Winter's OB project - feedback requested

Objectives:
110dB SPL continuously capable
All drivers minimally baffled
Utilize dipole tweeter
Utilize monopole subs for difficult last few octaves

Drivers:
Bass - AE Dipole 15, 70Hz-200Hz
Mid - 2 x Seas W15, 200Hz-1800Hz
Tweet - Mundorf AMT2310 1800Hz+

Mid/Tweeters arranged in correctly spaced MTM arrangement. High slope linear phase XO between mid and tweeter. Monopole JL Audio F112 subs used below 70Hz, 24db/oct LR XO.

Other equipment:
DEQX HDP-3
Bel Canto EVO200.6 (6 channel amp)

Construction:
3" MDF sandwich for the back brace, 1" ply for the bass panel, 3/4" for the MTM panel. 1" thick granite stone for base weight.

Any feedback on driver choice, XO points, construction, anything else?

This is my first full from scratch design, although I did put together and currently own the Linkwitz Orion. I have read as much as I can of the Linkwitz site and also John K's site, incl. reviewing his dipole design guide.

[IMGHTTPDEAD]http://www.acousticfrontiers.com/storage/Genie%20drawing.PNG[/IMGHTTPDEAD]
 
Last edited:
winter ob project

I like your design a lot. Especially the bass section. Love the back brace too. Could I suggest that you might want to curve that backside of the brace up to meet at the top of the upper panel? I think that being supported only at the bottom, no matter how thick the back brace below....it is going to wobble like crazy.

I have not listened to the driver selection, so cannot comment on that part of it. Good luck to you!
 

CLS

Member
2005-06-17 6:58 am
Taiwan
Beautiful:D

In such design, for "defending" the vibration, I think soft is better than stiff. It's nearly impossible to make a real mechanical grounding to make the whole speaker vibration free. On the contrary, soft mounting can lessen or totally eliminate the vibrations. The softest is by suspending the drivers by strings.

Keyword: "swing" ;)
 
I like the design, certinally looks much easier than the one I can up with. The information supplied by others in my thread suggested that the Dipole 15 would get down into the low 30's and save all of the extra work with the sub. I guess if you want to go below that then the subs a must. Oh my 15 is going in a H frame which makes a difference.

Sorry I cant be of much help as I still have my "L" plates on. good luck I will follow with interest.:)
 

StigErik

Member
2008-02-21 12:30 pm
Nice design, I like it. Mid and tweeter selection is right up my alley... :)

A few thoughts though...
1. To get the best (horizontal) off-axis response, the tweeter should be as narrow as possible. The best is using the driver "nude" - no baffle or frontplate.
2. Why not extend that "arm" holding the mid/tweeter baffle all the way up, and suspend (hang) the mids and tweeter from there? Much better decoupling of vibrations.
3. 200 Hz is too low XO point for the mids, I would go up to 300-350 Hz.
 
Thanks everyone for the feedback!

The 2510c has better off axis behavior but is larger. On the other hand you can remove the face plate.

Using the DEQX or anything else ?

The 2501 would be too big (too great driver separation for the MTM design and XO point)

Will be using the DEQX for linear phase XO

I like the design, certinally looks much easier than the one I can up with. The information supplied by others in my thread suggested that the Dipole 15 would get down into the low 30's and save all of the extra work with the sub. I guess if you want to go below that then the subs a must. Oh my 15 is going in a H frame which makes a difference.

My modeling shows the Dipole 15 is displacement limited on that baffle size at a 70Hz crossover point whilst meeting the 110dB SPL requirement. I will get quite a bit more SPL than the model shows though due to the 2pi radiation (+6dB). I wanted to keep a bit of flexibility with XO point experimentation. Running down to 30 would require a H frame or multiple drivers. In any event I am a strong convert to leaving low bass reproduction to a monopole subwoofer in the interests of space!

A few thoughts though...
1. To get the best (horizontal) off-axis response, the tweeter should be as narrow as possible. The best is using the driver "nude" - no baffle or frontplate.
2. Why not extend that "arm" holding the mid/tweeter baffle all the way up, and suspend (hang) the mids and tweeter from there? Much better decoupling of vibrations.
3. 200 Hz is too low XO point for the mids, I would go up to 300-350 Hz.

1) I have seen your nude speakers, what kept me from doing that rather than the minimal bafffle was the added complexity of driver mounting...
2) That is a possibility. How do you hang drivers?
3) My excel modeling using John Ks tool shows I can get 118dB SPL using a LR4 200Hz XO since the W15s have very good XMAX (6mm). Is there some reason why I couldn't? I looked at using a single W15 but although XMAX was good power handling would have been an issue.
 
To design a crossover that will lead to seamless transition between all those drivers is going to be a nightmare. Try reducing the number of drivers. Going with a larger pro midrange may be an option to meet your SPL needs.

Its hard to imagine that using the excellent AE driver gets you down to just 70 Hz on an OB. I've seen them flat to 20 Hz, but may be not at 110 db. I believe there is enough directional information at 70 Hz that a separate sub just does not cut it.
 
Here's what I'm currently listening to:

183802d1281625321-my-current-ob-experiment-img_1680.jpg


AE OB12, 18Sound 6ND430, B&G Neo3. XO @ 250Hz and 1.8kHz with a DCX2496, and then 60-80Hz to a sub. I'm not shooting for anywhere near your target SPL levels.

I would put some thought into mechanical decoupling / isolation. The MT panel on this vibrates with just the woofer playing.
 
Here's what I'm currently listening to:

183802d1281625321-my-current-ob-experiment-img_1680.jpg


AE OB12, 18Sound 6ND430, B&G Neo3. XO @ 250Hz and 1.8kHz with a DCX2496, and then 60-80Hz to a sub. I'm not shooting for anywhere near your target SPL levels.

I would put some thought into mechanical decoupling / isolation. The MT panel on this vibrates with just the woofer playing.

That must have been where I got the inspiration from - I looked at a lot of designs then just went into my CAD package and had a play around.

If you did your design again how would it be different (given that it is similar to mine from the construction side)?
 
To design a crossover that will lead to seamless transition between all those drivers is going to be a nightmare. Try reducing the number of drivers. Going with a larger pro midrange may be an option to meet your SPL needs.

Its hard to imagine that using the excellent AE driver gets you down to just 70 Hz on an OB. I've seen them flat to 20 Hz, but may be not at 110 db. I believe there is enough directional information at 70 Hz that a separate sub just does not cut it.

I'm going to use dual subs (I already own the JLs) because I agree with you on the fact that there is important directional information at 70Hz. I read an AES paper that said 50Hz was about the lowest we could hear directionality.

Not sure how I could reduce the number of ways without either compromising SPL, horizontal polar response or going to a different construction (h frame, u frame, w frame) for the o/b bass section which I don't want to do. The only other option I thought of was using a higher efficiency midrange like a PHL 1120 which would eliminate the need for an MTM and running the AE up higher. However I really like the Seas mag cones I think they are awesome in the midrange... I actually think a theoretically perfect open baffle would approach a 5 way design using conventional drivers, even doing a 3 way has drawbacks in the low bass (large excursion required from ob) and lower treble (increased directionality above baffle peak). The Orion is now a 4 way, and that is my reference, so a 4 way seems a good starting point. Always open to other ideas though!
 
Last edited:
That must have been where I got the inspiration from
Heh... I decided not to say "Hey you copied my design idea" ;)

If you did your design again how would it be different (given that it is similar to mine from the construction side)?

That's a good question. For starters, I had a friend with a DIY CNC machine cut the pieces for me - this is way beyond my carpentry skills. Consequently, I also had to do this in a way that fit the limitations of his CNC machine. Specifically, he doesn't have the accuracy to flip the piece over and cut on both sides, so I had to plan for cuts that only needed to be made on one side.

I think my biggest pain was during the glue-up - the bezier curves on the spine piece look good, but it also means there are no parallel surfaces to attach a clamp to.

Other than that, I've had a few thoughts:

* I did a poor-man's mortise/tenon joint - cut out slots in the panels, and the end of the spine piece is sized to fit into the slot. I'd do something a little better there.

* Plan for cable management - the best option would be to build each piece as a sandwich, and have internal 'tunnels' to send the wires through. This will make the construction more complex, since I won't want the channels to fill up with glue. Maybe run the cables during the glue-up.

* Think about separate spine / reinforcement pieces for the upper and lower baffles. Or some other technique for minimizing vibrations. Maybe magnet mounting.
 
:confused::confused:
Could you please elaborate on this ?
Linkwitz uses sealed monopole subs crossed at 50Hz in his Orion ++ setup (which is his normal everyday listening setup). Therefore it is a four way!

Don't agree in general because for conventional systems this comes at cost of the vertical response/directivity.
True. But then engineering is all about compromises and there are no hard and fast rules about what is more important. The models I have seen for vertical polar response errors due to MTMs or multiple drivers spread out vertically all assume that the listening room is basically a reverberant chamber (perfectly adding acoustic power from all directions), which it most definitely is not.