[bellicon]

This will be my first but I hope very successful venture into the OB world. I’m  very intrigued by what’s often said about their sound. When properly designed and built they can deliver sweet, airy highs, without shrill or brightness. Like ESLs! A lifelike and involving midrange. A large 3D sound stage. Robust and articulate midbass. And I suspect that there are also other sonic qualities that are unique to the OB sound. 

But there’s only one guy I know of who builds OBs; he’s actually local but didn’t get back to me about auditions.

However, can I have great sounding OBs my way? The biggest obstacle is that I prefer to avoid active crossovers. They certainly have their advantages, which is why most OB designers use them. But I’m in a condo and the extra cost, power and heat management issues would make them undesirable in my situation. Instead, I’m hoping that with careful driver selection and crossover design flat panel or H-Frame OBs will yield superb performance, with lows down to 70Hz.

I would think that limiting bass response to around 70 to 100Hz would minimize distortion in the critical midbass and prevent the woofers from sounding like slugs, as Danny Richie might say. My pair of Rythmik 12" sealed servo subs could take over below this.

The footprint of each OB can be as much as 26” wide and 15” deep; H-Frames or flat panels with wings.

I have no problem spending serious money on drivers and crossover parts (e.g. Dave Slagle’s coils and autoformers yes but not $800. each for Duelund caps).

My solid-state Class A amplifier does 30/60 wpc/8/4 ohms, handles 60 w peaks all in Class A, and delivers far more, if less alluring, power in Class AB. My Rythmik subs each have >300 watt Class AB amps. But my room is only 14 ft x 19 and my ears can’t take average midband SPLs much above 65db at 11 feet, if that much. So unless the OBs prove to be fairly heavy loads then there should be ample wpc from my main amp.

Among those drivers, which seem quite popular for OBs are Tang Band (midrange), Eminence (Beta 12” LTA, with dust cap/phase plug mod for midrange), Supravox (http://www.supravox.fr/kits/panneauplan.pdf ) , SEAS (tweeter), GPA  (for midbass http://www.greatplainsaudio.com/downloads/416-8B%20Spec%20Sheet.pdf ).

For a tweeter, some have suggested a 1” Radian compression drivers with beryllium diaphragms. But with 1” horns what you might gain in holographic imaging would you lose in sound stage extension?

If these and or other suggested drivers are used in passive OBs with system efficiency of at least 90db, impedance that doesn’t drop much below 7 ohms and bass limited to between perhaps 100 and 70Hz, my amp should be kept within its Class A bias region, and avoid its otherwise lean sounding midbass performance.
 
Thoughts? Suggestions? Design/Build services?

[JohnR]

My suggestion would be that you either find a done design or kit, or if you want to roll your own, start small (in cost/complexity not necessarily size).

[FullRangeMan]

The traditional success OB receipt are 1 Alpha15A and 1 Visaton B200 with an 4,7mH ribbon inductor that you can get at Sonic Craft w/Jeff.
http://www.soniccraft.com/product_info.php/pair-alpha-core-12-awg-470mh-p-5026
Also there is the Eminence Red Coat Big Ben 15 inches if you want try something new (inductor is the same).
W/the Alpha you will get response down to 60Hz.
Forget tweeters and 3 ways xovers.

[JCS]

Take a look at:

http://www.quarter-wave.com/OBs/OB_Theory.html

And here is a link to a discussion of my build using Fostex FF85k (now superseded by the even better FF85wk, although the sensitivity is slightly different) and Eminence Alpha 15a:

http://www.diyaudio.com/forums/full-range/155019-jim-shearers-fostex-ff85k-eminence-alpha-15a-open-baffle-speaker.html

[richidoo]

With passive crossover you have two options for bass: Use high Qts bass woofer with no EQ but sacrifice transient response (bass detail,) or use low Qts bass woofer with added EQ bass boost in your passive crossover. You'll have good bass detail, but low Q drivers are much more expensive, and the added complexity of the EQ in passive xo, and you need a lot more powerful amp to compensate for the headroom lost in the passive bass boost EQ. You can see why high Q woofers like Alpha are popular with OB diy experimenters.

Servo bass is not passive, but since you're open to it that's what I would recommend. You already have the amps, so just buy some dipole version servo woofers and you'll have the FR extension of high Q drivers, with the good transient response of low Q, and no additional EQ.

The "articulate bass" you mentioned is a benefit of dipole radiation pattern, but the poor transient response of high Q bass drivers negates much of the benefit, imo. If you spend the money to have well controlled bass drivers, either low Q w/EQ, or servo, then you get the full benefit of a dipole bass - good detail and good room acoustics. Dipole bass does lack the punch of traditional box woofers though, ymmv.

[PDR]

I'm (yes still) breaking in a set of TB 1808 for a DIY PAP trio.
I have the four Eminence sitting in boxes, just trying to find the time to get them all together.

I have managed to get the 1808s on two flat boards, and using the subs from my Super vs......
I use a 80Hz high pass....

If you have a set of GR OB subs.....I'd recommend the TB for a starter kit.
I think you'll be surprised what this combo has to offer.

Edit:  Just re-read and see your Rythmiks are sealed....

 

[MJK]

Use high Qts bass woofer with no EQ but sacrifice transient response (bass detail,) or use low Qts bass woofer with added EQ bass boost in your passive crossover.

I have used a few different high Qts drivers in dipole systems with a SS amp, passive and active crossovers, and never felt that I was sacrificing anything with respect to transient response. When you use a low Qts driver and EQ are you not just making it behave like a high Qts driver? The desired SPL curve is the same in both cases.

[JohnH12]

There are still a few of the Titan sub woofers available if you want to build your own

http://projectgallery.parts-express.com/speaker-projects/titan-ob/

The total project can come in at $300.  Which gives you a chance to try OB for little $.


John

[richidoo]

When you use a low Qts driver and EQ are you not just making it behave like a high Qts driver? The desired SPL curve is the same in both cases.

Hi Martin.    I don't think it's really behaving the same. The FR might look the same but EQ doesn't affect transient response, while damping does. A high Q driver plays low and loud without EQ by sacrificing mechanical and electrical damping, but when it blends with a more detailed midrange driver the lack of damping can be heard. I personally find high Qts bass drivers in dipole a little bit lacking in tonal texture, ymmv. But it is truly amazing how they can play flat to 30Hz with no box and no EQ!

[MJK]

I don't think it's really behaving the same. The FR might look the same but EQ doesn't affect transient response, while damping does. A high Q driver plays low and loud without EQ by sacrificing mechanical and electrical damping, but when it blends with a more detailed midrange driver the lack of damping can be heard. I personally find high Qts bass drivers in dipole a little bit lacking in tonal texture, ymmv. But it is truly amazing how they can play flat to 30Hz with no box and no EQ!

If the FR is the same then the transient response is the same, I don't think it matters how you get the FR. Take the inverse FFT of the SPL curve to get the transient response, the driver and the EQ are both contributing to the transient response.

The Qts of the driver only comes into play around fs, above fs where the driver is transitioning to a mid or full range the high Qts driver's response is controlled by the cone mass. The cone construction may lead to break-up modes which will ring and can be heard messing up the FR. In general large high Qts drivers tend to have light paper cones compared to low Qts drivers of the same size which have heavier thicker cones. The light cone helps the low Qts drivers play loud with a small magnet. The peaking at fs associated with a high Qts driver is mitigated by the rolling off caused by the baffle, the combined effects decrease the effective system Q to be in line with a boxed speaker (but with an 18 dB/octave roll off), there is no excessive ringing bass if you have made the correct choices for driver and baffle geometry.

My experience has been that using a light cone high Qts driver over a limited frequency range to produce only low bass (< 200 Hz for example) is a valid approach for a passive OB speaker system, you don't sacrifice anything. Playing these same drivers over an extended range to try and mate with another driver above say 500 Hz is a gamble, I am very skeptical of light cone high Qts dipole woofers that cross over to a small tweeter or compression driver in the 1000 Hz+ range.

It is all in the trade-offs.

[richidoo]

Very cool! Thanks for explaining. I'll digest.

My experience is with 2 different very high quality drivers with high Qts .7 designed for OB. Both had poor transient response throughout their pass band. I correlated that with the high Q. Even at low freq comparing the bass detail to a box speaker with low Q driver was much more detailed.

edit: added "different"

[bellicon]

Thanks all for chiming in with lots of great info, precautions and options. I’m gathering more info to follow one design path right now. I’ll be back to bounce it around here with more questions before I start buying parts.   

[matevana]

Hi Martin!

Great to see some posts from you again. Since the OP has already received some good advice from the community, may I ask what you are working on these days? The last time we spoke, I believe you were thinking of revisiting some type of horn design. Anything in open baffles that has tapped your interest lately?

Thanks,

Ed

[MJK]

Great to see some posts from you again. Since the OP has already received some good advice from the community, may I ask what you are working on these days? The last time we spoke, I believe you were thinking of revisiting some type of horn design. Anything in open baffles that has tapped your interest lately?

Hi Ed,

I don't read or post very much on public forums anymore. Occasionally I will add a comment if something strikes my interest.

I continue to think and work on horns starting back at square one and re deriving the acoustic theory in my own notation and with my own modifications. I have an outline of a presentation of the work that needs to be fleshed out and eventually loaded onto my site. It is looking like a long drawn out process.

I updated all of my MathCad worksheets a few times over the past year or two. All of the worksheets are at least two way with series and parallel passive crossovers or an active crossover option. I added a tilted baffle capability to the dipole worksheets in an attempt to mitigate the floor bounce at higher frequencies. I have done some more work on the baffle edge sources and improved the calculation of the baffle step or dipole response. I stay busy.

I have not built anything in a while, I had a serious knee injury just over two years ago and I have not fully recovered after they put it all back together (it may never be 100% again). So I have been slowed a little bit this past year, but this winter I should be back cutting wood.

Thanks for asking,

Martin