Jump to Dipole baffle width and cancellation frequencies?
Vasyakins Post
Bipolar operation
Posted By: Pete Schumacher © <pete_schumacher@yahoo.com>
Date:
Say you build an MTM for use out in the room, and full baffle step compensation is needed. Rather than contour the XO to accomplish this, design for flat and add a pair of woofers onto the rear of the enclosure. Baffle step will be taken care of.
However, now you have rear drivers that will also be filling the room with sound, just like the front drivers. Could something like that also be considered an omni?
Re: Bipolar operation
Posted By: Roman Bednarek <roman@twcny.rr.com>
Date:
In Response To: Bipolar operation (Pete Schumacher ©)
> Say you build an MTM for use out in the
> room, and full baffle step compensation is
> needed. Rather than contour the XO to
> accomplish this, design for flat and add a
> pair of woofers onto the rear of the
> enclosure. Baffle step will be taken care
> of.
> However, now you have rear drivers that will
> also be filling the room with sound, just
> like the front drivers. Could something like
> that also be considered an omni?
I haven't thought too deeply about this one but I would think that there would be some nulls in the horizontal off-axis direction when the tone from the rear-firing woofer is 180 degrees out of phase with the front-firing woofer. The thing is that those drivers would have to have good dispersion at those frequencies for this to happen. I read somewhere that the trick to having a bipolar design mimic an omni is to make the speaker very shallow so that the distance between the front and rear drivers is very small. I suspect that if a speaker is too deep the null that I spoke of before will move lower in frequency. These are just some thoughts off of the top of my head.
Bipole/Dipole....
Posted By: Wolf <shafbm01@hotmail.com>
Date:
In Response To: Re: Bipolar operation (Roman Bednarek)
> I haven't thought too deeply about this one
> but I would think that there would be some
> nulls in the horizontal off-axis direction
> when the tone from the rear-firing woofer is
> 180 degrees out of phase with the
> front-firing woofer.
While I do agree with what you stated, Pete said Bipole, so both arrays are in outward motion at the same time. Dipole is where they are in opposed motions, or opposing polarities.
Later,
Wolf
Re: Bipole/Dipole....
Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
In Response To: Bipole/Dipole.... (Wolf)
> While I do agree with what you stated, Pete
> said Bipole, so both arrays are in outward
> motion at the same time. Dipole is where
> they are in opposed motions, or opposing
> polarities.
> Later,
> Wolf
Roman is correct, Bipolar will produce some on-axis nulls. I can more later when I get home from work.
Re: More detail....
Posted By: Jeff B. <jeffb1836@yahoo.com>
Date:
In Response To: Re: Bipole/Dipole.... (Jeff B.)
> Roman is correct, Bipolar will produce some
> on-axis nulls. I can more later when I get
> home from work.
I see that John k. has replied too, and my reply is essentially the same. On a dipole the lobes are on the front and rear axis and the nulls are 90 degrees to the sides where cancellation occurs. On a bipole the combined lobes are the opposite and are at the sides; 90 degrees off-axis. Of course, this is also dependent on the radiation pattern of the driver. If very little high frequency energy is radiated at 90 degrees off-axis then they will not be a lot of energy to sum for the lobe. Unfortunately, with a bipole there will also be nulls on the front and rear axis at certain frequencies as delayed acoustical energy from the other bank of drivers makes it way around to join the on-axis sound. This delay will create out of phase conditions at certain frequencies. The frequencies this occurs at will be determined by the physical separation of the drivers and the width and depth of the enclosure. In other words, it is determined by how long it takes sound to wrap around the speaker and merge with the other axis. Again, driver radiation patterns play a role in the magnitude of these cancellations too.
I have built speakers like Pete is describing, and I didn’t notice anything other than the increased spaciousness of the speaker. One way to reduce the spaciousness and the occurrence of nulls would be to design it as a 2.5way where the rear drivers are rolled-off above the baffle step frequency. This way they will compensate for the baffle step and contribute to the low end response without changing the on-axis response in the upper range.
Jeff B.
Re: More detail....
Posted By: Wolf <shafbm01@hotmail.com>
Date:
In Response To: Re: More detail.... (Jeff B.)
I do recall seeing your small-box 2.5 way speakers in pictures. I have a slightly better grasp now. I guess I didn't know the cancellations would still be there in front and back, albeit at certain areas of the sound spectrum. Again- I learned something! =-)
Thanks,
Wolf
i get this very strange feeling !
Posted By: vasyachkin
Date:
In Response To: Re: More detail.... (Jeff B.)
> I see that John k. has replied too, and my
> reply is essentially the same. On a dipole
> the lobes are on the front and rear axis and
> the nulls are 90 degrees to the sides where
> cancellation occurs. On a bipole the
> combined lobes are the opposite and are at
> the sides; 90 degrees off-axis. Of course,
> this is also dependent on the radiation
> pattern of the driver. If very little high
> frequency energy is radiated at 90 degrees
> off-axis then they will not be a lot of
> energy to sum for the lobe. Unfortunately,
> with a bipole there will also be nulls on
> the front and rear axis at certain
> frequencies as delayed acoustical energy
> from the other bank of drivers makes it way
> around to join the on-axis sound. This delay
> will create out of phase conditions at
> certain frequencies. The frequencies this
> occurs at will be determined by the physical
> separation of the drivers and the width and
> depth of the enclosure. In other words, it
> is determined by how long it takes sound to
> wrap around the speaker and merge with the
> other axis. Again, driver radiation patterns
> play a role in the magnitude of these
> cancellations too.
> I have built speakers like Pete is
> describing, and I didn’t notice anything
> other than the increased spaciousness of the
> speaker. One way to reduce the spaciousness
> and the occurrence of nulls would be to
> design it as a 2.5way where the rear drivers
> are rolled-off above the baffle step
> frequency. This way they will compensate for
> the baffle step and contribute to the low
> end response without changing the on-axis
> response in the upper range.
> Jeff B.
that you all left madisound to secretly implement my ideas !
;)
http://www.madisound.com/forum/dc/dcboard.php?az=show_mesg&forum=2&topic_id=2159&mesg_id=2161&page=
in the above post i describe everything you just said as well as make some other points
Re: Bipole/Dipole....
Posted By: Roman Bednarek <roman@twcny.rr.com>
Date:
In Response To: Bipole/Dipole.... (Wolf)
> While I do agree with what you stated, Pete
> said Bipole, so both arrays are in outward
> motion at the same time. Dipole is where
> they are in opposed motions, or opposing
> polarities.
> Later,
> Wolf
I should have been a little clearer, but a dipole design will be fine 90 degrees off axis horizontally because the drivers are in phase. However, if you are 45 degrees off axis for example, the distance between you and the front speaker will be less than the difference between you and the rear speaker and if the differential between these distances is equal to half a wavelength of a particular frequency they will cancel completely at that frequency. So you'll basically get some sort of comb filtering effect.
ALMOST right
Posted By: vasyachkin
Date:
In Response To: Re: Bipole/Dipole.... (Roman Bednarek)
> I should have been a little clearer, but a
> dipole design will be fine 90 degrees off
> axis horizontally because the drivers are in
> phase. However, if you are 45 degrees off
> axis for example, the distance between you
> and the front speaker will be less than the
> difference between you and the rear speaker
> and if the differential between these
> distances is equal to half a wavelength of a
> particular frequency they will cancel
> completely at that frequency. So you'll
> basically get some sort of comb filtering
> effect.
you are ALMOST looking at it right.
the wave from the other side that will be combining with the front wave to cause an interference pattern will actually be the DIFFRACTED wave from the other side.
there is also a diffracted wave from the FRONT side which is OPPOSITE in phase to the diffracted wave from the back.
the two waves can potentially cancel each other COMPLETELY so instead of creating additional interference you would eliminate the diffraction nulls you already have.
this would only happen for a cabinet of zero depth however.
to summarize there are really 4 waves to be concerned with
1 front direct
2 front difracted
3 rear direct
4 rear diffracted
and out of these on-axis you will get a combination of 1, 2 and 4 (you will not get anything from 3). if done properly 2 and 4 will cancel each other and you will get a perfect 1 (without any baffle step and without any diffraction effects and without any nulls, nothing just perfect flat response)
of course you will not be able to get there because as i said it would require zero depth speaker but you can get halfway there by contouring the cabinet shape.
this is NOT too good to be true because all it really accomplishes is a speaker built into an infinite baffle except that this infinite baffle is virtual.
Re: Bipolar operation
Posted By: Æ
Date:
In Response To: Bipolar operation (Pete Schumacher ©)
> Say you build an MTM for use out in the
> room, and full baffle step compensation is
> needed. Rather than contour the XO to
> accomplish this, design for flat and add a
> pair of woofers onto the rear of the
> enclosure. Baffle step will be taken care
> of.
Isn't that something like Linkwitz's ORIONS?
Re: Bipolar operation
Posted By: Pete Schumacher © <pete_schumacher@yahoo.com>
Date:
In Response To: Re: Bipolar operation (Æ)
> Isn't that something like Linkwitz's ORIONS?
Slightly different. The Orion's are DIPOLE.
Re: Bipolar operation
Posted By: John k...
Date:
In Response To: Bipolar operation (Pete Schumacher ©)
> Say you build an MTM for use out in the
> room, and full baffle step compensation is
> needed. Rather than contour the XO to
> accomplish this, design for flat and add a
> pair of woofers onto the rear of the
> enclosure. Baffle step will be taken care
> of.
> However, now you have rear drivers that will
> also be filling the room with sound, just
> like the front drivers. Could something like
> that also be considered an omni?
Below the baffle step it will be omni. It's omni with just one driber below the baffle step too. Above the baffle step it depends on the distance between the sources and the off axis radiation pattern of the sources. But if the sourece remain fairly wide angled radiators the polsr response will develope a lobing pattern with peans and nulls on axis and side lobes at 90 degrees (as opposed to the null of a dipole). The number of lobes in the polar responser between 0 and 90 degrees will depend on the distance between the sources and the frequency.
Above baffle step
Posted By: Pete Schumacher <pete_schumacher@yahoo.com>
Date:
In Response To: Re: Bipolar operation (John k...)
> Below the baffle step it will be omni. It's
> omni with just one driber below the baffle
> step too. Above the baffle step it depends
> on the distance between the sources and the
> off axis radiation pattern of the sources.
> But if the sourece remain fairly wide angled
> radiators the polsr response will develope a
> lobing pattern with peans and nulls on axis
> and side lobes at 90 degrees (as opposed to
> the null of a dipole). The number of lobes
> in the polar responser between 0 and 90
> degrees will depend on the distance between
> the sources and the frequency.
Above baffle step, the drivers are much more directional. The rear drivers would not be interfering with the front drivers above baffle step, and thus wouldn't be creating peaks and nulls on-axis.
Where I could see that happening is in the transition region and below, where the drivers certainly are omni.
Re: Above baffle step
Posted By: J Kim <kim_woojae@hotmail.com>
Date:
In Response To: Above baffle step (Pete Schumacher)
> Above baffle step, the drivers are much more
> directional. The rear drivers would not be
> interfering with the front drivers above
> baffle step, and thus wouldn't be creating
> peaks and nulls on-axis.
> Where I could see that happening is in the
> transition region and below, where the
> drivers certainly are omni.
Essentially, you and john talk about the same thing. The lobes will occur in the transition region---a certain frequency band that is low enough for the waves to be omni and high enough for them to be phase-interfered. Yes, this will be affected by the baffle step point, distance between two sources, and directivity of drivers.
-jAy
Re: Above baffle step
Posted By: John k...
Date:
In Response To: Above baffle step (Pete Schumacher)
> Above baffle step, the drivers are much more
> directional. The rear drivers would not be
> interfering with the front drivers above
> baffle step, and thus wouldn't be creating
> peaks and nulls on-axis.
> Where I could see that happening is in the
> transition region and below, where the
> drivers certainly are omni.
Basically, what ever interference happens to the front from the rear depends on the 90 degree off axis response of the woofer. The severity will be driver dependent. A smaller diameter driver on a wider baffle could have significant effects. Big difference between talking about what can happen and the specifics of a particular case. It is important to understand what can happen and then determine to what degree it does happen.
Re: Related Question
Posted By: Bob Blickle <bblickle@mindspring.com>
Date:
In Response To: Bipolar operation (Pete Schumacher ©)
While on the subject, what about Sensitivity?
If a rear driver is significantly more than a wavelength from the front, it seems like they wouldn't "couple" to increase output.
To help me better grasp this, let's consider the extreme case: How about front and rear facing super tweeters that easily might be 10 wavelengths apart.
If they are wired parallel but in fact do not "couple" does that mean the net sensitivity change of adding the second tweeter is about +3db (rather than +6 if front mounted)?
If they are wired serial but do not "couple" does that mean the net sensitivity change is -3db (rather than 0dB if front mounted)?
bb
Ambient sound
Posted By: Pete Schumacher © <pete_schumacher@yahoo.com>
Date:
In Response To: Re: Related Question (Bob Blickle)
> While on the subject, what about
> Sensitivity?
> If a rear driver is significantly more than
> a wavelength from the front, it seems like
> they wouldn't "couple" to increase
> output.
> To help me better grasp this, let's consider
> the extreme case: How about front and rear
> facing super tweeters that easily might be
> 10 wavelengths apart.
> If they are wired parallel but in fact do
> not "couple" does that mean the
> net sensitivity change of adding the second
> tweeter is about +3db (rather than +6 if
> front mounted)?
> If they are wired serial but do not
> "couple" does that mean the net
> sensitivity change is -3db (rather than 0dB
> if front mounted)?
> bb
Since the tweeters would be effectively well above baffle step, they wouldn't couple at all, with each sending it's radiation away from the other.
On the other hand, woofers on the rear, say 12" from the front woofers, would directly couple as if they were all mounted on the front, up to a certain frequency, where wavelength and separation kicks in to cause amplitude addition anomalies. But that should allow good summation up into the multiple 100s of Hz.
Re: Ambient sound
Posted By: Bob Blickle <bblickle@mindspring.com>
Date:
In Response To: Ambient sound (Pete Schumacher ©)
Pete,
Back to the tweeters for a min, under the line of thinking I was pursuing, a series rear-mounted tweeter seems like a more attractive option for padding by -3dB than adding a resistor does, particularly if we're talking about something low cost like an ND.
Dipole baffle width and cancellation frequencies?
Dipole baffle width and cancellation frequencies?
bb
Dipole baffle width and cancellation frequencies?
Posted By: BrianP <caspian@peak.org>
Date:
The forward output of a dipole-mounted driver will roll off at 6dB/octave with decreasing frequency, as wavelengths exceed the baffle width and the out-of-phase rear output "wraps around" the baffle and phase cancellation occurs. What is the formula to predict the -3dB, -6dB, and -12dB points for a given baffle width, assuming the driver is centered horizontally on the baffle?
Actually . . .
Posted By: Pete Schumacher © <pete_schumacher@yahoo.com>
Date:
In Response To: Dipole baffle width and cancellation frequencies? (BrianP)
> The forward output of a dipole-mounted
> driver will roll off at 6dB/octave with
> decreasing frequency, as wavelengths exceed
> the baffle width and the out-of-phase rear
> output "wraps around" the baffle
> and phase cancellation occurs. What is the
> formula to predict the -3dB, -6dB, and -12dB
> points for a given baffle width, assuming
> the driver is centered horizontally on the
> baffle?
It's 24dB/octave asymptote at the lowest frequencies, just like any open box system, TL or HELMHOLZ. The rear wave cancels the front and you get a 4th order slope below cutoff.
Re: Actually . . .
Posted By: Roman Bednarek <roman@twcny.rr.com>
Date:
In Response To: Actually . . . (Pete Schumacher ©)
> It's 24dB/octave asymptote at the lowest
> frequencies, just like any open box system,
> TL or HELMHOLZ. The rear wave cancels the
> front and you get a 4th order slope below
> cutoff.
I'm a newbie in this realm but Jeff B's Response Modeler indicates a 6dB/octave rolloff.
Re: Actually . . .
Posted By: Pete Schumacher © <pete_schumacher@yahoo.com>
Date:
In Response To: Re: Actually . . . (Roman Bednarek)
> I'm a newbie in this realm but Jeff B's
> Response Modeler indicates a 6dB/octave
> rolloff.
I think that's for the area around baffle step. The final asymptote is 4th order for an open baffle woofer.
Parts Express Tech Talk
Re: Actually . . .
Posted By: Christian B <mazeroth@gmail.com>
Date:
In Response To: Actually . . . (Pete Schumacher ©)
It's actually 6 dB/octave. I would link to linkwitzlab.com but his site is down due to exceeding bandwidth (doh!). However, John K's site is up. Don't let his math scare you as you can design a great dipole without using too much math:
http://www.musicanddesign.com/Equivalent_Baffles.html
Re: Dipole baffle width and cancellation frequenci
Posted By: Roman Bednarek <roman@twcny.rr.com>
Date:
In Response To: Dipole baffle width and cancellation frequencies? (BrianP)
> The forward output of a dipole-mounted
> driver will roll off at 6dB/octave with
> decreasing frequency, as wavelengths exceed
> the baffle width and the out-of-phase rear
> output "wraps around" the baffle
> and phase cancellation occurs. What is the
> formula to predict the -3dB, -6dB, and -12dB
> points for a given baffle width, assuming
> the driver is centered horizontally on the
> baffle?
I don't know the formula but I know what works really well and that is Jeff B's Response Modeler Excel program. The baffle diffraction response modeler at the bottom gives the exact reponse. Your question is sort of hard to answer also because there is a ~2dB peak before the roll off kicks in.
I'm in the same boat right now because I'm beginning a dipole project of my own and have some decisions to make on the baffle design.
Re: Dipole baffle width and cancellation frequenci
Posted By: Roman Bednarek <roman@twcny.rr.com>
Date:
In Response To: Dipole baffle width and cancellation frequencies? (BrianP)
> The forward output of a dipole-mounted
> driver will roll off at 6dB/octave with
> decreasing frequency, as wavelengths exceed
> the baffle width and the out-of-phase rear
> output "wraps around" the baffle
> and phase cancellation occurs. What is the
> formula to predict the -3dB, -6dB, and -12dB
> points for a given baffle width, assuming
> the driver is centered horizontally on the
> baffle?
I don't know the formula but I know what works really well and that is Jeff B's Response Modeler Excel program. The baffle diffraction response modeler at the bottom gives the exact reponse. Your question is sort of hard to answer also because there is a ~2dB peak before the roll off kicks in.
I'm in the same boat right now because I'm beginning a dipole project of my own and have some decisions to make on the baffle design.
Re: Now I'm confused . . .
Posted By: max andrews <madmax@Nyu.edu>
Date:
In Response To: Now I'm confused . . . (Pete Schumacher ©)
Maybe it's 4th including the XO Pete?
Also, isn't the cutoff frequency the total distance from the front to the back of the woofer? So a 6" driver on a 12" wide by .75" thick open baffle with no "wings" would have a total distance of 3"+3"+.75"= 6.75" = 1900hz? Or it is half that at 950hz?
Re: Now I'm confused . . .
Posted By: Ted Wilt <ted4412wilt@msn.com>
Date:
In Response To: Re: Now I'm confused . . . (max andrews)
> Also, isn't the cutoff frequency the total
> distance from the front to the back of the
> woofer? So a 6" driver on a 12"
> wide by .75" thick open baffle with no
> "wings" would have a total
> distance of 3"+3"+.75"=
> 6.75" = 1900hz? Or it is half that at
> 950hz?
Actually I believe this is a complex subject since all of the sound from the rear is not going to follow one simple path; however, I think you would take the difference between the shortest path from the cone rear to the ear and the path from the cone front to the ear. If that difference is half wave length you get double output. Quarter wave length gives full output. If there is no difference you get complete cancellation.
I placed a 12" car woofer on the floor propped against a foam mattress. At 1000Hz it was slightly louder than my AR3. At 500Hz down 6dB, 250Hz down 12dB, 125 Hz down 18dB, 64 Hz slightly audible, 32 Hz cone moved but no sound. All those are rough guesses but I thought I had a bad driver. It is rated at 91.2dB/1 watt with 10mm xmax. I then taped it to an 8' length of 11.5" Quik-Tube leaned up against a corner at 45 degree and bisecting the corner angle ending 6" down from the corner so it is like loading a horn. I have to keep my bass amplifier turned way down or it shakes the upstairs floor. The thin walls of the Quik-Tube vibrate too much at about 62-70Hz producing huge distortion so I need to reinforce the tube quite a bit or add an outer tube an inch larger and fill with sand etc. I am also going to experiment extending another 4' and then add a 4' length on the other side of the driver to make a Bose Cannon. My chief limitation is the $33 woofer with 35Hz Fs; however the 130 oz magnet appears to give a lot of power for the money.
http://www.partsexpress.com/pe/showdetl.cfm?&Partnumber=293-483
3rd Order?
Posted By: Pete Schumacher <pete_schumacher@yahoo.com>
Date:
In Response To: Re: Dipole baffle width and cancellation frequenci (Roman Bednarek)
> I don't know the formula but I know what
> works really well and that is Jeff B's
> Response Modeler Excel program. The baffle
> diffraction response modeler at the bottom
> gives the exact reponse. Your question is
> sort of hard to answer also because there is
> a ~2dB peak before the roll off kicks in.
> I'm in the same boat right now because I'm
> beginning a dipole project of my own and
> have some decisions to make on the baffle
> design.
So, what this is saying is that
Re: 3rd Order?
Posted By: Roman Bednarek <roman@twcny.rr.com>
Date:
In Response To: 3rd Order? (Pete Schumacher)
> So, what this is saying is that
> will have 3rd order roll-off when the
> natural response of the woofer is added in?
I would presume so because that is how I modeled it (combining the baffle response with the infinite baffle measurement). It isn't an issue for me though because I will be crossing at 200 Hz with 4th order slopes and the XG18 doesn't start rolling off until below that point.
John did a pretty good job of explaining things below for the original poster as I hoped he would (thanks John). My next project (dipole mid/vented bass) should be fun and a great learning experience for me. I'm actually going to be recycling the bass module from my Daedalus design by adding a dipole top using the XG18 and RS28a (It's a bit too cold to use the table saw outside right now and the dipole top will be simple enough to use a circular saw in the basement). I'm pretty excited about this project but I'm sure you'll see some questions in the future regarding issues that I am not experienced with.
Re: Dipole baffle width and cancellation frequenci
Posted By: John k...
Date:
In Response To: Dipole baffle width and cancellation frequencies? (BrianP)
> The forward output of a dipole-mounted
> driver will roll off at 6dB/octave with
> decreasing frequency, as wavelengths exceed
> the baffle width and the out-of-phase rear
> output "wraps around" the baffle
> and phase cancellation occurs. What is the
> formula to predict the -3dB, -6dB, and -12dB
> points for a given baffle width, assuming
> the driver is centered horizontally on the
> baffle?
The relationship is a little complicated. First of all, at lower frequencies it can be shown that an arbitrary shaped baffle is equivalent to a circular baffle of some effective radius, even if the drive ris mounted off cennter. See http://www.musicanddesign.com/Equivalent_Baffles.html for discussion of a simple rectangular baffle. Once you know the equivalent radius then on axis the phase difference between front and rear is gives as:
Phi = 180 + 360 x (R/C) x F where R is the equivalent radius, C is the speed of sound and F is frequency.
Then the amplitude on axis is given as A = 2 sin(Phi/2) x cos(Phi/2).
This give a 6dB/octave roll off with a zero dB point (dipole = monopole) at R/C x F = 1/6, or Feq = C/(6R). Now that you know the 0 dB point you can shift the frequency an ovate lower at a time to get 6dB/octave to get the - 6, - 12, -18 dB, etc frequencies. F(-6) = Feq/2, F(-12)= Feq/4, F(-18) = Feq/8.....
This is the basic dipole roll off which must then be superimposed on the driver's native response.
Re: ... just when I thought I was understanding it
Posted By: Undefinition <Undefinition@aol.com>
Date:
In Response To: Re: Dipole baffle width and cancellation frequenci (John k...)
John, can I be excused from class to go to the nurse? My brain hurts!
Another reference that covers this topic well
Posted By: manzanita <jbusch@sti.net>
Date:
In Response To: Re: I love it when you talk like that... (Jeff B.)
Re: about the trigonometry...
Posted By: Undefinition <Undefinition@aol.com>
Date:
In Response To: Re: Dipole baffle width and cancellation frequenci (John k...)
Surprisingly, once I got over seeing so much calculation, I started to understand the methodology (or at least how to use it for my own benefit). HOWEVER, what I want to know, John, is how did the trig functions become a part of this?
Sorry if that's a stupid question, but I ask it because learning the calculations behind speakers has really enhanced my understanding of electronics (and sound, obviously)--something I've wanted to get, but grappled with for probably 15 years. Anyway, up until now, all the formulas I've used in regards to speaker design involved mostly algebra... but not trigonometry. In no way am I doubting your ability or knowledge, but I am just curious where you picked up these sort of equations. (I do admire your work, BTW. It was Dan Neubecker's version of the NaO minis that made me want to go down the dipole path... and now it's hard to turn back!)
-Paul
periodic waveforms
Posted By: Pete Schumacher © <pete_schumacher@yahoo.com>
Date:
In Response To: Re: about the trigonometry... (Undefinition)
> Surprisingly, once I got over seeing so much
> calculation, I started to understand the
> methodology (or at least how to use it for
> my own benefit). HOWEVER, what I want to
> know, John, is how did the trig functions
> become a part of this?
> Sorry if that's a stupid question, but I ask
> it because learning the calculations behind
> speakers has really enhanced my
> understanding of electronics (and sound,
> obviously)--something I've wanted to get,
> but grappled with for probably 15 years.
> Anyway, up until now, all the formulas I've
> used in regards to speaker design involved
> mostly algebra... but not trigonometry. In
> no way am I doubting your ability or
> knowledge, but I am just curious where you
> picked up these sort of equations. (I do
> admire your work, BTW. It was Dan
> Neubecker's version of the NaO minis that
> made me want to go down the dipole path...
> and now it's hard to turn back!)
> -Paul
The trig is there because of the relationship of the physical size of the baffle to the wavelength of the signal. The trig function could be re-written in algebraic format. It's just simpler to use the trig functions.
With periodic signals, you can express them as vectors that describe a circle. When you have a circle, you can use polar or rectangular coordinates. The math can just become much simpler by using polar instead.
Re: about the trigonometry...
Posted By: John k...
Date:
In Response To: Re: about the trigonometry... (Undefinition)
> Surprisingly, once I got over seeing so much
> calculation, I started to understand the
> methodology (or at least how to use it for
> my own benefit). HOWEVER, what I want to
> know, John, is how did the trig functions
> become a part of this?
> Sorry if that's a stupid question, but I ask
> it because learning the calculations behind
> speakers has really enhanced my
> understanding of electronics (and sound,
> obviously)--something I've wanted to get,
> but grappled with for probably 15 years.
> Anyway, up until now, all the formulas I've
> used in regards to speaker design involved
> mostly algebra... but not trigonometry. In
> no way am I doubting your ability or
> knowledge, but I am just curious where you
> picked up these sort of equations. (I do
> admire your work, BTW. It was Dan
> Neubecker's version of the NaO minis that
> made me want to go down the dipole path...
> and now it's hard to turn back!)
> -Paul
As was stated below, we start with the front and rear radiation as
Front = sin(wt), Rear = sin(wt + Phi). The the sum of two sine functions of equal strength is a simple trig idenity. Actually I should have written the result as
A = 2 sin(wt+Phi/2) x cos(-Phi/2),
Many thanks, all!
Posted By: BrianP <caspian@peak.org>
Date:
In Response To: Dipole baffle width and cancellation frequencies? (BrianP)
And especially John K, for giving me some useful formulae to crunch some numbers in.
I will download Jeff's calculator and play with it. Software like that is wonderful, and predicts far better results than I could, working with textbook equations and graph paper. Still, I have this old-fashioned desire to understand HOW and WHY things work the way they do.
I knew this would be a lot more complicated than baffle diffraction step on a conventional box. One complication is that the rear wave output of a dynamic driver (as opposed to maybe a planar) will NOT be identical to its forward output. The chassis will bollox things up, as will the cutout in the baffle if it has any thickness at all.
Another issue with dipole response is the contribution of room reflections to the sound heard at the listening position. This is also an issue with box speakers, but the custom there is to use quasi-anechoic measurements and modelding, and pretend the room doesn't exist. Can't do that with a dipole--the room is an essential element in the "spacious" sound that dipole lovers love.
So I suppose the best approach is the empirical one: measure the response of the actual mid drivers in their actual baffles, in their intended positions in the room, from the intended listening position. Plot their LF rolloff, and calculate a crossover to the sealed-box woofer from there.
"Is anyone familiar with this concept"
| | Hi Everyone - Hi Everyone - |
In response to Reply # 0
| | the problem though is that your speaker would not be any more omnidirectional than any normal speaker. just think about it ... |
| |
|
| Wed Feb-20-08 10:33 AM | |
#8. "Spacial Localization (the left and right of sound)"
In response to Reply # 3
Wed Feb-20-08 10:41 AM by smokinjimmyv
| | Think of directional frequencies as the 'high' ones. Low frequencies have very very little directionality. |
| Sat Feb-09-08 05:41 AM | |
#2. "RE: Is anyone familiar with this concept"
In response to Reply # 0
Sat Feb-09-08 05:44 AM by Andy_G
| | http://www.linkwitzlab.com/pluto-wa_2.htm |
|
#5. "RE: Is anyone familiar with this concept"
In response to Reply # 2
| >http://www.linkwitzlab.com/pluto-wa_2.htm |
#9. "RE: Is anyone familiar with this concept"
In response to Reply # 2
| | ||||
| Member since | ||||
#6. "Yes . . . check these"
In response to Reply # 0
GR research has a couple of kits of this description, the A/V-O and the O-3:
http://www.gr-research.com/index.asp?PageAction=VIEWCATS&Category=2
Samadhi acoustics offers the "Magic Cube" and the "Ichiban" which are rather pricey assembled systems, apparently no longer available as kits:
http://www.samadhiacoustics.com/
Other posters have mentioned the Linkwitz Plutos, which are an active bi-amplified system. Apparently they sound wonderful, but WAF is low, as they have a face only a plumber could love.
I have built a couple of designs along these lines with inexpensive drivers, and they sound WAY better than they have any right to. The main advantage is that they cast a wide, deep, layered soundstage when placed well out into the room with plenty of air around them.
The advantage of this configuration is that you extend the omnidirectional response pattern much higher in frequency than is possible with a forward-firing design, but gradually narrow to a directional pattern at the highest frequencies, where imaging cues reside. Thus imaging is more precise than with a design that radiates omnidirectionally all the way up, but you still get the big soundstage.
The approach works best with smaller midbass drivers (4" to 5") because you're using the 90 degree off-axis response, and this needs to extend flat high enough to cross to a tweeter at a reasonable frequency. The rolloff of the midbass is quite steep at this angle, once the response starts to drop, so you can get by with a shallow electrical crossover slope but may need to incorporate some response shaping.
Dan Neubecker has built a couple of MTM systems on this principle, and has done some excellent technical analysis of how they work. Look for discussion of his HOSS and SoundRounds on the PE board.
| | ||
| Sat Feb-09-08 12:27 PM | ||
#7. "And here are a couple of reviews"
In response to Reply # 6
| http://www.blackdahlia.com/samadhi/ichibanrel.htm |
#10. "RE: Yes . . . check these"
In response to Reply # 6
| |
Jump to top of page
