Gino,
if you want to reproduce a square wave, what you need first and foremost is bandwidth. That's because a square wave can be thought as a combination of sine waves of ever higher frequencies.
With this in mind any decent widerange driver will reproduce a square wave more exactly than the worlds best woofer - if we look at the single driver only.
With the same in mind, even the best crossover needs to be bad in reproducing a square wave. That is because the main duty of a crossover is to limit bandwidth in some way.
If you combine drivers with suitable bandwidths, crossed over in the right way, you will get a decent square wave with any sort of loudspeaker - ESL, dynamic, whatever. You will recognise those loudspeakers by a linear frequency response within the ears hearing limits and a deep notch in response at the crossover frequencies, when inverting the phase of one driver.
In fact this is what the Quad 63 does. It is a coaxial two-way loudspeaker with a well done crossover. No secrets inside.
Rudolf
if you want to reproduce a square wave, what you need first and foremost is bandwidth. That's because a square wave can be thought as a combination of sine waves of ever higher frequencies.
With this in mind any decent widerange driver will reproduce a square wave more exactly than the worlds best woofer - if we look at the single driver only.
With the same in mind, even the best crossover needs to be bad in reproducing a square wave. That is because the main duty of a crossover is to limit bandwidth in some way.
If you combine drivers with suitable bandwidths, crossed over in the right way, you will get a decent square wave with any sort of loudspeaker - ESL, dynamic, whatever. You will recognise those loudspeakers by a linear frequency response within the ears hearing limits and a deep notch in response at the crossover frequencies, when inverting the phase of one driver.
In fact this is what the Quad 63 does. It is a coaxial two-way loudspeaker with a well done crossover. No secrets inside.
Rudolf
This definition alone is not sufficient unfortunately. One must clearly define that they also have to be phase-linear throughoutr their passband. Although the expression "linear frequency response" would actually imply this - most people (and high-gloss brochuers as well) actally only mean a linear amplitude response when they mention "linear frequency response".
Regards
Charles
Rudolf, I owe you an explanation
I do not listen often to Square Wave music
Maybe in the future
No seriously, I listened to these Quads and I was captured by its quickness in the midband
So I start gathering info on the web about them
Then I found this review on Stereophile
They performed this SWT on the esl63 and showed the result as an exceptional one, almost unique
So I started mumbling about the sound I heard and this result
And asking here if these two things can be related
On principle I like the rule what you put in you must get out
If you put in a SW you also muste get a SW at the output
More or less, something that resembles to a SW let's say
Maybe this is trivial
But if you send in a SW and you get another wave with different shape this means distortion, at least to me
It is something like when you send in an amp a 1kHz tone and you get also 2, 3 , 4 ... kHz at the output
You get something generated by the amp but not present in the test signal
This is again distortion, not fidelity
To end all this rumblings I believe the SW is a very crtical test for a speakers, a test that has no relation with natural sound but still could be very significative
This is a very important point
One think that is not clear to me is why they performed this SWT only on the esl63
If this this test is so important they should perform it always
I am confused about this fact
If it is not so important why bother with it on the Quads ???
I think I understand and thank for your explanation
Maybe I am overrating this ability of reproducing a good SW
It is a coaxial two-way loudspeaker with a well done crossover.
No secrets inside.
Rudolf
I must confess I did not know this. I thought it were a single diaphragm transducer, like most headphones, just a little bigger
And I also thought that this performance had to do with the physical properties of the transducer
Do you know which is the crossover frequency ? which slope they adopted ?
this is extremely important to me
This speakers intrigue me a lot
I understand they have some limitations, as all others, but what I heard amazed me
Thank you very much for your kind and helpful advice
Kind regards,
gino
Last edited:
Hi Gino,
As John said, rise time has nothing to do with T/S parameters. It's simply the high frequency response. If the driver can play a higher frequency, it will have a faster rise time. Here is the impulse response of a dome tweeter. It looks pretty good but it wouldn't sound very good without a woofer to fill in the low frequencies.
Attachments
As others have said, I think the magic of ESLs is more about the dipole radiation than any specifics of drivers and/or crossovers. Cone-based dipoles with steep-slope crossovers (bad-looking square wave) can do the magic sound quite well. Popular examples include the Linkwitz Orion and JohnK's NaO.
Hi Gino,
As John said, rise time has nothing to do with T/S parameters.
It's simply the high frequency response.
If the driver can play a higher frequency, it will have a faster rise time.
Here is the impulse response of a dome tweeter.
It looks pretty good but it wouldn't sound very good without a woofer to fill in the low frequencies.
Hello and good morning !
Of course I do not compare woofers and tweeters
Let's talk about woofers ... take two different 10" woofers
Perform the Impulse Response Test (let's call it ITR)
It is possible that they have a different rise time
This rise time can be related to ANY mechanical and/or electrical measurable property of the drivers ?
That is my question
Let's take a tweeter and a woofer now
Of course we will have a much lower rise time for the tweeter
Why is that ?
Again, can this behaviour be related to ANY mechanical and/or electrical measurable property of the drivers ?
As a paragon, in a car we have for instance the ratio power/weight
Okkei ... there is also the gear box ... but with same gear box the car with the greatest ratio should have the best acceleration ...
I am everything but an expert
But I suspect, and stress suspect, that the T/S parameters can tell us something of very important about the dinamic behaviour of a driver
Please note that I am talking only of the driver as measured to get the TS parameters
For instance I have understood that a high Qts implies a fast damping after an impulse
I believe this is not trivial at all
I like a driver that when the impulse is finished stops quickly, without tails
If anyone could direct me to a rigorous but also not too difficult explanation of the T/S parameters I'd appreciate indeed
Are you sure about that ? this is my fundamental question and doubt
Are you saying that any dipole present a behaviour similar to that of ESL ?
Well, long life to dipoles then !
I think I will study them much more deeply in the next days
I know almost nothing about dipoles
Nevertheless I suspect that some ribbon non-dipole drivers could exhibit an extremely similar performance to that of the ESL63
But I do not know anything of the ribbon drivers as well
I just see some waterfall test and they seem to have very good IRT performances
I will do searches about them
Thank you very much indeed for your kind and very helpful advice
Kind regards,
gino
In both cases, high frequency response affects rise time. John said this in post#2, IIRC.
T/S parameters are defined at resonance and only valid near resonance. They (especially Cms and Rms, and to some extent Mms) actually change with frequency and drive level....
Thank you very much indeed for your kind and helpful reply
Please excuse my belated reply but I was out visiting parents and had no access to the web
I did not know of the limited validity of the T/S paremeters
Thanks again and kind regards,
gino
Gin,
Traditionally, CSD plots are used to evaluate drivers pulse response. Might visit the ZAPH site to see some. I have been playing with pure three cycle pulses as well and seeing some inteestring dynamics both on start and stop. As I have neither space nor can afford Quad's, I live in the small box domain. Other "fast" drivers, would be Heil, small planers, and true ribbons. All are not cheap if any good. But just for experiments, some of the cheap planers, usually called ribbons but are not, may provide insight to the issues.
Never found square waves of must interest in driver testing. My 2 cents.
Traditionally, CSD plots are used to evaluate drivers pulse response. Might visit the ZAPH site to see some. I have been playing with pure three cycle pulses as well and seeing some inteestring dynamics both on start and stop. As I have neither space nor can afford Quad's, I live in the small box domain. Other "fast" drivers, would be Heil, small planers, and true ribbons. All are not cheap if any good. But just for experiments, some of the cheap planers, usually called ribbons but are not, may provide insight to the issues.
Never found square waves of must interest in driver testing. My 2 cents.
Thank you very much for your very interesting reply
I think that this is a very promising testing field to assess the drivers' properties
I saw some ribbon drivers doing very well in the SWT
I saw some waterfall plots on Stereophile site
Same situation here. I found the esl63 width too much for my room
I very much prefer tower shape speakers, or satellites plus subs
A solution like the beautiful Martin Logan ... narrow front and a electrostatic unit above a good woofer for the lows
Of course is a different thing compared to the quad ... but also more practical
Actually the SWT is very uncommon in the lab reports of the speakers' reviews in the magazines
And I found particularly strange that just in the case of the esl63 they were performed and so published
I mean, if this test is important and telling why do not perform it on a regular basis ?
Thanks again and kind regards,
gino
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.