Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Multi-Way Conventional loudspeakers with crossovers

Phase shift due to dipole D and its effect upon crossovers
Phase shift due to dipole D and its effect upon crossovers
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 7th August 2012, 11:20 AM   #1
gainphile is offline gainphile  Australia
diyAudio Member
 
gainphile's Avatar
 
Join Date: Dec 2007
Location: Melbourne the sunny city!
Default Phase shift due to dipole D and its effect upon crossovers

What are the practical implications of this

linkwitzlab.com/frontiers_7.htm#D2

It seems that a dipole loudspeakers with varying baffle width (Like NaO Note) can experience problems when standard crossover is used.

I did brief experiment before which although measures well, never sounded right and was quite suspicious about the baffle size transition. Although never been able to explain it.

www.diyaudio.com/forums/multi-way/161681-my-s13-ob-uniform-polar-response-tweeters-last-2.html#post2097027

Click the image to open in full size.


And what about no-baffle dipoles ??
__________________
http://gainphile.blogspot.com
  Reply With Quote
Old 7th August 2012, 04:15 PM   #2
john k... is offline john k...  United States
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
The axial response of any unequalized dipole will with have a phase response will be zero at the dipole peak and approach 90 degrees at DC. Above the dipole peak the response will exhibit the usual peak/null behavior. The higher the dipole peak in frequency (smaller D), the more rapidly the phase approaches the DC value of 90. However, once the response is equalized to some band pass, the axial response of that band pass will be the same as that of a boxed driver with the same band pass because the axial response of a dipole, equalized or not, is minimum phase. The net result is that axial sum is no different than that of summing monopoles in that the different AC positions must be considered. Of course, the problem with using true dipoles above the dipole peak is that the nulls can not be eq'ed. But driver directionality is usually used to counter that effect.

You brought up my NaO Note. Just where are these problems SL alludes to in it's axial response?

Click the image to open in full size.

The bottom line is that this is no different than a typical box speaker. You measure the axial amplitude and phase response of each source. That response will be minimum phase plus some contribution for AC offset. That there may be a phase difference due to the position of the dipole peak (or D) is no different that the observation that an LP filter with cut off of 1k Hz Hz has different phase at 100 Hz than an LP filter with cut off of 300 Hz. It's much ado about nothing; SL is just making things seem more complex than they are, as he typically does.
__________________
John k.... Music and Design NaO dsp Dipole Loudspeakers.
  Reply With Quote
Old 7th August 2012, 05:21 PM   #3
ra7 is offline ra7  United States
diyAudio Member
 
Join Date: Feb 2009
Phase shift due to dipole D and its effect upon crossovers
John, that looks almost anechoic flat. What type of measurement is this? Anechoic on axis, quasi-anechoic, spatially averaged, or is it the simulated response? Do you have a listening position response in your room?
  Reply With Quote
Old 7th August 2012, 06:50 PM   #4
speaker dave is offline speaker dave  United States
diyAudio Member
 
speaker dave's Avatar
 
Join Date: Nov 2009
Location: The Mountain, Framingham
Linkwitz is talking about the delay of the combined front and back radiation (as seen from the front) of an open baffle dipole. He is essentially saying that since the rear radiation that gets around to the front is significantly delayed the combined output is delayed as the vector sum of the two. This is all true and is another way of describng why the unequalized dipole rolls off at lower frequencies. It is still minimum phase and equalization of the dipole response to flat will improve the phase.

The curves he shows are typical summation of a delayed woofer and less delayed midrange. this looks like a lot of delay to me. Except for the case of adding a short tweeter horn to a long mid horn I never see multiple rotations (the ripple) as he is showing. Recessing the mid could bring them well into phase. I guess SL's argument is that the dipole section will need more delay compensation (electrical or physical) but I'm not sure I would generalize to that extent.

David S.
  Reply With Quote
Old 7th August 2012, 07:52 PM   #5
john k... is offline john k...  United States
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by ra7 View Post
John, that looks almost anechoic flat. What type of measurement is this? Anechoic on axis, quasi-anechoic, spatially averaged, or is it the simulated response? Do you have a listening position response in your room?
The response shown is a combination of a 1/6 octave smoothed far field (1M quasi-anechoic) measurement combined with a dipole corrected near field measurement below 385Hz.

Quote:
Originally Posted by speaker dave View Post
Linkwitz is talking about the delay of the combined front and back radiation (as seen from the front) of an open baffle dipole. He is essentially saying that since the rear radiation that gets around to the front is significantly delayed the combined output is delayed as the vector sum of the two. This is all true and is another way of describng why the unequalized dipole rolls off at lower frequencies. It is still minimum phase and equalization of the dipole response to flat will improve the phase.

The curves he shows are typical summation of a delayed woofer and less delayed midrange. this looks like a lot of delay to me. Except for the case of adding a short tweeter horn to a long mid horn I never see multiple rotations (the ripple) as he is showing. Recessing the mid could bring them well into phase. I guess SL's argument is that the dipole section will need more delay compensation (electrical or physical) but I'm not sure I would generalize to that extent.

David S.
As I said, he is making it seem more complex than it is. The summed front and rear dipole response measured at any position (on or off axis) is minimum phase. The result that a dipole on a circular baffle of diameter D has different phase than one on a baffle of diameter D/x is simply a consequence that the dipole peak is shifted, yielding different amplitude, hence, minimum phase.

If you have a dipole source on a 10" circular baffle and another on a 3" dia. baffle and you could eq both to the exact same band pass response then they will have the exact same phase regardless of the dipole delay. The problem is that due to the nulls above the dipole peak the response above the peak can not be made identical in practice.

For example, in theory you could take the 10 dipole and eq it to a perfect B1 LP at 500 Hz and take the 3" dipole and eq it to a perfect B1 HP at 500 Hz and they would sum perfectly. The dipole delay would be negated by the eq.
__________________
John k.... Music and Design NaO dsp Dipole Loudspeakers.
  Reply With Quote
Old 7th August 2012, 09:17 PM   #6
speaker dave is offline speaker dave  United States
diyAudio Member
 
speaker dave's Avatar
 
Join Date: Nov 2009
Location: The Mountain, Framingham
Quote:
Originally Posted by john k... View Post

If you have a dipole source on a 10" circular baffle and another on a 3" dia. baffle and you could eq both to the exact same band pass response then they will have the exact same phase regardless of the dipole delay. The problem is that due to the nulls above the dipole peak the response above the peak can not be made identical in practice.

For example, in theory you could take the 10" dipole and eq it to a perfect B1 LP at 500 Hz and take the 3" dipole and eq it to a perfect B1 HP at 500 Hz and they would sum perfectly. The dipole delay would be negated by the eq.
Yes, both assuming the drivers have alligned acoustic centers. (probably not the same baffle planes) It still isn't clear to me why the SL example shows as much phase rotation as it does.

David S.
  Reply With Quote
Old 7th August 2012, 10:56 PM   #7
john k... is offline john k...  United States
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by speaker dave View Post
Yes, both assuming the drivers have alligned acoustic centers. (probably not the same baffle planes) It still isn't clear to me why the SL example shows as much phase rotation as it does.

David S.
The thing is Dave, ask yourself, "How is this different from ordinary diffraction?" The answer is there really isn't any difference except the strength of the inverted wave. It makes no difference is the wave come from the rear and wraps around the baffle or if it's a surface wave that travels across the baffle and is then diffracted at the edge; the delay is the same. The difference is only in the strength of that diffracted wave. So the same type of argument can be made for any speaker, dipole, box, or otherwise. It's a simply matter of multiple minimum phase sources being summed to provide flat response on axis. You do what is necessary, add delay, slope the baffle, offset the drivers, what ever. The idea that this is some how a dipole problem is a misrepresentation.

I'll put up some figures later if I feel like it.

As for SL's phase differences, I think he is off by a factor of 2. I get about 16 degrees between mid and woofer for a 10 and 40 cm equivalent delay. Interestingly, if I do a crude box simulation for baffle step with similar baffles I get about the same "baffle diffraction induced" phase difference at 100 Hz. But again, this is corrected by baffle step compensation. The important thing to note that this is not the same as AC offset. This is phase difference due to differences in the amplitude response and the consequence of minimum phase. So I do not think it is correct to look at the phase difference between unequalized dipoles on different size baffles, at some particular crossover frequency, and conclude that that phase difference carries over to a necessary delay compensation for the crossover frequency.
__________________
John k.... Music and Design NaO dsp Dipole Loudspeakers.
  Reply With Quote
Old 8th August 2012, 09:37 AM   #8
gainphile is offline gainphile  Australia
diyAudio Member
 
gainphile's Avatar
 
Join Date: Dec 2007
Location: Melbourne the sunny city!
So in summary the problem does not exist if the drivers are equalised appropriately on-axis ?

What software do you use to calculate simulate this?
__________________
http://gainphile.blogspot.com
  Reply With Quote
Old 8th August 2012, 01:46 PM   #9
john k... is offline john k...  United States
diyAudio Member
 
john k...'s Avatar
 
Join Date: Aug 2004
Location: US
Quote:
Originally Posted by gainphile View Post
So in summary the problem does not exist if the drivers are equalised appropriately on-axis ?

What software do you use to calculate simulate this?
Basically that is it, as long as you are concerned only with the forward axial response. Things get very complicated if you are trying to design a front-rear symmetric dipole.

I put togeter a short write up this morning; Dipole-mp
__________________
John k.... Music and Design NaO dsp Dipole Loudspeakers.
  Reply With Quote
Old 8th August 2012, 03:27 PM   #10
speaker dave is offline speaker dave  United States
diyAudio Member
 
speaker dave's Avatar
 
Join Date: Nov 2009
Location: The Mountain, Framingham
An interesting tutorial but I'm not comfortable with the "no delay" conclusion.

As I understand, you are starting with a wide band driver of no depth and a second similar unit with 10cm recess and you end up with a continuous downward roll of phase (that cycles with every null). I would have thought that this downward phase droop was evidence of the delay of the simulated dipole.

Comparing the unit's phase to the Hilbert transform of the same response isn't proof of no delay if the delay is inherent in the response itself.

David
  Reply With Quote

Reply


Phase shift due to dipole D and its effect upon crossoversHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
X-over Phase shift question kiwilistener Multi-Way 2 29th May 2012 11:49 AM
Phase Shift PWM Andrew Eckhardt Class D 6 14th November 2010 08:56 AM
does early effect shift much within a device family? unclejed613 Solid State 18 20th January 2010 03:12 AM
Why am I getting a positive phase shift sigpl Solid State 22 29th December 2009 05:31 PM
Zero phase-shift gareth Solid State 11 9th May 2008 05:45 PM


New To Site? Need Help?

All times are GMT. The time now is 12:28 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki