First time OB builder here. I've been trolling this thread looking for a no compromise OB speaker I could attempt to build that was in line with my overall tastes and talents. I am no speaker designer, but woodworking and electronics I can handle, so its time to replace my big box towers.
I saw Perry's build of these live edge dipoles and loved the natural wood look, design decisions, and his description of the end product certainly sounds sublime. I am strongly considering something exactly like this for my build.
I reached out to Perry with a few questions specifically related to his speakers with a relatively novice builder in mind. His thoughts were to make the information public if anyone else out there had similar questions, so we could share information. (love this forum btw)
I was hoping he might share the rough frame dimensions, or any schematics that could be helpful.
I was also curious about the recessed drivers and the grills. If the grills were custom made or sourced specific to the drivers. They are very slick.
Also was curious about his crossover and the specifics related to capacitors. If he recommended a certain brand for his crossover components.
I cant wait to get my hands dirty with a project like this.
Btw, a big thanks to all like Perry who willingly share information in the pursuit of expanding the field for everyone.
I am extremely happy with these. I have a hard time thinking of anything I would change. One thing I could change is triamp instead of biamp, and delete the passive xover components. That's certainly viable but it's a lot of extra complexity.
Originally I used the Beyma 15XA38ND for the coax, but its polar pattern is a hot mess between 4K and 7K and so I switched in the B&C 15CXN88s which are much better.
The grilles are hoops made by my carpenter on a CNC machine, I believe they are 1/8" shallower and 1/8" smaller diameter than the allotted space so after the cloth is stretched over they fit snugly.
The schematics are here: https://evo2.org/wp-content/uploads/2021/06/Bitches_Brew_Open_Baffle_Live_Edge_Speakers.pdf
Crossover components: Goertz copper foil inductors from Madisound. Large capacitors are 1/3 of capacitance contributed by mylar or film caps and 2/3 from electrolytic caps.
Originally I used the Beyma 15XA38ND for the coax, but its polar pattern is a hot mess between 4K and 7K and so I switched in the B&C 15CXN88s which are much better.
The grilles are hoops made by my carpenter on a CNC machine, I believe they are 1/8" shallower and 1/8" smaller diameter than the allotted space so after the cloth is stretched over they fit snugly.
The schematics are here: https://evo2.org/wp-content/uploads/2021/06/Bitches_Brew_Open_Baffle_Live_Edge_Speakers.pdf
Crossover components: Goertz copper foil inductors from Madisound. Large capacitors are 1/3 of capacitance contributed by mylar or film caps and 2/3 from electrolytic caps.
I'd just like to add: People conjure up with all kinds of ideas as to what an "ultimate" system might look like. Often they are extremely complex designs with lots of excess drivers or technology or whatever.
I wanted something that at least in outline form was simple and elegant. Well, with the "Two and a half way" crossover design and the coaxial driver, the very focused, well-behaved polar pattern and the dipole configuration, I think these achieve that. I don't think throwing money or components at them would make them better. I suppose extreme bass-heads might add a sub-subwoofer for flapping your pant legs below 20Hz but 99.5% of music lovers would find that entirely unnecessary.
These are stiff competition to anything you can buy in a store, or see at an audio show for less than $100,000; and they'll give most of the designs north of $100K a good run for their money.
I don't know of any other system I've ever seen that has this combination of attributes - constant directivity across the entire band 20-20K, flat phase response, pristine step and impulse response, dipole pattern, 100dB efficiency, 500 watt power handling, dynamic range to burn. The imaging and transparency are unreal. You can walk ALL around the room and the stereo image is stable wherever you go.
And most women (including my wife) think they're gorgeous.
I encourage you to build these and show them off here on the forum.
I wanted something that at least in outline form was simple and elegant. Well, with the "Two and a half way" crossover design and the coaxial driver, the very focused, well-behaved polar pattern and the dipole configuration, I think these achieve that. I don't think throwing money or components at them would make them better. I suppose extreme bass-heads might add a sub-subwoofer for flapping your pant legs below 20Hz but 99.5% of music lovers would find that entirely unnecessary.
These are stiff competition to anything you can buy in a store, or see at an audio show for less than $100,000; and they'll give most of the designs north of $100K a good run for their money.
I don't know of any other system I've ever seen that has this combination of attributes - constant directivity across the entire band 20-20K, flat phase response, pristine step and impulse response, dipole pattern, 100dB efficiency, 500 watt power handling, dynamic range to burn. The imaging and transparency are unreal. You can walk ALL around the room and the stereo image is stable wherever you go.
And most women (including my wife) think they're gorgeous.
I encourage you to build these and show them off here on the forum.
Awesome stuff Perry. Really, thanks.
I agree with your philosophy on design. Plus, I love a big bold speaker that can somehow maintain in-home acceptability. Thats where the wood shines in my book.
I was wondering how you ended up at your total capacitance when I was reviewing your crossover schematic. The very little crossover work I've done followed builds that used lower uf values and could be mapped out with just a few simple film caps.
I agree with your philosophy on design. Plus, I love a big bold speaker that can somehow maintain in-home acceptability. Thats where the wood shines in my book.
I was wondering how you ended up at your total capacitance when I was reviewing your crossover schematic. The very little crossover work I've done followed builds that used lower uf values and could be mapped out with just a few simple film caps.
The values used crossing from woofers to midrange (roughly 6mH and 350uF as I recall) generate a Q somewhat greater than 1, leading to an RLC filter having an effective acoustic crossover frequency of ~110Hz. This is a series type crossover which isn't often used, but has much less interaction with driver impedance peaks than the standard parallel topology most people use.
perrymarshall,
Really nice work!
Did you connect the low frequency and high frequency drivers in series?
And then, did you connect 350uF across the low frequency driver, and 6mH across the high frequency driver?
Well, I answered my question above, because I read your earlier posts, and saw your schematics.
I would like to have a set of those speakers, but I would only do that if I had someone else build them for me.
I once did such a series crossover for a 6-1/2 in driver, and a 'phenolic' tweeter (with a capacitor and an inductor set to about 2.5kHz).
Really nice work!
Did you connect the low frequency and high frequency drivers in series?
And then, did you connect 350uF across the low frequency driver, and 6mH across the high frequency driver?
Well, I answered my question above, because I read your earlier posts, and saw your schematics.
I would like to have a set of those speakers, but I would only do that if I had someone else build them for me.
I once did such a series crossover for a 6-1/2 in driver, and a 'phenolic' tweeter (with a capacitor and an inductor set to about 2.5kHz).
Last edited:
Hello all. I've been home for enough time to start a couple buildsz to cut down on the little stashes of gear in my apartment. I'd like
Some open baffle bass that goes a bit deeper than some 3mm excursion instrument speakers, and in a smaller package.
I won't be buying more drivers.
At hand are two nos Ariel by Scanspeak 10" paper 8 ohm sub drivers, iirc, about a 0.44 qts and 8.5 mm X-Max. 4 Dayton RSS 265HF-4 10" drivers, about 11mm X-Max, and a 0.41 qts.
I don't rattle the neighbor's anymore, but would like to add to or crossover to a pair of open back main speakers I'm using now.
I'm wondering if a lx521 styled bass bin or two may be fine with movies in a 14x 25' room with a regular 8 foot ceiling.
I have a few choices in power amps, ranging from 300, 500, and 1000 watt plate amps, and a 250 watt stereo non bridgeable power amp.
These are spares from backups for sold models over the years, and the doctors tell me I can't return to work, so there'll be no purchases for this.
If necessary, I could put an old equalizer in the line level to an amp, and or an old active crossover with up to a 10db lift at 50 hertz, and a second order filter to highpass lower bass at 32 hertz.
Just looking for some ideas before making sawdust.
Some open baffle bass that goes a bit deeper than some 3mm excursion instrument speakers, and in a smaller package.
I won't be buying more drivers.
At hand are two nos Ariel by Scanspeak 10" paper 8 ohm sub drivers, iirc, about a 0.44 qts and 8.5 mm X-Max. 4 Dayton RSS 265HF-4 10" drivers, about 11mm X-Max, and a 0.41 qts.
I don't rattle the neighbor's anymore, but would like to add to or crossover to a pair of open back main speakers I'm using now.
I'm wondering if a lx521 styled bass bin or two may be fine with movies in a 14x 25' room with a regular 8 foot ceiling.
I have a few choices in power amps, ranging from 300, 500, and 1000 watt plate amps, and a 250 watt stereo non bridgeable power amp.
These are spares from backups for sold models over the years, and the doctors tell me I can't return to work, so there'll be no purchases for this.
If necessary, I could put an old equalizer in the line level to an amp, and or an old active crossover with up to a 10db lift at 50 hertz, and a second order filter to highpass lower bass at 32 hertz.
Just looking for some ideas before making sawdust.
Can anybody suggest a way to simulate a drive unit in open space, i.e. not on a baffle?
In Basta when I simulate a very small baffle around the driver I can see the baffle step response as expected but when I move the driver off the baffle into free space the usual baffle step effect becomes almost flat, which seems strange to me.
In Basta when I simulate a very small baffle around the driver I can see the baffle step response as expected but when I move the driver off the baffle into free space the usual baffle step effect becomes almost flat, which seems strange to me.
I haven't used Basta, but is there Open baffle toggle somewhere like in The Edge?
The Edge is the easiest way for that. Just toggle Open baffle! Yes you must have a baffle, but overlay baffle and driver with tiny bit smaller diameter! Baffle with 12 corners behaves just like a circular baffle with same diameter.
VituixCAD 2 is total speaker/xo simulation package with several modules eg. baffle effect simulation.
Attachments
Last edited:
Even a naked driver actually has a baffle, as the basket normally has an bigger diameter than the cone.
So for good simulation results of a naked driver in Edge, I would take the cone diameter as driver diameter in Edge and create a circular baffle (or use the method described by Juhazi) in the size of the overall speaker diameter.
At least for my measurements, this method gave me the most realistic results.
Satellites
Still burning up spare parts.
This is a follow-up to an earlier experiment to address comb filtering. It doesn't quite do it by the numbers, but is far closer than conventional tweeter woofer pairs.
Crossover points are a 2700 hertz highpass to a 2" Seas fabric dome tweeter, 2500 and 800 hertz bandpassed 3" metal dome tweeters, and lowpassed at 800 hertz MarkAudio 12PW full rangers, one lowpassed at about 375 hertz solen split .
Crossover points are just guesses based on similar designs, and will likely be tweaked.
Deeper bass will be provided by a plate amp and existing panels to be used as open back bass players.
Just a first dry fit, enough for one day.
3/4" light stock for U wings, MDF laminated to 1 inch thick baffle. OD 10" x 16" x 25".
Sort of an ugly duckling, but I wanted a tweeter that's an inch smaller than the bandpassed pair.
Still burning up spare parts.
This is a follow-up to an earlier experiment to address comb filtering. It doesn't quite do it by the numbers, but is far closer than conventional tweeter woofer pairs.
Crossover points are a 2700 hertz highpass to a 2" Seas fabric dome tweeter, 2500 and 800 hertz bandpassed 3" metal dome tweeters, and lowpassed at 800 hertz MarkAudio 12PW full rangers, one lowpassed at about 375 hertz solen split .
Crossover points are just guesses based on similar designs, and will likely be tweaked.
Deeper bass will be provided by a plate amp and existing panels to be used as open back bass players.
Just a first dry fit, enough for one day.
3/4" light stock for U wings, MDF laminated to 1 inch thick baffle. OD 10" x 16" x 25".
Sort of an ugly duckling, but I wanted a tweeter that's an inch smaller than the bandpassed pair.
Attachments
Answer: They create a short transmission line that generates peaks between 150Hz and 250Hz (more for the bottom woofer than the middle one). But since the crossover is at 100Hz, those peaks don't have a large effect on the overall response. Also because the speaker is as wide as it is deep, the peaks aren't very large.
You can see a more pronounced version of this effect here:
https://psma-website-assets.s3.amazonaws.com/aegean_wave_dipoles-1.pdf
See the EQ curve on page 2 of the above PDF, you see large dips in the digital xover that compensate for peaks at 230Hz and 380Hz. The peaks are large because the transmission line is long and narrow.
What the "wings" buy me is much deeper bass extension. This system starts rolling off via Dipole cancellation at around 45Hz rather than at 80Hz which would be the case if it was just a flat baffle. So with the xover at 110Hz and 1100, I get true dipole radiation and bass extension at the same time.
You can see a more pronounced version of this effect here:
https://psma-website-assets.s3.amazonaws.com/aegean_wave_dipoles-1.pdf
See the EQ curve on page 2 of the above PDF, you see large dips in the digital xover that compensate for peaks at 230Hz and 380Hz. The peaks are large because the transmission line is long and narrow.
What the "wings" buy me is much deeper bass extension. This system starts rolling off via Dipole cancellation at around 45Hz rather than at 80Hz which would be the case if it was just a flat baffle. So with the xover at 110Hz and 1100, I get true dipole radiation and bass extension at the same time.
Last edited by a moderator:
Some posts split to here - Hello!
Wanted to take advantage of the multitude of OB knowledgeable folks around this thread and request input about a couple of options I'm considering towards ultimate bass/midbass in a dipole system.
Currently I have two 12" Rythmik servo subs in DIY sealed boxes below 65Hz in mono, and per side: one 18" Faital baffleless dipole on a swing from 65 to 275Hz, twin 8" from 275Hz to 1.8kHz (bottom 8" on U baffle to reach to 275Hz) and AMT above. Active 4-way, digital xo and room correction.
Subs are great and thought about adding two more to get a distributed bass array. And I came across GR Research OB servo subs in H or U baffle, getting great comments, and can be xo up to 200 or 250Hz.
I'm wondering if I would be better off:
I love the bass articulation I'm getting today with the 18" Faitals. The twin 12" GR would offer about a third less Sd at 65Hz (maybe less sense of impact?), but servo twin 12" in OB is said to be even more articulate than a naked 18".
Does anybody have experience with these?
Wanted to take advantage of the multitude of OB knowledgeable folks around this thread and request input about a couple of options I'm considering towards ultimate bass/midbass in a dipole system.
Currently I have two 12" Rythmik servo subs in DIY sealed boxes below 65Hz in mono, and per side: one 18" Faital baffleless dipole on a swing from 65 to 275Hz, twin 8" from 275Hz to 1.8kHz (bottom 8" on U baffle to reach to 275Hz) and AMT above. Active 4-way, digital xo and room correction.
Subs are great and thought about adding two more to get a distributed bass array. And I came across GR Research OB servo subs in H or U baffle, getting great comments, and can be xo up to 200 or 250Hz.
I'm wondering if I would be better off:
a) Adding 2 Rythmik servo sealed subs and leave the rest as-is.
b) Adding 2 GR Research servo OB subs (with two drivers each) and run them 20-65Hz with the sealed subs, getting the DBA, leaving the rest as-is.
c) Adding 2 GR Research servo OB subs with 2 drivers each, run them 20 to say 150Hz, let the sealed subs complement 20-65Hz (getting DBA under 65Hz), and replace my 18" Faitals with maybe a 12" or 15" naked driver and xo to the 8" higher up and remove its U baffle.
b) Adding 2 GR Research servo OB subs (with two drivers each) and run them 20-65Hz with the sealed subs, getting the DBA, leaving the rest as-is.
c) Adding 2 GR Research servo OB subs with 2 drivers each, run them 20 to say 150Hz, let the sealed subs complement 20-65Hz (getting DBA under 65Hz), and replace my 18" Faitals with maybe a 12" or 15" naked driver and xo to the 8" higher up and remove its U baffle.
I love the bass articulation I'm getting today with the 18" Faitals. The twin 12" GR would offer about a third less Sd at 65Hz (maybe less sense of impact?), but servo twin 12" in OB is said to be even more articulate than a naked 18".
Does anybody have experience with these?