"Music Room" Open Baffle Project

[ErinH]

I recently started up an open baffle project. The idea was implanted by a friend who builds speakers. He mentioned the OB idea to me one day and it got my brain going. After hearing a set he built I was sold so I started it up.

The end goal is to have a true reference system (or as close as I can reasonably get). The setup will be in my upstairs room, which is roughly 17x22ft (recalling from memory). The motivation to do this actually came from my wife, who after listening to a pair of bookshelves I recently built, encouraged me to turn our upstairs space into a music room. Don’t have to ask me twice.
I plan to move my computer upstairs and have all my media accessible right there and possibly use my netbook as my controls via VNC.

The idea behind open baffle projects (aka: fullrange/true dipole) is that you remove the box from the equation, thus getting a truer representation of the music. There’s plenty of info on this kind of build out there so if you’re interested get to googling.
If you have any specific questions I’ll do my best to answer, but as noted below, this is a work in progress, hopefully to be completed by July.


The final setup will consist of the following gear though:
Tang Band w8-1772 w/Adcom gfa-2535 @ 2x60w (only using 2 channels)
Acoustic Elegance OB15 w/Behringer 2500 @ 2x800w

The below are BY NO MEANS the final product. This was simply one portion of the build: experimentation and measurement. The final baffle will likely be about 4ft tall and 2ft wide at the top (after many, many simulations with the Edge program). Of course, the subs' will also dictate baffle size and shape, too, so the 4x2ft baffle is only a rough idea. I'm going to kerf the edges so that I can have some curvature rather than a flat board. I plan on making a few different baffle types after using software to narrow it down to about 3 or 4. After that I'll do in home testing to see which works the best.

The baffles in the below pictures were pretty much built solely for this purpose and also some random measurement purposes for comparing to baffle size influence (which correspond well to the simulations I modeled). As I said above, this is just a glimpse of what's to come. The sub in the picture will be replaced by a single ae OB15 on each side and maybe 2 if I find that one is not enough.
So far, though, I'm very impressed.

Once everything is said and done I'll update you guys and try to provide some of the 'science' that caused me to arrive at my final design such as baffle design vs frequency response and also selective placement of rear wave acoustic absorption/diffraction. :thumbsup:



Anyway... just a few pictures...

Sub amp. Already has the fan mod for quieter fan and the guy sent me XLR/rca cables so I don't have to make my own now.

As a side note, it's crazy awkward to be behind speakers and still have an image in front of you.
These drivers need some break in time. The first listen was pretty rough and there was serious harshness. After about a day of music things are much better.

 

[j_garcia]

My next sub is going to be open baffle. Similar to a previous sub I built, but that was dual free air subs in a sealed box with a 1.5" baffle. This one will be duals as well, but will be OB.

 

[ErinH]

Which ones do you plan to use?

I was looking at the emminence alpha 15a, but after some digging, I've found some unsatisfactory reviews so decided to go with the AE since I'm already using the IB15 in my car.

 

[j_garcia]

GR Research SW-12-16FRs with the Rythmik servo amp. Not sure what kind of output I will get, but I think with duals and the performance he is saying it is capable of with the servo amp driving them, I think it will be just fine.

 

[TLS Guy]

I recently started up an open baffle project. The idea was implanted by a friend who builds speakers. He mentioned the OB idea to me one day and it got my brain going. After hearing a set he built I was sold so I started it up.

The end goal is to have a true reference system (or as close as I can reasonably get). The setup will be in my upstairs room, which is roughly 17x22ft (recalling from memory). The motivation to do this actually came from my wife, who after listening to a pair of bookshelves I recently built, encouraged me to turn our upstairs space into a music room. Don’t have to ask me twice.
I plan to move my computer upstairs and have all my media accessible right there and possibly use my netbook as my controls via VNC.

The idea behind open baffle projects (aka: fullrange/true dipole) is that you remove the box from the equation, thus getting a truer representation of the music. There’s plenty of info on this kind of build out there so if you’re interested get to googling.
If you have any specific questions I’ll do my best to answer, but as noted below, this is a work in progress, hopefully to be completed by July.


The final setup will consist of the following gear though:
Tang Band w8-1772 w/Adcom gfa-2535 @ 2x60w (only using 2 channels)
Acoustic Elegance OB15 w/Behringer 2500 @ 2x800w

The below are BY NO MEANS the final product. This was simply one portion of the build: experimentation and measurement. The final baffle will likely be about 4ft tall and 2ft wide at the top (after many, many simulations with the Edge program). Of course, the subs' will also dictate baffle size and shape, too, so the 4x2ft baffle is only a rough idea. I'm going to kerf the edges so that I can have some curvature rather than a flat board. I plan on making a few different baffle types after using software to narrow it down to about 3 or 4. After that I'll do in home testing to see which works the best.

The baffles in the below pictures were pretty much built solely for this purpose and also some random measurement purposes for comparing to baffle size influence (which correspond well to the simulations I modeled). As I said above, this is just a glimpse of what's to come. The sub in the picture will be replaced by a single ae OB15 on each side and maybe 2 if I find that one is not enough.
So far, though, I'm very impressed.

Once everything is said and done I'll update you guys and try to provide some of the 'science' that caused me to arrive at my final design such as baffle design vs frequency response and also selective placement of rear wave acoustic absorption/diffraction. :thumbsup:



Anyway... just a few pictures...

Sub amp. Already has the fan mod for quieter fan and the guy sent me XLR/rca cables so I don't have to make my own now.

As a side note, it's crazy awkward to be behind speakers and still have an image in front of you.
These drivers need some break in time. The first listen was pretty rough and there was serious harshness. After about a day of music things are much better.

You can Google it, but most of the stuff out there is poor science and nonsense.

An open baffle dipole design like that is among the most difficult.

Your design is off to a bad start.

The baffle is too small and you need at least a partial box, otherwise the negative and positive pressure waves either side of the driver will cancel too severely.

Loss of LF progressively is the Achilles heel of this approach. You need a lot of power, sturdy expensive drivers with a linear motor.

Then you need an active crossover and Eq. A passive solution is impossible, no matter what anybody else may say on the net.

The only guy who has made this approach work is Siegfried Linkwitz, who has made the the major opus of his life's work.

You need to study and understand all his research and painstaking observations, if you want to turn this into a reference system.

It will be a steep climb for you but worthwhile. Even with DIY capital outlay will be significant.

 

[Ito]

Correct me if I am wrong, but wouldn't the out of phase waves from the back of the driver reflect off of the back walls and cause negative interference? I always thought that was one of the primary reasons of using an enclosure.

 

[TLS Guy]

Correct me if I am wrong, but wouldn't the out of phase waves from the back of the driver reflect off of the back walls and cause negative interference? I always thought that was one of the primary reasons of using an enclosure.

Reflection off the back wall is the least of the problem.

When the cone moves out there is positive pressure in front of the cone and negative pressure behind, and vice versa. The OP has a very short path and there is going to be total neutralization of the pressure waves either side of the cone. Estimating the size of his baffle, sharp acoustic roll off will start at 1 kHz, so not only will he not have a lick of bass, he will be rolling off well in the mid range. That won't change no matter how long he tries to break his driver in!

 

[ErinH]

Alright, dude… your post was completely unwarranted and rude considering you have no idea what I’ve been up to and obviously did not read my OP. With that said, I felt a reply was in order…

You can Google it, but most of the stuff out there is poor science and nonsense.

Not all of it, but that’s the nature of research. You have to learn to weed through the BS, the non-applicable, and the downright stupidity to find good info. That’s what I’ve been doing the past two weeks, in addition to talking with engineers and other hobbyists who have successfully built OB setups.

You need to study and understand all his research and painstaking

You assume WAY too much. Read the above.

Did you even bother to actually read my initial post?

Your design is off to a bad start.

The baffle is too small and you need at least a partial box, otherwise the negative and positive pressure waves either side of the driver will cancel too severely.

Where did I say this was the final product? Did you catch where I said the purpose of the pictured setup was only for measurement purposes and to compare with my simulated response would be via Edge?
Did you not also see the portion about how I likely will wind up with curved edges (to help serve as some extra wave separation)? That covers your bit about a pseudo-enclosure.

Further testing will involve about 3-4 baffle designs, once I've narrowed it down the most relevant versions. My RTA (with a calibrated mic) will be used for measurements in the listening environment (ie: not outside away from 'reflected surfaces' because that's not applicable and furthermore is completely and utterly useless to me and this project). I'll post that stuff up once I get a bit further along, though.

Loss of LF progressively is the Achilles heel of this approach. You need a lot of power, sturdy expensive drivers with a linear motor.

You don’t need “lots of power”. Did you bother to look up the sensitivity of the drivers I’m using? The TB’s are 95dB @ 1m/1w, IIRC. That means with about 10w they’ll have an output of ~112dB. That’s about 20dB higher than most average listening levels. Ten friggin’ watts…

Check the specs on the subs I said I’m using. Power is needed for the low end output, and I can use the term power because I understand what my goal is and what it will take to get me there (about 500w, based on rough calculations) but ‘power’ is relative and a function of many things (enclosure size, desired output level, sensitivity, etc)… it’s not a measuring tool on its own and should never be viewed as such.

Then you need an active crossover and Eq. A passive solution is impossible, no matter what anybody else may say on the net.

Re: Active crossover
You don't have to have an active crossover but it sure is a big help so you can determine the nominal crossover points and also to help you correct any issues you find. Of course, I'm using the miniDSP to help me set these up and, once finalized, I will make the passive crossover version so I can free the miniDSP up for other projects.

I also took some RTA measurements to get me a rough idea of the baffle's effect (and none of that windowed BS, either) but didn't post them because I didn't have the ability to post correlating Edge program simulations.




As I said in my first post: This is just a small sampling. It’s meant to be the beginning of product design, and I plan to (unless I keep getting replies like this and realize that my efforts go un-noted and people only want to assume the worst) keep updating it as I go with useful information so that others can see why I chose to do what I did in the end. It’s a work in progress… not ‘hey, check out my speakers’. I take my time to post these threads because it’s a hobby I LOVE as much as I love the reason I do it (music) and I want to help others further their love for this hobby as well… even if it’s only by simply seeing a link I post.
I'll update the thread with relevant info once I've finalized the design and I will post necessary links/measurements/simulations which helped me arrive at my final design.





FWIW, I spent last Saturday with a recording engineer listening to his 6 different speaker setups in 4 different rooms for about 3 hours. Quite an amazing array of speakers.
The following day I listened to a set of open baffle speakers with these full range drivers and a pair of peerless xxsls 12" (in an enclosure). The OB build had much more detail, 'better' transients, and a much deeper soundstage. All rooms are not built equal, but I think it can be done. And as I find stepping stones to get me there, I'll post it up.

Criticism is welcomed, but negative assumptions are not. I implore you, in the future, make sure you read a person's post before assuming they have no idea what they are doing. If you have legitimate questions or concerns, please share. I’m here to learn, but I don’t appreciate the negative temperament of posts like yours.


- Erin

 

[j_garcia]

For drivers that are designed for free air operation, IMO, the effects of an OB design are much less.

I was actually quite surprised what kind of bass the previous free air drivers I used in that other sub I built had with no box at all, but they sounded a lot better in the box I built. I knew a guy who was selling some of his own custom designs back then and I had him model the box up for me and it sounded great.

 

[ErinH]

The OP has a very short path

How do you know? Did you come to my house and measure? You say 'short path'... well, to where?

Please tell me the distance of my seated position to the rear wall, my seated position to the speakers, and the distance of the speakers to the each wall. If you can tell me that, then I'll send you an e-hug.

Estimating the size of his baffle, sharp acoustic roll off will start at 1 kHz, so not only will he not have a lick of bass, he will be rolling off well in the mid range.

You're wrong. Rolloff starts at about 400hz according to my measurement. Though, I was indeed a bit surprised as simple calculation (speed of sound/frequency) will tell you that the shortest distance (12in, in this case) would equal a wavelength of ~1150hz (full wave). This is why the final product should be measured because what you expect isn't what you get always. I have my suspicions as to why I get what I get, but I won't elaborate.

FWIW, here's a great link for anyone who wants to do the calculation quickly:
http://www.mcsquared.com/wavelength.htm

"sharp"... 6dB/octave is standard for an open baffle. I'd hardly call that sharp.

^ measured system response. The sub doesn't go high. You can clearly see the fullrange driver's rolloff, though.

That won't change no matter how long he tries to break his driver in!

And what does this have to do with anything? Is this just some sort of cheap jab? I've measured driver break in. Want me to show it to you with an impedance sweep?

The baffle size has nothing to do with break in, you are correct. But where did I EVER imply that that break in was related at all to the baffle size?







So, again, I'd appreciate you not making assumptions. I'll be happy to answer questions, but answering for me as though you are me is presumptuous and, quite frankly, aggravating.

If you want to reply to me, do so via PM. If you want to have a positive discourse on this subject, then we can do so. However, I'm not going to junk up my thread with anymore useless posts. I'll ask that each of your replies be deleted, too, if you want to continue this argument. In the future, I suggest you read and talk with the OP instead of insinuating he/she hasn't done due diligence in their efforts and assume he/she is a complete moron.

Thanks again,
Erin

 

[TLS Guy]

Alright, dude… your post was completely unwarranted and rude considering you have no idea what I’ve been up to and obviously did not read my OP. With that said, I felt a reply was in order…



Not all of it, but that’s the nature of research. You have to learn to weed through the BS, the non-applicable, and the downright stupidity to find good info. That’s what I’ve been doing the past two weeks, in addition to talking with engineers and other hobbyists who have successfully built OB setups.



You assume WAY too much. Read the above.

Did you even bother to actually read my initial post?




Where did I say this was the final product? Did you catch where I said the purpose of the pictured setup was only for measurement purposes and to compare with my simulated response would be via Edge?
Did you not also see the portion about how I likely will wind up with curved edges (to help serve as some extra wave separation)? That covers your bit about a pseudo-enclosure.

Further testing will involve about 3-4 baffle designs, once I've narrowed it down the most relevant versions. My RTA (with a calibrated mic) will be used for measurements in the listening environment (ie: not outside away from 'reflected surfaces' because that's not applicable and furthermore is completely and utterly useless to me and this project). I'll post that stuff up once I get a bit further along, though.



You don’t need “lots of power”. Did you bother to look up the sensitivity of the drivers I’m using? The TB’s are 95dB @ 1m/1w, IIRC. That means with about 10w they’ll have an output of ~112dB. That’s about 20dB higher than most average listening levels. Ten friggin’ watts…

Check the specs on the subs I said I’m using. Power is needed for the low end output, and I can use the term power because I understand what my goal is and what it will take to get me there (about 500w, based on rough calculations) but ‘power’ is relative and a function of many things (enclosure size, desired output level, sensitivity, etc)… it’s not a measuring tool on its own and should never be viewed as such.



Re: Active crossover
You don't have to have an active crossover but it sure is a big help so you can determine the nominal crossover points and also to help you correct any issues you find. Of course, I'm using the miniDSP to help me set these up and, once finalized, I will make the passive crossover version so I can free the miniDSP up for other projects.

I also took some RTA measurements to get me a rough idea of the baffle's effect (and none of that windowed BS, either) but didn't post them because I didn't have the ability to post correlating Edge program simulations.




As I said in my first post: This is just a small sampling. It’s meant to be the beginning of product design, and I plan to (unless I keep getting replies like this and realize that my efforts go un-noted and people only want to assume the worst) keep updating it as I go with useful information so that others can see why I chose to do what I did in the end. It’s a work in progress… not ‘hey, check out my speakers’. I take my time to post these threads because it’s a hobby I LOVE as much as I love the reason I do it (music) and I want to help others further their love for this hobby as well… even if it’s only by simply seeing a link I post.
I'll update the thread with relevant info once I've finalized the design and I will post necessary links/measurements/simulations which helped me arrive at my final design.





FWIW, I spent last Saturday with a recording engineer listening to his 6 different speaker setups in 4 different rooms for about 3 hours. Quite an amazing array of speakers.
The following day I listened to a set of open baffle speakers with these full range drivers and a pair of peerless xxsls 12" (in an enclosure). The OB build had much more detail, 'better' transients, and a much deeper soundstage. All rooms are not built equal, but I think it can be done. And as I find stepping stones to get me there, I'll post it up.

Criticism is welcomed, but negative assumptions are not. I implore you, in the future, make sure you read a person's post before assuming they have no idea what they are doing. If you have legitimate questions or concerns, please share. I’m here to learn, but I don’t appreciate the negative temperament of posts like yours.


- Erin

I did read your whole post. Your current baffle will only provide mid range data.

With your proposed baffle your roll of will start sharply at 500 Hz. And yes it does take a lot of power even with sensitive speakers because of the amount of Eq required.

You really do need to look a Siegfried Linkwitz's data to see what is going on.

He has all the designs and circuits you need to really make this work, and it can work well if you listen to the master of this approach.

 

[ErinH]

You really do need to look a Siegfried Linkwitz's data to see what is going on.

He has all the designs and circuits you need to really make this work, and it can work well if you listen to the master of this approach.

I already have, and in fact, have listened to the youtube video (just sayin'). A friend of mine is staying in one of his cabins next month, I believe.

I was really scouring his website when I was building my binaural mic setup. That guy's site is a treasure trove.

Estimating the size of his baffle, sharp acoustic roll off will start at 1 kHz, so not only will he not have a lick of bass, he will be rolling off well in the mid range. That won't change no matter how long he tries to break his driver in!

With your proposed baffle your roll of will start sharply at 500 Hz. And yes it does take a lot of power even with sensitive speakers because of the amount of Eq required.

Which one is it?

You got the 500hz right on the second guess.
And, again, what exactly is "sharp"? 6dB is the furthest thing I would call from 'sharp', which is exactly what standard OB design depicts as does my actual measurements.

 

[TLS Guy]

I already have, and in fact, have listened to the youtube video (just sayin'). A friend of mine is staying in one of his cabins next month, I believe.

I was really scouring his website when I was building my binaural mic setup. That guy's site is a treasure trove.








Which one is it?

You got the 500hz right on the second guess.
And, again, what exactly is "sharp"? 6dB is the furthest thing I would call from 'sharp', which is exactly what standard OB design depicts as does my actual measurements.

The big issues you have to solve are these.

Because of diffraction losses the combined Eq required below the wavelength of the path from front to back ends up being 18 db per octave.

At half the path distance the rear output will be in phase with the front output and create a peak. This is where an active solution becomes just about mandatory.

The power equation is a problem. A dipole cone requires 1000 times the cone excursion at 50 Hz than it does at 500 Hz. A monopole requires 100 times the excursion at 50 Hz than it does at 500 Hz. So that is sobering.

These problems are solvable, but suspect you have a steeper hill to climb than you think.

 

[ErinH]

good, quality post. Thanks.

I'm not being naive... I realize it's not going to be easy. I have a gut feeling that I'll ultimately wind up having to buy a 2nd set of subwoofers and stagger the crossovers, but before I go out and buy four, I'm going to try two. My room actually lends itself quite a bit to low end response.

This should prove interesting, and fun. My goal of July was set for a reason, though.

 

[TLS Guy]

good, quality post. Thanks.

I'm not being naive... I realize it's not going to be easy. I have a gut feeling that I'll ultimately wind up having to buy a 2nd set of subwoofers and stagger the crossovers, but before I go out and buy four, I'm going to try two. My room actually lends itself quite a bit to low end response.

This should prove interesting, and fun. My goal of July was set for a reason, though.

You are braver than I am. I have to say I would not attempt a dipole design from scratch using Moving coil loudspeakers.

If I wanted a dipole, I would build electrostatics, which believe it or not would be easier and give better results. Electrostatics are not that complicated really, once you get into it.

The other problem is splicing a standard subwoofer to a dipole. Peter Walker found that out. Gradient of Finland came to his rescue with a dipole sub designed for his ESL 63.

 

[ErinH]

Having heard a great dipole setup, I'll just say this: hearing is believing.

You don't happen to be anywhere near Alabama, do you? I'm about 10 miles off I-65. You're more than welcome to stop by after I get it all up and running. Assuming, of course, it's worth me auditioning.

As I said earlier, I heard a set of dipoles the day after I heard $15k Mark-Daniel speakers at an audio engineer's house. I preferred the dipole setup. Not sure if that tells you anything or simply says I don't know what I'm hearing.
Though, the mark-daniels use hiel atm tweeters which are flat out amazing. Beautifully defined and crisp top end without harshness. Basically like taking my favorite dome tweeters and sprinkling some 'awesome' on top.

 

[TLS Guy]

Having heard a great dipole setup, I'll just say this: hearing is believing.

You don't happen to be anywhere near Alabama, do you? I'm about 10 miles off I-65. You're more than welcome to stop by after I get it all up and running. Assuming, of course, it's worth me auditioning.

As I said earlier, I heard a set of dipoles the day after I heard $15k Mark-Daniel speakers at an audio engineer's house. I preferred the dipole setup. Not sure if that tells you anything or simply says I don't know what I'm hearing.
Though, the mark-daniels use hiel atm tweeters which are flat out amazing. Beautifully defined and crisp top end without harshness. Basically like taking my favorite dome tweeters and sprinkling some 'awesome' on top.

No, I'm in North Central Minnesota.

 

[90gstman]

Good to see someone bucking grade and doing something they are passionate about even though it may not be easy.

I'm currently putting together a semi-dipole 3 way that originaly started off as a typical boxed project so I'm realy interested in seeing how your project progresses. You are a much braver man then I in attempting to put together a dipole low end (who am I kidding, I want to try that too).

I am also taking a similar aproach as you in that I am currently using an active crossover with the final goal to be completely passive. Right now I have an active three way and am supplimenting with passive correction and equalization.

Linkwizt's site is very informative, I spent hours reading. There is also a good dyi semi dipole project on Troels Gravesen website called the Acapella (he uses the same Seas W18EX001 midrange as I have and thats what helped divert me to going dipole, well that and listening to a pair of Magnapan 3.6 at a local dealer).

 

[ErinH]

Good to see someone bucking grade and doing something they are passionate about even though it may not be easy.

I'm currently putting together a semi-dipole 3 way that originaly started off as a typical boxed project so I'm realy interested in seeing how your project progresses. You are a much braver man then I in attempting to put together a dipole low end (who am I kidding, I want to try that too).

I am also taking a similar aproach as you in that I am currently using an active crossover with the final goal to be completely passive. Right now I have an active three way and am supplimenting with passive correction and equalization.

Linkwizt's site is very informative, I spent hours reading. There is also a good dyi semi dipole project on Troels Gravesen website called the Acapella (he uses the same Seas W18EX001 midrange as I have and thats what helped divert me to going dipole, well that and listening to a pair of Magnapan 3.6 at a local dealer).

Thanks, man. I appreciate it.

I'm working on my car's system and (long story short) gutted my IB wall. Since I had that taken apart, I figured I'd throw the subs on the baffle from my car which is about 33x17" and houses two of the AE IB15 subs. I carried it upstairs, powered 'em up and measured the response. The subs were good to about 50hz and then rolled off at about 12dB/oct. That was comforting, though, given the extremely small size of the baffle.
Since the room is a dedicated music room, I can build as large a baffle as I want. I went out today and picked up a 4x4' sheet of 3/4" MDF that I will be putting a hole in for a single 15".
On the full range driver website, you can find some good predictions on baffle size required vs. frequency response.
http://fullrangedriver.com/singledriver/baffle.html
This says I need about 90" to get 40hz. I have almost half of that, lol. So, we'll see. With the DSP I have, I should be able to EQ the bottom end if needed. Further testing is all part of the game, though. We'll see. I'll post back what I find.

Gravesen's site is awesome. A friend sent me that link about a week ago and I've been surfing it a lot. Great source of info there for all sorts of things. Really cool site!


Thanks again.

- Erin

 

[ErinH]

Good read on use of H-frame and U-frame OB setups compares to standard OB:
http://www.quarter-wave.com/OBs/U_and_H_Frames.pdf

I’ve still got some testing of my own to do (specifically, assymetrical h-frames) but this is good info.