Speakerplans.com Homepage
Forum Home Forum Home > General > Advanced Discussion
  New Posts New Posts RSS Feed - Controlling Xmax of loudspeakers
  FAQ FAQ  Forum Search   Events   Register Register  Login Login

Controlling Xmax of loudspeakers

 Post Reply Post Reply Page  <123
Author
Message
Elliot Thompson View Drop Down
Old Croc
Old Croc
Avatar

Joined: 02 April 2004
Location: United States
Status: Offline
Points: 5172
Post Options Post Options   Thanks (0) Thanks(0)   Quote Elliot Thompson Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 6:34am
Originally posted by odc04r odc04r wrote:

Sorry Elliott, but you're a little off base here. An impedance plot can tell you a lot of things but it does not strictly correlate with excursion.

Consider a sealed enclosure, one resonant peak but at frequencies under the resonance the driver quickly becomes excursion limited. At an equivalent frequency abover resonance it will not be. However impedance at both points will be roughly the same. You cannot directly link an impedance plot to driver excursion unless you  understand the electro-mechanical equivalent circuit of your enclosure.

1 - Acoustic, deals with transferring cone velocity to volume velocity via radiation impedance
2 - Mechanical, deals with conversion of current I to loudspeaker force via F=Bl*i
3 - Electronic, includes external resistance/impedance such as voicecoil, cable, amplifier output

All of these 3 things act to give you a total impedance measured impedance curve. If you want to know excursion you solve the model to find mechanical velocity and integrate the expression to obtain displacement. This is how WinISD or any other TS based software does it.

If you want to use an impedance plot as a rough guide, then a common sense approach is that under the lowest system resonance the driver excursion will rise considerably. You would set a high pass approximately here, but there are several variables to consider as to what. And as mentioned excursion is good because it cools the voice coil. You need to strike a balance between that and the rest of the musical content's power spectrum. As there are so many variables to consider it is no surprise that a lot of it is done by ear, after making a reasonable first guess with the info you have to hand. The ultimately most sensible approach is to bring more gear than you need and run it all with plenty of headroom.

You could implement a solution as suggested in the article, it's a nice academic concept but I fear in a real life situation it would probably be too much faff for little gain. Just bring the right gear in the first place and don't try to milk every last dB. Low freq SPL drops off drastically if you can't provide the displacement necessary to create it anyway, so not sure how real world viable the idea is.

This is the problem with simulators. They never co-relate to the impedance results using ARTA and/or REW. This is due to the fs, VAS in addition to Qts of the driver changes once it is placed in the box.

 

If you measure the TS Parameters of a driver in free air versus the Manufactures TS Parameters you will attain different results on the impedance curve.

 

If you measure TS Parameters in the box versus the simulated TS Parameters (Measured or Manufacture) in the box, you will attain different results on the impedance curve.  

 

Simulators are like Protein Powder. It may appear to offer all the necessary nutrients in terms of proteins however Protein Powder cannot give you the same results as eating real food supplied with protein.  

 

The more I measure loudspeakers, the more I realise Simulators are a good starting point to familiarise yourself with loudspeaker design. However, at some point you will need to start literally measuring loudspeakers to get the in depth analysis on the results.

 

We also need to take into consideration exceeding the Xmax does not damage the driver. It is exceeding the Xmech limit (One way Peak or Peak to Peak) that damages the driver.

 

Best Regards,

 

Elliot Thompson
Back to Top
Crashpc View Drop Down
Registered User
Registered User


Joined: 26 February 2008
Location: Czech Republic
Status: Offline
Points: 463
Post Options Post Options   Thanks (0) Thanks(0)   Quote Crashpc Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 7:49am
Originally posted by gen0me gen0me wrote:

Crashpc according to the method from the first page(flat excursion) you could actually cool driver on frequencies that are usually under your high pass filterWink.


Eeee I don´t think so. Not really.
It is doable, but it would bring too much trouble that it is probably not worth. I was thinking about this before.
Firstly, If you move the speaker below its usable pass band, it doesn´t mean there is not significant SPL output. That alone might cause some trouble. My 29-33Hz tuned bins shake the ceiling at 14-25Hz, even though, there is not much sound to hear.
Secondly, if you du it in regular music, you offset the cone (irregular fashion makes this even worse), so you lower your cone excursion range for your usable pass band.
Then You need to put significant amount of power to the coil to move it, yet you don´t have any usable sound from it. Many enclosures have impedance minimums under the HPF subsonic, so that would also add another heat.
Then cooling the cone with low frequency is less efficient. You really WANT to cool it down at lets say 40+Hz. That is efficient.

You still could solve SOME of these problems, but at a cost of advanced processing and programming, which would rather add so much price for an individual, that it is lightyears easier to just buy better speaker or build multiple bins...


Nikon and Canon people should not be married to each other. Why did you let this happen?
Back to Top
gen0me View Drop Down
Young Croc
Young Croc
Avatar

Joined: 20 February 2016
Location: UK
Status: Offline
Points: 999
Post Options Post Options   Thanks (0) Thanks(0)   Quote gen0me Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 8:42am
Originally posted by Crashpc Crashpc wrote:

Firstly, If you move the speaker below its usable pass band, it doesn´t mean there is not significant SPL output. That alone might cause some trouble. My 29-33Hz tuned bins shake the ceiling at 14-25Hz, even though, there is not much sound to hear.
It doesnt have to sound. It has to move. But there might be something about colorizing lower frequencies with harmonic oscilator resonance frequency.
Originally posted by Crashpc Crashpc wrote:

Secondly, if you du it in regular music, you offset the cone (irregular fashion makes this even worse), so you lower your cone excursion range for your usable pass band.
I dont see any cones in this. If in the music there is already such material then there are no higher bass frequencies. Simply on studio mix level under 40hz bass means giving up on any higher notes.
I appreciate every like :)) https//www.facebook.com/genomesoundsystems
Mixes: https://www.mixcloud.com/gen-ome/
Back to Top
Crashpc View Drop Down
Registered User
Registered User


Joined: 26 February 2008
Location: Czech Republic
Status: Offline
Points: 463
Post Options Post Options   Thanks (0) Thanks(0)   Quote Crashpc Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 8:45am
Elliot: Never say never. WinISD simmed my 21DS115 box within usable passband of 25-100Hz VERY acurately. Both phase and impedance. It is not THAT bad always. On the other hand, Hornresp failed in this regard. It was little more accurate in frequency response graph though.
Yet there are things which are not being simmed, and also very different things happen with large signals, which are used most often for real work with the speaker. Of course...

I hope nobody ever recommends a simulator or protein powder as a final decisive factor/solution.

Of course cone excursion is related to the impedance, no doubt. But it is not shackled to it.
I can present it on my development box too. There are many aspects playing the role in the cone excursion, and accoustical impedance is only one of these, altering the speaker impedance.
Not that you didn´t have a good point, it is just not that black and white.


Nikon and Canon people should not be married to each other. Why did you let this happen?
Back to Top
Crashpc View Drop Down
Registered User
Registered User


Joined: 26 February 2008
Location: Czech Republic
Status: Offline
Points: 463
Post Options Post Options   Thanks (0) Thanks(0)   Quote Crashpc Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 8:55am
Originally posted by gen0me gen0me wrote:

Originally posted by Crashpc Crashpc wrote:

Firstly, If you move the speaker below its usable pass band, it doesn´t mean there is not significant SPL output. That alone might cause some trouble. My 29-33Hz tuned bins shake the ceiling at 14-25Hz, even though, there is not much sound to hear.
It doesnt have to sound. It has to move. But there might be something about colorizing lower frequencies with harmonic oscilator resonance frequency.


Eeee, I don´t think you understood. It doesn´t have to sound, but IT DOES.
And that is quite a problem.  In order to not make it do sounds, you need to move the speaker waay way down under the pass band. And it usually means a lot of power due to the low impedance. That is contraproductive, because at low frequencies, cooling is not that effective.

Originally posted by gen0me gen0me wrote:

Originally posted by Crashpc Crashpc wrote:

Secondly, if you do it in regular music, you offset the cone (irregular fashion makes this even worse), so you lower your cone excursion range for your usable pass band.
I dont see any cones in this. If in the music there is already such material then there are no higher bass frequencies. Simply on studio mix level under 40hz bass means giving up on any higher notes.


Any cones? I don´t understand here. If you cool down your speaker by cone movement, and the cone is offset by that movement, and audio signal comes to the speaker in phase at that moment, the cone will push even further. But as the cone was in offset position already, and the speaker is limited by Xmax, or power, you leave very little cone excursion available for usable content, because you used it for the cooling.
It is quite "rare", that there is musical sub 25Hz material in the music. Especially for PA.
But I would not risk thinking that there is nothing all the time. That might end up in a catastrophy again.


Edited by Crashpc - 31 December 2018 at 8:58am
Nikon and Canon people should not be married to each other. Why did you let this happen?
Back to Top
gen0me View Drop Down
Young Croc
Young Croc
Avatar

Joined: 20 February 2016
Location: UK
Status: Offline
Points: 999
Post Options Post Options   Thanks (0) Thanks(0)   Quote gen0me Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 9:42am
Originally posted by Crashpc Crashpc wrote:

But as the cone was in offset position already, and the speaker is limited by Xmax, or power, you leave very little cone excursion available for usable content, because you used it for the cooling. 
It is quite "rare", that there is musical sub 25Hz material in the music. Especially for PA.
But I would not risk thinking that there is nothing all the time. That might end up in a catastrophy again.
It wont happen. You cant have both type of basses simultaneously in the mix. One bassline near 0db fulfills the hole mix. And if there comes the kick - bass is sidechained.

What were your port input parameters for winisd and hornresp?


Edited by gen0me - 31 December 2018 at 9:42am
I appreciate every like :)) https//www.facebook.com/genomesoundsystems
Mixes: https://www.mixcloud.com/gen-ome/
Back to Top
gen0me View Drop Down
Young Croc
Young Croc
Avatar

Joined: 20 February 2016
Location: UK
Status: Offline
Points: 999
Post Options Post Options   Thanks (0) Thanks(0)   Quote gen0me Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 9:49am
Even going smarter you can make generator 20hz that is sidechained under djs output and delayed right way, with right release time.

Although the difference in all mechanical ts would be ridiculous at the end of the show.Embarrassed


Edited by gen0me - 31 December 2018 at 9:57am
I appreciate every like :)) https//www.facebook.com/genomesoundsystems
Mixes: https://www.mixcloud.com/gen-ome/
Back to Top
odc04r View Drop Down
Old Croc
Old Croc
Avatar

Joined: 12 July 2006
Location: Sarfampton
Status: Offline
Points: 5483
Post Options Post Options   Thanks (0) Thanks(0)   Quote odc04r Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 10:50am
Originally posted by Elliot Thompson Elliot Thompson wrote:

This is the problem with simulators. They never co-relate to the impedance results using ARTA and/or REW. This is due to the fs, VAS in addition to Qts of the driver changes once it is placed in the box.

If you measure the TS Parameters of a driver in free air versus the Manufactures TS Parameters you will attain different results on the impedance curve.

If you measure TS Parameters in the box versus the simulated TS Parameters (Measured or Manufacture) in the box, you will attain different results on the impedance curve.  

Simulators are like Protein Powder. It may appear to offer all the necessary nutrients in terms of proteins however Protein Powder cannot give you the same results as eating real food supplied with protein.  

The more I measure loudspeakers, the more I realise Simulators are a good starting point to familiarise yourself with loudspeaker design. However, at some point you will need to start literally measuring loudspeakers to get the in depth analysis on the results.

We also need to take into consideration exceeding the Xmax does not damage the driver. It is exceeding the Xmech limit (One way Peak or Peak to Peak) that damages the driver.

Best Regards,



Yep. TS properties are just a starting point, they describe a simplified small signal lumped linear model after all. It would be unrealistic to expect anything else. But as we say in the UK (sometimes), still better than a kick in the balls.

A driver's TS parameters do not change when placed in an enclosure, the net sum (of e.g. compliance) is modified with respect to the properties of the enclosure such as a sealed enclosures volume compliance adding to the drivers compliance to shift up the original Fs and Qts, but the drivers individual properties should not change at small signal levels. If you were to construct a full circuit model and decompose the total system impedance plot to it using regression you'd see that, although there is a problem in that some of the model properties will inevitable be able to compensate for others. There is a good read on the subject here: http://www.volucres.fr/AudioHighEnd/resources/SciAudioBoxEn/Vented-box-parameter.pdf

It would be a sensible approach to assume that there will be up to 10% deviation in free air TS from any manufacturers published specs to allow for manufacturing variation imo, especially Fs which is subject to multiple manufacturing variations that could pull it either way.


Edited by odc04r - 31 December 2018 at 10:54am
Back to Top
Crashpc View Drop Down
Registered User
Registered User


Joined: 26 February 2008
Location: Czech Republic
Status: Offline
Points: 463
Post Options Post Options   Thanks (0) Thanks(0)   Quote Crashpc Quote  Post ReplyReply Direct Link To This Post Posted: 31 December 2018 at 11:15am
gen0ne: Not that it wouldn´t be doable, but it is very bad way to do that...


Nikon and Canon people should not be married to each other. Why did you let this happen?
Back to Top
 Post Reply Post Reply Page  <123

Forum Jump Forum Permissions View Drop Down

Forum Software by Web Wiz Forums® version 12.06
Copyright ©2001-2023 Web Wiz Ltd.

This page was generated in 0.125 seconds.