are really easy. See Linkwitz.
Most of the software is freeware. Measure, measure, measure.
Someone said that already.
Let me add, measuring gets you to find what is wrong that you are hearing easier and quicker than the old way: Lots of folks milling around, lots of listening, and lots of swapping parts. Some decent speakers have been made that way. I venture no great speakers were. Hearing is truth. I do final voicing in stereo, with my subs, in the room where I am going to listen...
I have never tuned a speaker out of an imaging problem, but I have tuned rooms out of imaging problems.
Everything you need to know about measurement is in Joe D'Appolitos book, Measuring loudspeakers. Understand what you are doing, so when you use WooferTester, LIMP, HOLM or whatever, you understand what you are measuring.
You must learn to interpret your measurements. You don't get simple numbers.
TO really appreciate the tools, you can always measure T/S parameters with a signal generator ( PC) a resistor and a voltmeter. It takes hours and then yo plot it on graph paper. I believe it is worth doing once to really understand what the computer is doing for you. I won't give up my trig calculators, but I know how to do it by hand too.
Read all the documentation on HOLM, ARTA, LIMP, TrueAudio RTA and even read the manual for SoundEasy. Read sights like Zaph and True Audio.

Sorry, do you mean that you voice your speakers by ear ? Really ?
changing elements in the x-over maybe ?
Thanks and regards,
gino
 
TO really appreciate the tools, you can always measure T/S parameters with a signal generator ( PC) a resistor and a voltmeter. It takes hours.....

Hours? To measure T/S manually it takes about 10 minutes. It might take hours if you include construction of a test box and searching out all of the formulas, but the actual measurement is rather quick.

I agree that doing it the "hard way" makes you appreciate what Speaker Workshop does. I use speaker workshop for simulations and impedance measurements, but it doesn't seem to do as well for me on acoustic measurements on any of the PC's I have tried. I use HOLM for acoustic measurements and import into SW.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.