al2002 said:
Of course there will be cancellation! How else will you get the figure 8 radiation? The tuned boxes will still go JUST AS LOW to the front and back as as a monopole! There will be a deep notch at the path length distance and much lower radiation off to the sides of the cabinets like all other dipoles = People often look at theory rather than practice - This is one reason why I don't contribute to these websites very often! Go build a Linkwitz (distortion box) baffle with eq - then compare it to this, when you do this please report back - bye for now😉
johninCR said:
John,
there isn´t something like a velocity wave or a pressure wave. Every wave has alternating regions of maximal pressure and regions of maximal velocity. And both are radiating from the driver with the speed of sound. A dipole is just called a velocity TRANSDUCER because it best feeds energy into a wave at the waves velocity maximum. A compression transducer feeds energy best at the waves pressure maximum.
I too had a hard time to get this straigt in my mind. 😉
Rudolph,
Thanks for trying on the velocity vs pressure thing. I'll build a box and try the isobarik layout vs 2 drivers in a W baffle. Hopefully the drivers will behave differently enough and/or exciting the wave at its pressure max instead of velocity maximum will result in more or even equal efficiency with the same number of drivers. I know the increase in "d" will be a benefit without having to go the U baffle route or a very deep H alignment.
I do know that I'm not going the double box route, except maybe for the bottom octave for HT which is very intriguing. High impact in the teens at the listening position that nets mostly to zero outside the room is very interesting indeed. My front and back walls are concrete, so keeping it in room is a real possiblity and I have 2 tempests designated for sub 28hz duty that I've been debating over the best way to implement.
Thanks for trying on the velocity vs pressure thing. I'll build a box and try the isobarik layout vs 2 drivers in a W baffle. Hopefully the drivers will behave differently enough and/or exciting the wave at its pressure max instead of velocity maximum will result in more or even equal efficiency with the same number of drivers. I know the increase in "d" will be a benefit without having to go the U baffle route or a very deep H alignment.
I do know that I'm not going the double box route, except maybe for the bottom octave for HT which is very intriguing. High impact in the teens at the listening position that nets mostly to zero outside the room is very interesting indeed. My front and back walls are concrete, so keeping it in room is a real possiblity and I have 2 tempests designated for sub 28hz duty that I've been debating over the best way to implement.
Incredibly brilliant scientists like Magnetar often have difficulties explaining things, which is why they don't/can't contribute to these websites very often. Filled with too many idiots like myself. Their theories are so complex, words in the english language cannot do them justice. Unfortunately, the laypersons like myself have to rely on the published (for all to see) data on websites like that nimwit Siegrfried Linkwitz. I actually fell into the trap of building one of his distortion generators. I feel so foolish now. I'll junk them as soon as Magnetar publishes his many AES papers that refute/embarass fools like SL.
His Magnapole woofer system must sound incredible. It probably uses a Magnetar-Riley crossover. A Magnetar-Transform possibly enhances low end extension. The baffle shape probably looks like the Magnetar Artistry Beethoven speakers (which received rave reviews).
I can't wait to see all the measurements. Distortion. FR. Data, data, data. What methods for used to get them. The scope of the science will be so far above my head it will hurt 😱 .
But I still want to see it. Please. The world anxiously awaits your genius Magnetar. Please don't keep us waiting forever. I've got music to listen to. All I'm hearing is distortion.
Cheers,
AJ
p.s. - could you give us your real name so I can start researching your vast number of AES papers? I only know the last three letters are PHD. Can you clue us on the rest?
His Magnapole woofer system must sound incredible. It probably uses a Magnetar-Riley crossover. A Magnetar-Transform possibly enhances low end extension. The baffle shape probably looks like the Magnetar Artistry Beethoven speakers (which received rave reviews).
I can't wait to see all the measurements. Distortion. FR. Data, data, data. What methods for used to get them. The scope of the science will be so far above my head it will hurt 😱 .
But I still want to see it. Please. The world anxiously awaits your genius Magnetar. Please don't keep us waiting forever. I've got music to listen to. All I'm hearing is distortion.
Cheers,
AJ
p.s. - could you give us your real name so I can start researching your vast number of AES papers? I only know the last three letters are PHD. Can you clue us on the rest?
Magnetar said:
NO! The same cancellation that produces the dipole pattern will also result in a LF roll off needing eq, just like an open baffle dipole. There will not be just a notch in the FR; the cancellation will result in a LF roll off. For purposes of analysing the back-to-back box mounted dipole you will still need a path length difference and use the f3 and Q of the box mounted drivers instead of free air fs and Qts for an open baffle dipole.
Magnetar said:
Not sure what you mean by this remark. The theory of radiation from dipole sources - both acoustic and electromagnetic - is on sound theoretical footing. If you don't want to have obvious errors pointed out, then, indeed, you should not post often.
Magnetar said:
So are you are implying that you have built both: a closed box dipole and a Linkwitz dipole of similar physical size, and have a set of comphrensive, VALID, comparative measurements under anecohic (or anechoic-like) conditions?
I doubt very much you have done this. But just in case you have, please post your data so we can settle this once and for all.
by al2002,
Not sure what you mean by this remark.
`````````````````````````````````````````````````````````````````````
what magnetar is saying is that the results of practice are often different than what theory
would suggest
Not sure what you mean by this remark.
`````````````````````````````````````````````````````````````````````
what magnetar is saying is that the results of practice are often different than what theory
would suggest
I'm assuming he was reffering to 2 bass reflex speakers (with a well chosen alignment) designed exactly the same, one wired out-of-phase and physically placed behind the other.
At low freq.s dipole cancelation is the same (just look at the length of the wavelength to see if there will be cancelation) when comparing dipole baffle size or Mag's. dipole box placement.
Whats different is that you have gained the benefits (and disadvanteges) of a good bass reflex design vs. that of "free air" operation. (at the expense of added box complexity and size).
Lets look at the acoustic disadvatage 1st:
1. Goup Delay is significantly worse.
2. Low freq. extension is limited.
The acoustic advantages are:
1. The alignment allows flexiblity with the design. Create a port resonance higher in freq. and you can effectly compensate for some (even all) of the sp loss of dipole operation for a given low freq. cut-off (i.e. the more compensation the less extension).
2. The port's spl contribution typically reduces harmonic distortion in its bandwidth by providing a secondary source of spl. Additionally, because the port only works within a limited bandwidth and because a sealed alignment has a sp loss of 12 db an octave (as opposed to a dipole's 6 db), harmonic distortion is further decreased (due to a decrease in excursion) because the spl at very low freq.s is more attenuated than what a traditional dipole provides.
3. You have the ability to move the two (out-of-phase) boxes to varying locations (depending on the upper freq. limit of operation) to achieve a more uniform in room freq. response.
At low freq.s dipole cancelation is the same (just look at the length of the wavelength to see if there will be cancelation) when comparing dipole baffle size or Mag's. dipole box placement.
Whats different is that you have gained the benefits (and disadvanteges) of a good bass reflex design vs. that of "free air" operation. (at the expense of added box complexity and size).
Lets look at the acoustic disadvatage 1st:
1. Goup Delay is significantly worse.
2. Low freq. extension is limited.
The acoustic advantages are:
1. The alignment allows flexiblity with the design. Create a port resonance higher in freq. and you can effectly compensate for some (even all) of the sp loss of dipole operation for a given low freq. cut-off (i.e. the more compensation the less extension).
2. The port's spl contribution typically reduces harmonic distortion in its bandwidth by providing a secondary source of spl. Additionally, because the port only works within a limited bandwidth and because a sealed alignment has a sp loss of 12 db an octave (as opposed to a dipole's 6 db), harmonic distortion is further decreased (due to a decrease in excursion) because the spl at very low freq.s is more attenuated than what a traditional dipole provides.
3. You have the ability to move the two (out-of-phase) boxes to varying locations (depending on the upper freq. limit of operation) to achieve a more uniform in room freq. response.
AJ,
I think that will only ensure Magnetar doesn't return.
Nuuk,
The only thing I really followed was that Magnetar said he tried both with and without the divider in the bass section with better than dipole results. However, I believe most of what he is talking about is with 2 sealed boxes. I don't have interest in sealed except maybe for an HT sub and a several meter "d", or ported for that matter for the extreme bottom end. The notch null he mentioned I guess must be at wavelength/2 = "d" , which at d=30" I'm getting close to my 200hz limit.
Tomtt,
I take it to mean that problems in theory are often exaggerated compared to real world results. A good example for me is driver spacing for line arrays. With the same 4" drivers I've done arrays with 4.5", 5", and 5.5" CTC spacing with no noticeable comb filtering effects and I run the drivers full range. All I notice is some attentuation of the top end compared to a single driver which falls off above 10khz. This coincides with what I hear at concerts (attentuated highs).
Al,
I think we can all agree that the dipole roll off point will still exist and is dependant upon "d". I am intrigued by how a loose isobarik coupling will affect cone movement and system Q. Also, does that alignment stimulate the waves differently and have the potential for greater sensitivity?
Rudolph or ScottG,
If you can step in and discuss the last point above, please do so.
I think that will only ensure Magnetar doesn't return.
Nuuk,
The only thing I really followed was that Magnetar said he tried both with and without the divider in the bass section with better than dipole results. However, I believe most of what he is talking about is with 2 sealed boxes. I don't have interest in sealed except maybe for an HT sub and a several meter "d", or ported for that matter for the extreme bottom end. The notch null he mentioned I guess must be at wavelength/2 = "d" , which at d=30" I'm getting close to my 200hz limit.
Tomtt,
I take it to mean that problems in theory are often exaggerated compared to real world results. A good example for me is driver spacing for line arrays. With the same 4" drivers I've done arrays with 4.5", 5", and 5.5" CTC spacing with no noticeable comb filtering effects and I run the drivers full range. All I notice is some attentuation of the top end compared to a single driver which falls off above 10khz. This coincides with what I hear at concerts (attentuated highs).
Al,
I think we can all agree that the dipole roll off point will still exist and is dependant upon "d". I am intrigued by how a loose isobarik coupling will affect cone movement and system Q. Also, does that alignment stimulate the waves differently and have the potential for greater sensitivity?
Rudolph or ScottG,
If you can step in and discuss the last point above, please do so.
@johninCR
I do know that magnetar has an impressive history as a loudspeaker system builder and does not settle for low goals.
So I would really appriciate if you try one of his proposals and report here. While the isobarik layout would be easier to build, the double box route should show even better, whether any advantages of that CB design can be made available in a dipole.
Keep us informed.
Rudolf
I do know that magnetar has an impressive history as a loudspeaker system builder and does not settle for low goals.
So I would really appriciate if you try one of his proposals and report here. While the isobarik layout would be easier to build, the double box route should show even better, whether any advantages of that CB design can be made available in a dipole.
Keep us informed.
Rudolf
Am I correct in thinking that you would need to decide whether to go for open-baffle or sealed box before you choose the drivers? 
tomtt said:
That can happen. However, in this instance no data has ben presented, so any further discussion of theory vs. practice is unlikely to be productive.
Rudolph,
I don't doubt at all that sealed will give us more output down to a certain point. SL talks about the point where dipole = sealed SPL, and it is much higher than the point where the dipole bass roll off starts. I'm really not interested in fighting a sealed roll-off starting in the 60hz range and I have no interest in bipole bass. If my drivers fit in the one set of sealed boxes I have then I'll be able to do a 3 way comparison since I have plenty of the same driver. Of course I'll report back results.
Nuuk,
Higher Q drivers make bass more easily on OB, but they require larger boxes if you go sealed. Lower Q drivers have to be forced into making OB bass because their cone movement is more tightly controlled by their motor, but for sealed the boxes can be smaller.
I don't doubt at all that sealed will give us more output down to a certain point. SL talks about the point where dipole = sealed SPL, and it is much higher than the point where the dipole bass roll off starts. I'm really not interested in fighting a sealed roll-off starting in the 60hz range and I have no interest in bipole bass. If my drivers fit in the one set of sealed boxes I have then I'll be able to do a 3 way comparison since I have plenty of the same driver. Of course I'll report back results.
Nuuk,
Higher Q drivers make bass more easily on OB, but they require larger boxes if you go sealed. Lower Q drivers have to be forced into making OB bass because their cone movement is more tightly controlled by their motor, but for sealed the boxes can be smaller.
johninCR said:
John,
I wasn´t referring to simple sealed boxes but to a dipole simulated by two sealed boxes spaced "d" apart and wired 180° out of phase. This should - in theory - work like ONE driver in an OB with a width of "2d". I could imagine that this configuration, done with bass reflex instead of sealed as assumed by ScottG, might raise efficiency.
BassAwdyO did already propose that somehow here: http://www.diyaudio.com/forums/showthread.php?s=&threadid=50964&perpage=10&highlight=&pagenumber=2
BTW: I can´t see how two drivers spaced "d" away in a isobarik design should work different than one driver in an OB with a width of "2d".
It´s midnight over here and I call it a day. Bye.
Rudolf
(JPK) Sorry to jump in so late but the bottom line here is that dipole is dipole. Two sources separated by a distance d and operating out of phase will generate the same response only when they have the same volume velocity (cone area times cone displacement). Whether on an opened baffle, a separate sealed boxes, or sharing an isobaric box. The spl is basically a function of cone excursion. All the different configurations do is change the alignment. If the have separate boxes then the individual driver response will be what ever the driver in box is aligned to. If you want the response to follow some target you need to add a response shaping network if the in box alignment isn't what you need. If you place the drives in a shared isobaric chamber then it's basically the same thing a single driver on an open baffle with the same front to back separation once the response is shaped and eq'ed to the target. With two drivers on an open baffle you can pick up 6dB in max SPL capability over either dual driver in box configuration. As one of the other posts pointed out, one difference is if an H frame is used you may have to deal with a cavity 1/4 wave resonance. The other difference is that they may be different power requirements because if individually loaded by a sealed box the the impedance will be different than for a OB or isobaric type arrangement.
There is one other consideration. If you place the drivers in individual sealed boxes you may (may) be able to achieve some of the required dipole EQ by aligning the drivers to a high Qts. Of course, you will sacrifice output below Fb in that case but it could potentially be more "satisfying" than the same drives on an open baffle without equalization.
If I repeated anything already posted, I apologize. I wasn't able to read all the posts in detail.
There is one other consideration. If you place the drivers in individual sealed boxes you may (may) be able to achieve some of the required dipole EQ by aligning the drivers to a high Qts. Of course, you will sacrifice output below Fb in that case but it could potentially be more "satisfying" than the same drives on an open baffle without equalization.
If I repeated anything already posted, I apologize. I wasn't able to read all the posts in detail.
Rudolph,
I understand what you mean about 2 sealed in a dipole config. The point Magnetar was making, with which I agree and doesn't conflict with info on SL's site, is that a single driver is generally going to have a higher SPL in a box design that in free air operation at 1watt. This is why you see multiple drivers used in IB setups. I think Magnetar's point is that since you start at a point of higher SPL then even with the dipole rolloff that you still have higher output. This doesn't consider max SPL at the bottom end because that is limited by excursion.
I made an incorrect statement earlier. At the point of d=1/2 wavelength, the front and rear waves reinforce each other exactly and the output is +6db higher than the single monopole. From that point down the SPL decreases at slightly less than 6db per octave but is not linear until you reach the point where the combined output is equal to the front side only output. SL calls this point Fequal and is calculated with the formula Fequal= .17 x 343m/s / "d" where "d" is the extra distance of travel by the rear wave, in meters. Below this point the dipole bass roll-off is 6db/octave assuming a flat response in output.
The reason I bring this up is because no matter how you achieve your dipole response, the decrease in SPL due to dipole must be addressed and EQ'd in some fashion to get a flat response, so Magnetar was incorrect about that part.
In addition, if you start plugging baffle dimensions into the formula above you will understand why big baffles are just a waste of wood if your goal is a flat response because you need eq anyway.
I understand what you mean about 2 sealed in a dipole config. The point Magnetar was making, with which I agree and doesn't conflict with info on SL's site, is that a single driver is generally going to have a higher SPL in a box design that in free air operation at 1watt. This is why you see multiple drivers used in IB setups. I think Magnetar's point is that since you start at a point of higher SPL then even with the dipole rolloff that you still have higher output. This doesn't consider max SPL at the bottom end because that is limited by excursion.
I made an incorrect statement earlier. At the point of d=1/2 wavelength, the front and rear waves reinforce each other exactly and the output is +6db higher than the single monopole. From that point down the SPL decreases at slightly less than 6db per octave but is not linear until you reach the point where the combined output is equal to the front side only output. SL calls this point Fequal and is calculated with the formula Fequal= .17 x 343m/s / "d" where "d" is the extra distance of travel by the rear wave, in meters. Below this point the dipole bass roll-off is 6db/octave assuming a flat response in output.
The reason I bring this up is because no matter how you achieve your dipole response, the decrease in SPL due to dipole must be addressed and EQ'd in some fashion to get a flat response, so Magnetar was incorrect about that part.
In addition, if you start plugging baffle dimensions into the formula above you will understand why big baffles are just a waste of wood if your goal is a flat response because you need eq anyway.
Thank you John K! An actual EE has spoken. Makes perfect sense. Magnetars "theories" explained. Just say you don't like someones design from a personal standpoint. You have a preference for 400lb horns,etc. Thrashing someonelse's design because you dont understand the principles at work are a bad idea. SL's open baffle design is terrific. John K's NaO is another superb OB design. If you say something ridiculous about a solid competent design that is simply false because of your lack of knowledge - you will catch my wrath. I make no apologies to Magnetar for rambling then running. Stand up for what you say and back it up with facts. Hiding doesn't help. If Magnetar does not post here again, so be it. What a loss. Yawn. Got to get back to buiding my OB coaxials. See Ya.
Cheers,
AJ
Cheers,
AJ
AJinFLA said:
I understand the principles just fine, I have had both U and W dual 15" baffles here for years (built one pair back when the article first appeared in speaker builder - but with four JBL 2235H's ) with an assortment of drivers (from eminence 1.0 QTS 15's down to JBL 222O's @ .21 QTS) and eq's - None of the combinations with these baffles/drivers/eq's perform as well as the enclosed (tuned) dipoles - and the enclosed dipoles are 3db down from 200 at 40 cycles (EVM15B, front driver tuned to 40 Hz, EVM15L rear with 2 variovents) with no eq. To "fine tune" the system to my room I simply use a pair of Paradigm X30's and adjust the phase relationship between the front and back drivers. If I really want to excite the room I simply switch them over to bipoles (say with some Michael Stearns @ 125 db) by reversing the phase 180 degrees on the back woofer amp. When you build your coaxial dipoles remember that it needs to have dipole operation with a damped backwave throughout it's full range to sound best - contrary to what the experts say! LOL
😉 Have fun and enjoy your compromises.Other Compromises
Magnetar,
Why is it that guys like yourself (horn designers too for that matter) love to treat things like they are top secret and remain as vague as possible for as long as possible? Is it because you plan to go commercial or just want everyone else to put in the same sweat equity that you obviously have ?
BTW, I'm on your side vs the dipole purists, I prefer some attentuation of the rear wave. It seems to make placement much more flexible with no ill effects. I hope you hang around to pick your brain with a few questions.
1. I have to assume that in your dipole (hybrid) mode that you have room modes under control. How much bass to you have at the sides of your speakers? I guess my real question is how much can you fudge that pure dipole null at the sides and still keep ceiling and side wall modes under control?
2. I've planned to go the W baffle route with lots of drivers, since my target is OB down close to 28hz where my remaining room mode comes into play. I've often thought of using stuffing in the rear openings as a way to dial in more bass with a W baffle, resulting in comething somewhere between Aperiodic and dipole. Did you ever play around with that? If not, do you think it makes sense?
3. Another thought to generate more low end with OB, is to use a big panel absorber (possibly a helmholtz absorber that can be tuned to specific frequency ranges) directly behind by OB's. The idea is to absorb as much of the rear wave hitting the wall as possible. The ceiling and side nulls would remain pretty much intact, but whatever gets absorbed in back has to equate to more net output in the front. I would get the added benefit of a bass trap to help address my one remaining room mode. Does that idea hold water at all ?
4. For an HT sub to be used below 28hz use a very low tuned ported sub near the rear wall wired out of phase, then a similarly tuned sub 12ft closer and very near the listening position in the form of a coffee table. I come up with an Fequal of 16hz, plus at 1m distance vs over 4 meters the coffee table sub will be about 12db louder at the seating position. This should give me tremendous seating area bass, but a drastic reduction of what leaves the room. I can also play around with some stuffing in the vents for fine tuning. As long as I keep my low pass quite low and all output is below my lowest room mode, it seems like a great idea. Is there some problem that I'm not considering ?
Thanks for your time,
John
Why is it that guys like yourself (horn designers too for that matter) love to treat things like they are top secret and remain as vague as possible for as long as possible? Is it because you plan to go commercial or just want everyone else to put in the same sweat equity that you obviously have ?
BTW, I'm on your side vs the dipole purists, I prefer some attentuation of the rear wave. It seems to make placement much more flexible with no ill effects. I hope you hang around to pick your brain with a few questions.
1. I have to assume that in your dipole (hybrid) mode that you have room modes under control. How much bass to you have at the sides of your speakers? I guess my real question is how much can you fudge that pure dipole null at the sides and still keep ceiling and side wall modes under control?
2. I've planned to go the W baffle route with lots of drivers, since my target is OB down close to 28hz where my remaining room mode comes into play. I've often thought of using stuffing in the rear openings as a way to dial in more bass with a W baffle, resulting in comething somewhere between Aperiodic and dipole. Did you ever play around with that? If not, do you think it makes sense?
3. Another thought to generate more low end with OB, is to use a big panel absorber (possibly a helmholtz absorber that can be tuned to specific frequency ranges) directly behind by OB's. The idea is to absorb as much of the rear wave hitting the wall as possible. The ceiling and side nulls would remain pretty much intact, but whatever gets absorbed in back has to equate to more net output in the front. I would get the added benefit of a bass trap to help address my one remaining room mode. Does that idea hold water at all ?
4. For an HT sub to be used below 28hz use a very low tuned ported sub near the rear wall wired out of phase, then a similarly tuned sub 12ft closer and very near the listening position in the form of a coffee table. I come up with an Fequal of 16hz, plus at 1m distance vs over 4 meters the coffee table sub will be about 12db louder at the seating position. This should give me tremendous seating area bass, but a drastic reduction of what leaves the room. I can also play around with some stuffing in the vents for fine tuning. As long as I keep my low pass quite low and all output is below my lowest room mode, it seems like a great idea. Is there some problem that I'm not considering ?
Thanks for your time,
John
johninCR said:
Its just a bit of arrogance (perhaps unintended) coupled with a desire to keep replies short (..which usually back-fires). He does it all the time regardless of which forum he posts on.
BTW, what it sounds like he has just described is not completly a dipole (i.e. the spl at all freq.s for the in-phase and out-of-phase signals are likely NOT matched in level - particularly if you start playing with phase rotation). Call it then a Quasi-dipole (to be more accurate). In any event a dipole is not inherently superior in operation (just the opposite is true) - instead it just happens to be a method that can provide a smoother in-room low freq. response because of its ability to circum"vent" room nodes.
- Status
- Not open for further replies.