3-way OB

[Twisted85]

I'm still finishing my 2-way MTM project but I'm aldready planning my next project, which would be my first 3-way system.

I've been thinking of a using a open baffle midrange unit in this project. I've never made any ob speakers so I need some help and advices on this one.

I have something like this in my mind:

http://koti.mbnet.fi/twisted-/3way/model.gif

The front baffle would be tilted to compansate the acoustic center differences. I would use a 8" woofer probably in a reflex box. The midrange would be in an open baffle, the lighter gray in the picture is meant to be just a possible side brace (like a small U-frame).

At this moment actual issues would be to choose drivers, think appropriate placements for them in the baffle and chooce approximate xo point.

The woofer and tweeter are the easier ones, the midrange is more difficult. As a woofer I've been thinking maybe Peerless 830868 or 830884. I haven't had much thought with the tweeter, but ScanSpeak D2010 has been in my mind as an option.

I'd like to cross the midrange pretty high, somewhere around 3-5Khz, and midrange/woofer xo point would be around 400-500Hz.

Here's a simulation of Seas MCA15RCY in a baffle described above (without the side braces) as a possible midrange driver:

http://koti.mbnet.fi/twisted-/3way/response.gif

The Qts value is quite low, but if the xo point is around 400-500Hz this wouldn't be a problem, right?

Now questions about the OB that is unclear to me: How does it affect the driver sensitivity? Is the point where the frequency response starts to go down the same point where baffle step happens in a sealed box? above this point, is the sensitivity with OB and sealed box driver the same? If a driver would be set off axis (like in my scetch above) it would cause lobing to other side, is the effect similar with ob driver setup?
Why I've been thinking to cross the woofer/midrange around 400-500Hz is that this is the point where the midrange driver response starts to go down because of the OB. Is this a good idea?

Now about the midrange driver. Above I said that I'd like to cross the midrange/tweeter pretty high. I like the idea that one driver covers most of the midrange. The thing I've been thinking is the off axis response. Now let's take this driver as an example: Seas MCA15RCY. Let's say I would cross this driver at 4Khz. The on axis response is smooth, but the off axis responses are at lower level at this frequency. How does this affect in a open baffle design? Should I choose a driver and xo point where the response is smooth even measured from an angle (in an infinite baffle) or does this make a difference in a ob? If I would use the Seas driver mentioned, would it be more wise to cross it where the off axis is more at the same level, for example 3Khz?

All help, ideas and thoughts are appreciated.

 

[ttan98]

OB design

Last time I put my OB design here I got zero response.

If you get unsatisfactory response try this site there are many helpful people there.

http://www.audiocircle.com/circles/index.php?board=90.0

cheers.

 

[DougL]

I'd like to cross the mid range pretty high, somewhere around 3-5Khz, and mid range/woofer xo point would be around 400-500Hz.

I have a 3 way OB using am MCA15RCY. I cross it over on the low end at 320 Hz LR4, and currently have the high pass at 3000 LR4.Since I have a DCX2496, I have tried a wide range of values.
I really couldn't tell much difference crossing between 2 k and 4 k to a NEO3. My mid to woofer xo point was set by the desire to not excite the 2.1k resonance peak on my RS270 woofers, balanced with higher excursion at a low XO.
The size of your mid range baffle will also affect this.
I really think this driver works for an OB mid range, as long as you have a big enough baffle for your chosen XO point.

Doug

 

[Twisted85]

Re: OB design

Last time I put my OB design here I got zero response.

If you get unsatisfactory response try this site there are many helpful people there.

http://www.audiocircle.com/circles/index.php?board=90.0

cheers.

Thanks. I made a post to that site as well and waiting for replys.

If someone is willing to help I probably made my questions more clear at this another post so you could read this as well:

http://www.audiocircle.com/circles/index.php?topic=48622.0

 

[DougL]

Now questions about the OB that is unclear to me: How does it affect the driver sensitivity? Is the point where the frequency response starts to go down the same point where baffle step happens in a sealed box? above this point, is the sensitivity with OB and sealed box driver the same? If a driver would be set off axis (like in my scetch above) it would cause lobing to other side, is the effect similar with ob driver setup?

The Short answer is No, an OB isn't as simple as assuming the sealed box model works, and at about the same point.

An OB affects driver sensitivity for about 4 octaves above "F Equal"
there will be a peak of around 6 db about 2 octaves above "F Equal", transitioning back to infinite baffle sensitivity above that.

Below "F Equal", the driver has a roll off of 6 db per octave.
Below Fs, an additional 12 DB per octave roll off occurs.
This assumes a Qts of .7 Drivers of lower Q will have additional roll off as they approach Fs.

There are several sites that explain the basics, and several inexpensive tools that predict LF dipole behavior. MJK and JohnK come to mind.

Now about the midrange driver. Above I said that I'd like to cross the midrange/tweeter pretty high. I like the idea that one driver covers most of the midrange. The thing I've been thinking is the off axis response. Now let's take this driver as an example: Seas MCA15RCY. Let's say I would cross this driver at 4Khz. The on axis response is smooth, but the off axis responses are at lower level at this frequency. How does this affect in a open baffle design? Should I choose a driver and xo point where the response is smooth even measured from an angle (in an infinite baffle) or does this make a difference in a ob? If I would use the Seas driver mentioned, would it be more wise to cross it where the off axis is more at the same level, for example 3Khz?

Its a trade-off, just like other designs And like other designs, your polar pattern will affect your power response. I personally can't hear much difference in my current design. However, I have drivers of a similar size so the affect may be less pronounced. As always, YMMV.

Doug

 

[navin]

I've been thinking of a using a open baffle midrange unit in this project. I've never made any ob speakers so I need some help and advices on this one...All help, ideas and thoughts are appreciated. [/B]

One of the advantages of OB is the dipole bass that a true fullrange OB provides. The dipole bass has a lot less BOOM than traditional sealed and ported boxes. So if you considering OB it might be worth while to consider OB for the bass section too.

For sometime I too have been contempalting building a speaker that will replace my circa 1994 Audio Concepts SUBs + Scanpeak MTM combo. It sems most likely that I'll end up making a 3 way based around a superb wide range with helper drivers to cover the extremes.

Since the wide range is the fundamental driver it is this driver that is most important to choose. Common choices are Jordan (Jx92), Fostex (16x and 20x), the Visaton B200, Supravox 165 and Veravox 7x drivers. One could also consider co-ax drivers like the Tannoy, Ciare, or Hawthorne.

One option I am toying with is 2 10" Hawthorne Augie drivers mated to either a Visaton B200 + tweeter option or a Hawthorne 10" coax.

The music I listen to encompasses classical, jazz, blues and a lot of the popular music from the 60s and 70s (rock, pop, metal, and punk). I enjoy the coherrence of fullranges but dont want to completely sacrifice SPL or the ability to occasinally play Led Zepp at "live levels". I would love to know more from the Gurus here.

 

[SamL]

I would use a 8" woofer probably in a reflex box. The midrange would be in an open baffle, the lighter gray in the picture is meant to be just a possible side brace (like a small U-frame).

The single open baffle mid SPL will not be able to match the 8" woofer in reflex box. Unless you use a very sensitive mid or a low sensitivity woofer. Alternatively You can use 2 mid to increase sensitivity or you can reduce the sensitivity of the woofer but this will require an active system.

rabbitz who is active in this forum and audiocircle have build something similar using SS9700, 18W/8531G00 in open baffle and Vifa M21WO-39-08 in ported box. He uses sub amp to drive the Vifa. Unfortunately his site is no longer there.
AndyG who is also active in this forum and audiocicle have build something similar call Blackwood. His post about it.
http://www.audiocircle.com/circles/index.php?topic=38016.0

Have fun

 

[moray james]

here is some info...

http://www.t.kth.se/andzak54/speakersIntro.htm

 

[moray james]

here are two more sites...

http://www.digitalakuten.com/audio/ob/beyond_endorphine.html

http://www.lampizator.eu/SPEAKERS/Loudspeakers.html

 

[DougL]

The single open baffle mid SPL will not be able to match the 8" woofer in reflex box.

SamL, can you elaborate on this statement? The Seas MCA15RCY that is referenced has 90 db sensitivity, which should be sufficient to match an 8" woofer in any enclosure.

Doug

 

[SamL]

SamL, can you elaborate on this statement? The Seas MCA15RCY that is referenced has 90 db sensitivity, which should be sufficient to match an 8" woofer in any enclosure.

Doug

The 90db sensitivity is when the driver is place in a box. When place it in open baffle, it will lost 6db in sensitivity. Which is quite low to match with a woofer in a box.
With that and limited xmax from a mid driver, the mid will likely to bottom up before it reach the SPL you wanted.

 

[DougL]

The 90db sensitivity is when the driver is place in a box. When place it in open baffle, it will lost 6db in sensitivity.

This is not correct. Drivers on an open baffle do not loose sensitivity above F equal. The easiest way to convince you of this is to download a copy of xbaffle and run the simulations yourself.
Depending on the dimensions of the baffle, the output of a mid-range will be up to 6 db higher than a comparable mid-range in an enclosure.

To reinforce this, Seas measures most of its drivers on an open baffle. The results are the published data for the drivers on their web site.

Regards;

Doug

 

[SamL]

To reinforce this, Seas measures most of its drivers on an open baffle. The results are the published data for the drivers on their web site.

Are you sure? Maybe you are thinking about their tweeter but not the rest.

From Seas MCA15RCY PDF.
The frequency responses below show measured free field sound pressure in 0, 30, and 60 degrees angle using a
standard baffle (IEC 268-5) and a 1.7 litres rear chamber.

From Seas W21E002 PDF
Response curve recorded in anechoic chamber (Free-field, 4 pi radiation) with 0.5m microphone distance.
The loudspeaker is mounted in a closed box of 20 l net. volume

From Seas P25REX PDF
Response curve recorded in anechoic chamber (Free-field, 4 pi radiation) with 0.5m microphone distance. The
loudspeaker is mounted in a closed box of 28 l net volume.

 

[SamL]

This is not correct. Drivers on an open baffle do not loose sensitivity above F equal. The easiest way to convince you of this is to download a copy of xbaffle and run the simulations yourself.
Depending on the dimensions of the baffle, the output of a mid-range will be up to 6 db higher than a comparable mid-range in an enclosure.

Must say this is the 1st time I hear this. Mid in open baffle can be 6db louder than a box.

I though xbaffle only simulate the frequency response of a baffle, not SPL.

Have fun

 

[DougL]

Must say this is the 1st time I hear this. Mid in open baffle can be 6db louder than a box.

The 6 db reinforcement only happens at a single frequency.
This is about 2 octaves above F equal, and is dependant on baffle size. This is where the front and back waves have maximum reinforcement. Above this point, the baffle looks more like an infinite baffle. Below F equal, the front and rear start to cancel at 6 db per octave.

Regards;

Doug

 

[Twisted85]

I've now spent quite many days reading of open baffle speakers and tell you the truth I'm not sure if I can at this point make a well made OB speaker or should I still stick with more traditional designs. Open Baffle speakers really seems to be quite complicated to design.

But I got some thing more clear.

Music Design has quite a lot of good info about OB speakers at their website. I haven't had the time to read all of them but I've read some. Also quite a lot of it is quite mathematical and goes over my head.

But this graph about the radiation patterns became more clear to me:

http://www.musicanddesign.com/images/DPOB3.gif

D is the speration distance and W is wavelength. In one of my books about loudspeaker design it says that the D should be a bit more than half of the width of the front baffle. (assuming the driver is in the cented)

So I tried this with John K:s DipoleDesign program:

http://koti.mbnet.fi/twisted-/response3.gif

The baffle width is 100cm, so lets use 65cm for the D. The first null is at d/w=1. 0.65/0.65=1 and 344/0.65=529Hz, which seems to match with the response.
Now at d/w=0,5 there sould be a peak. 0.65/x=0.5 so x is 1.3. 344/1.3=264Hz which also matches with the response.

Now this I can understand, but setting the driver offset from the center makes all a lot harder. (I guess thats why I have the DipoleDesign to simulate the baffles )
Also I have no idea how it will affect off axis responses measured from different sides if the driver is set off-center.

Now I've also thought the option of making a full OB speaker as suggested. I got to know a little bit to the U and H baffles and found out that the U baffle isn't a real dipole because of the assymmetric design. There was also disscussion about dampened U-Baffle but I didn't really get hang of it.

Does anyone know how the separation is defined in John K's DipoleDesign with U and H baffles?

For example, if I want to simulate these U/H frames:

http://koti.mbnet.fi/twisted-/uhbaffles.gif

How much is the separation in this H or U frame?

If I use very high sensitivity woofer, can I lower it down with L-Pad or is there a problem with this? I was thinking I could use a high Qts driver (like 1-1.2) so I wouldn't need any equalization.

At this point If I'm sticking with similiar design I presented in my first message but use a open baffle woofer the 100-300Hz area becomes problematic. A big high Qts woofer like Eminence alpha 15" could be used under 100Hz, but a 3-6" midrange driver won't go to 100Hz in a open baffle width of 40cm. This whole design is still under heavy development and has quite many problems to dissolve if I'm going to ever make it

 

[navin]

I've now spent quite many days reading of open baffle speakers and tell you the truth I'm not sure if I can at this point make a well made OB speaker or should I still stick with more traditional designs. Open Baffle speakers really seems to be quite complicated to design....

Now I've also thought the option of making a full OB speaker as suggested....

I was thinking I could use a high Qts driver (like 1-1.2) so I wouldn't need any equalization....

At this point If I'm sticking with similiar design I presented in my first message but use a open baffle woofer the 100-300Hz area becomes problematic. A big high Qts woofer like Eminence alpha 15" could be used under 100Hz, but a 3-6" midrange driver won't go to 100Hz in a open baffle width of 40cm. This whole design is still under heavy development and has quite many problems to dissolve if I'm going to ever make it

With OB as with any other design proper choice of drivers is essential. You cold mate your Eminence woofer to a widerange like the Visaton (which also has a Qts of about 0.8) and have a simple 2 way full range system with a bafle width of about 40-50cm. If you feel the need for a bit more air in the HF there are quite a few tweeters you can try (Fountek/Aurum Cantus ribboins are worth looking at here).

If you then want to upgrade to a souped up version of the woofer and widerange the Hawthrone 15" Augie (they have 2 versions) mated to a Supravox 215 Signature can be one option.

Hope this helps.

 

[Rudolf]

Also I have no idea how it will affect off axis responses measured from different sides if the driver is set off-center.

The dipole eight will be bent to the "longer" side of the baffle, but polar radiation will be more homogenous to the "shorter" side. Have a look at some simulations I did.

Does anyone know how the separation is defined in John K's DipoleDesign with U and H baffles?

DipoleDesign has already calculated that separation distance for you. It´s the "Dia of equivalent circular, cm" in "Baffle dimensions"

How much is the separation in this H or U frame?

It is the way the sound from the back of the cone has to travel until it is in the same plane as the front of the cone.
For the U-frame this would be about 40 cm to the back of the frame plus 40 cm to the front baffle.
Fr the H-frame it would be about 20 cm to the back of the frame plus 20 cm to the baffle with the driver.

 

[Twisted85]

The dipole eight will be bent to the "longer" side of the baffle, but polar radiation will be more homogenous to the "shorter" side. Have a look at some simulations I did.

This was very interesting, thanks. So setting the driver offset will smooth the on axis response and it will radiate more to the longer side, but the response is better to the shorter side.

I simulated B200 driver so it would be placed as high as possible in a 100x42cm baffle and this seems to smoot a bit the first null point, but you mentioned in the other thread that there is also problems with this?

What about if any midrange driver is set to the center in horizontally and 10-20cm from the top of the baffle, how is the peak and null normally controlled in this situation? By active equalization? At this point I want to make fully passive speaker without equalization and I try to find out how (if it is) it would be possible.

What about if midrange driver is set to U-frame baffle? Does this resolve any of the peak/null problems?

The Visaton B200 is an interesting driver. It is rated 96dB/2.83V which sounds great, but judging by the frequency response this seems to come true only at frequencys above 1000hz. Does the top end usually need attenuation with this driver?

 

[DougL]

The Visaton B200 is an interesting driver. It is rated 96dB/2.83V which sounds great, but judging by the frequency response this seems to come true only at frequencies above 1000hz. Does the top end usually need attenuation with this driver?

The B200 is designed for OB. On a baffle with optimum D, the Baffle will provide a boost to frequencies under 1000 Hz to have a flatter response over a range of about a 100 Hz to over 8 K Hz. There are people putting either correction circuits (active or passive) or phase plugs to tame the high end.

My B200 are in a box for a couple more days / weeks.

Congratulations on your research. Your questions show understanding of the issues.

Regards;

Doug