Bipolar FR's in a Sealed Enclosure

[joeyb462]

Great forum, first time poster.

As I am currently in the research phase of a DIY build, I thought I would come to the masters. I have five goals.

1. Must be bipolar – not for the purist, but I like the sound stage.
2. Big sound – because Miles Davis deserves nothing less.
3. Must be relatively compact – I do have a wife.
4. Crossover-less – I am a beginner to speaker design. It would seem that full range and compact don’t play nice.
5. Can be driven by my DIY set of 45 Watt KT120 monoblocs

After much research and fiddling with freeware programs like WinISD and the like, I was convinced that my efforts were severely misguided until I ran into the Blueberry Audio Rhapsody in 4D. Anybody looked at this system? Appears to use Fostex FE208EZ Sigma FR drivers in a very compact sealed enclosure. Highs are handled with what looks like a Fostex TA500. While not necessarily crossover-less, this design seems simple enough due to the FR drivers and meets most of the criteria. The tweeter may only have the attenuator and cap in its network, but I don’t know.

I am very interested in this concept, but have no idea how to model a system like this. Any takers out there that want to point me in the right direction? I just finished reading the entire internet and it appears silent on the subject.

Thanks for your kind opinions.

Joey

[JonMarsh]

If they supply T/S parameters, you can model the box reposes as for any other speaker. If it's truly crossover less, it would have to have a natural BSC built in to the frequency response, but from what I've heard of most full range systems, many of those who favor that type just live with the frequency response. But many don't.

Note, if you were listening to the speaker I believe you were, that includes a bass system, with a servo controlled amplifier, so I really don't think between that and the tweeter setup that you could call it crossover-less. Developing something like this could be fairly straightforward, IF you have test equipment and speaker modeling software, and quite a bit of spare time on your hands. Have you showed your wife a picture of the Rhapsody's? My GF is an ex IBM systems engineer, so she's somewhat tolerance of technology, but I'm not sure she would go along with this in our family room...



This rear view shows the rear facing midrange and the power amp and crossover for the woofer system.



A close up of the woofer plate amp/crossover (off the shelf DirectServo model from Rhythmik)



The tweeter is a bit more adjustable than average, with both a rheostat (with a rather wide range!!) and an angle adjustment- that suggests to me that

A) the setting is fairly critical depending on where you sit

B) the tweeter may not have the widest dispersion, OR, the way it blends with the full range driver has issues with regards to off axis behavior.

Note, the suggested retail on this system is $25K.

The "full range" driver looks a lot like this Fostex 6.5" unit

FE168EZ

A number of visible resonances in the impedance curve.



T/S parameters for enclosure design/modeling:



IF that's the IB (infinite baffle) response, then it does need baffle step correction (AKA crossover filter) to shape the response.

you can see the standard Rhythmik Direct Servo plate amps here...

http://www.rythmikaudio.com/amplifiers.html

The one Blueberry is using looks a lot like this one:



The one Blueberry uses is likely customized for him, and with a servo interface board specific to the woofer drivers he's using. Which may or may not be one of Rhythmik's drivers.

[5th element]

Making a bipolar loudspeaker is something that I've been considering for a while but as usual I have absolutely no use for it apart from satisfying my desire to try it :scratchhead: This seems to be my biggest problem with loudspeakers, well heck amplifiers too I've got plenty of those sitting on shelves and DACs...

Anyway the nice idea behind this is that in theory the bipolar arrangement should compensate perfectly for baffle step losses but they will not (in theory) work well positioned close to the rear or side walls.

Usually in a monopole loudspeaker you get half space/hemispherical radiation where all the energy from the driver radiates forwards into a hemisphere. This happens at high frequencies where the width of the loudspeaker cabinet is large in relation to the wavelength of the sound.

As frequency drops and the wavelengths increase, they become significant in relation to the cabinet width and they then start to diffract around the cabinet edges, some energy now starts being lost around the back of the cabinet. Keep going down in frequency and you get to a point where exactly half of the energy is now passing behind the cabinet as the wavelength of sound is so large that the sound waves simply ignore the presence of the cabinet entirely. Energy is now radiated spherically.

This creates a frequency response that looks like this.



Now theoretically if you place a driver on the rear of the cabinet, if there are no reflective surfaces nearby, it should compensate perfectly for baffle step losses in the usual forward listening position. At low frequencies the spherical radiation from the rear driver will add to the spherical radiation of the front driver and reinforce what is lost. But as frequency increases and the rear driver starts to radiate hemispherically the cabinet will start to act as an acoustic filter and block those frequencies from the listener.

How well this would work in a room I do not know. Obviously the reflected rear radiation would be imposed upon the room but it would be uncorrelated with the radiation from the front. You're definitely going to get some sort of comb filtering but if the room has a decent amount of furnishings it's unlikely to do anything more than give you 3dB of baffle step compensation rather than the full 6 if I'm perhaps remembering theory correctly. Of course this is if the driver has omnidirectional radiation up till 20kHz, which it doesn't so you would get a downward tilted response anyway - which is a good thing.

If you don't want to use any crossover at all this wont do any harm and will only help things in my opinion. The rear radiation wont be to everyone's cup of tea, but if this is what you are specifically wanting anyway then it's perfect :W

[joeyb462]

Mr. Marsh and 5th Element,

Thanks for your timely responses and your willing interest. I am going to digest what both of you have said as it is abundantly clear that you guys are way ahead of me. A few clarifications to address some of the points both of you made.

1. I have a pair of 12” RS1000K Dayton Audio Kit subs powered by plate amps to handle the bottom end. I plan to keep those so a roll off @ 100 hz is a-ok. Although, those Rythmiks look interesting.
2. My listening area is 23’x20’ w/ 13’ ceilings. Not ideal, but plenty of space to separate from the wall.
3. Current system consist of aforementioned subs, Bose Series III 901’s (let the judging begin) driven by the monoblocs and Fostex TA90 Horn Tweeters with a virtual x-over set at 5 KHZ driven by a $20 Leipi 2x20 amp. Actually sounds decent but a bit of a hodgepodge.
4. The rhapsody is an ugly beast, no doubt. Since my wife was kind enough to give up the home for her fancy car in our 3-car garage (err-woodshop), she supports what I can produce out of the cave and whatever I create will look nothing like the rhapsody. Form will not be an issue. Function however….
5. Agreed, Mr. Marsh, the tweeter is likely very directional if it is anything like the TA90 Horn. Interesting Commentary from the Blueberry site “One of the most important specifications of an audio driver is angle of horizontal dispersion. Midrange drivers have an angle of horizontal dispersion significantly larger than tweeters. In all existing Hi-End Stereo Systems midrange and tweeter drivers position is locked. Toeing-in speakers in the search for the best tweeter performance - imaging, sound stage, depth, etc. can degrade the performance of midrange driver(s) and visa versa. Therefore, midrange and tweeter drivers must have the option of being independently adjustable.”
6. Regarding Crossover-less and FR drivers, also from Blueberry site “RHAPSODY in 4D Speaker System employs two 8" FULL RANGE drivers and one super tweeter in each channel. All drivers are internally wired to dedicated speaker terminals for direct coupling with amplifier(s) without crossovers. Parallel connected FULL RANGE drivers raise system sensitivity to 100 db.” Also stated is 4 ohms so these are paralleled 8 ohmers I think.
7. And yeah, $25K is a bit much for $6k in components and a few $hundred in plywood.

Speaking of diffraction, 5th Element, do you suppose that the cylindrical design of the bipole section is related?

Thanks again for the input

[cjd]

I'm curious about the "crossoverless" requirement - that puts some very clear restrictions on what can be done. I do imagine that Miles will sound good enough, but in my experience, there's always something you're giving up. It may only sound good from one place, for example (and I think that's often the case.) If it's just the "beginner" aspect, have no fear - if you can take and post photos, we can help with crossovers. If it's other more obscure obsession around a sound that only lives in wide-range drivers, carry on.

I was only mildly disappointed with the resulting sound provided by the Dayton PS180 - it's been less recent that I heard a Fostex but it was not pleasing to me. On the other hand, I don't like bipole either - completely messes things up.