[FredT300B]

These speakers already exist in the form of the DR912a speakers, which are the same speakers in a bass reflex configuration. I've decided to build a pair of open baffle speakers using the same drivers: nine BG Neo8 PDR tweeters with twelve 7" Dayton Audio RS180 midwoofers per side. The baffles are completed and I'm in the process of designing and fabricating a support assembly, which will consist of a base with 4' diagonal braces to damp the forward/backward motion of this very tall baffle. In the current bass reflex version the midwoofers are crossed to the planar tweeters at 1.2khz using a 48dB/octave slope. The subwoofer, a bass reflex TC Sounds Epic 12 is crossed to the midwoofers at 120hz, also using a 48dB slope.

Both the old bass reflex and the new open baffle versions are/will be tri-amplified using a DEQX digital processor for control preamp, active crossover, speaker equalization and room correction functions. For the open baffle DR912ad version I plan to replace the TC Sounds Epic 12 subwoofer with two H-frame Eminence Alpha 15's. The crossover frequency will be determined after measuring the natural rolloff of the midwoofers and subwoofers. If the concept is successful I plan to add one additional Alpha 15 per side.

Both the old and new versions stand 7' 8" tall. The old baffle is 14" wide. This was increased to 16" for the OB version to provide additional support alongside the 6' tall tweeter slot. Hidden behind the baffle on the OB version are three full length 3/4" mdf vertical braces plus a full width 3/4" crossbrace between every other midwoofer for additional support.

For those who haven't figured out the name, the DR stands for Dayton Reference, the 912 is for the nine tweeters and twelve midwoofers, the a is for active crossover, and you figure out what the d signifies in the dipole version. 

[this_is_vv]

.interesting build...

[Brad]

Hey Fred - those are barely going to fit into the door at LSAF 

Nice project - any estimate on the sensitivity?

[FredT300B]

Nice project - any estimate on the sensitivity?

The sensitivity is limited to the sensitivity of the Neo 8 PDR tweeters, which is in the low 90's. I'm using tri amplification, so matching sensitivity for the woofer and tweeter arrays isn't required, but I've discovered their loudness levels are about the same when the woofer arrays are wired for 12 ohms (three parallel-wired groups of four series-wired speakers per group) and the tweeter arrays are wired for four ohms (three parallel-wired groups of three series-wired speakers per group). In spite of this relatively high sensitivity, these speakers prefer a moderately powered solid state amp. In their present state each speaker is driven by a 120w/ch stereo amp, and at very high volume levels with highly dymamic classical music, I suspect double that power would be better.

[FredT300B]

Finally nearing completion. I'm waiting for a 100' spool of OFC wire and 100 quick disconnects to arrive from Parts Express. Twenty four woofers and eighteen tweeters take a lot of wire. For those who haven't been following the thread, they stand 8' 9" tall, 16" wide, by 2-1/2" deep. The twelve Dayton Audio RS180 midwoofers and nine BG Neo8 PDR tweeters on each side are bi amplified using a DEQX preamp, equalizer, room correction and active crossover, and four channels of an Emotiva XPR5 amplifier. The subwoofer is a single mono TC Sounds Epic 12 driven by a Datyon Audio 500w plate amp. I'm building a couple of H-frame OB Eminence Alpha 15's to audition in place of the ported Epic 12. The midwoofer to tweeter crossover is a 48dB/octave at 1,200hz. The sub to midwoofer frequency will be determined after testing the midwoofers' rolloff in their ob configuration.   

[Russell Dawkins]

This might be a configuration where you should consider multiple subs. As Duke LeJeune pointed out to me, mixing a line source or any large source with a point source for part of the spectrum will give you a system that can only be correct at one distance. Due to the fact that the sound pressure from a long line source falls of at a linear rate when you are in the nearfield (as you will be in this room) and falls off at a logarithmic rate from a point source, if you balance the outputs of the lines and the sub at one particular distance they will be too loud if you move farther away and too quiet when you move closer.
The only really kosher solution is stereo line array subs, but a reasonable compromise could be multiple cheap subs scattered around the room. This would also carry the benefit of minimizing the dominance of any particular set of room modes, i.e., the bass would sound more or less the same everywhere.

[FredT300B]

The only really kosher solution is stereo line array subs, but a reasonable compromise could be multiple cheap subs scattered around the room. This would also carry the benefit of minimizing the dominance of any particular set of room modes, i.e., the bass would sound more or less the same everywhere.

Russell, is this true if you're using dipole subs? Also, to what extent would the room effect problem be mitigated by the DEQX room correction feature?

[Russell Dawkins]

Russell, is this true if you're using dipole subs? Also, to what extent would the room effect problem be mitigated by the DEQX room correction feature?

It doesn't matter whether they are dipole or not, it's to do with the size of the sound source. I don't think that DEQX would have any effect that relates to this. It's all to do with the fact the line arrays are 8' tall - effectively 16' when the floor reflection is accounted for, so until the listening distance is more than 16', the attenuation rate as you move away is linear at 3dB per doubling of distance. The size of the sub source is about 2 feet, including the floor reflection, so once you are more than 2 feet away from the sub, the spl falls at a rate of 6dB per doubling of distance.
Math is not my strong suit, but this gives you the general idea.
DEQX will be able to do it's thing at any particular distance but, more than conventional arrays, results will hold over a more limited area.
If it were me, I would look into multiple cheapish subs, after getting used to what it sounds like with the single sub - unless you have the time, space and money to build a line array of subs.

This is all theory and speculation, since I have not actually been in your position, but it makes good sense to me and I thought I would alert you to the idea, because it was  a to me when Duke told me.

P.S. If you own the house and funds permitted, the simplest solution, since you have DEQX, especially, might be to create 8' line arrays of subs in the wall behind the main speakers, as long as what was behind the wall allowed this. Alternately two OB line arrays of 4 18" Goldwoods in elongated H-baffles would be fun!

Here's a little more info on line arrays: http://www.prosoundweb.com/article/print/everything_you_wanted_to_know_about_line_arrays_and_thanks_for_asking

[FredT300B]

Russell, thanks for the information. A distributed "swarm" of subs are the preferred solution, but they are not an option for this room. Neither are in the wall speakers. A tower of H frame subs would be unacceptable ugly in this room. I can have one or two subs positioned near the speakers.

[Russell Dawkins]

Well at least you will have some idea of what to expect. Have fun!

[JohnR]

Hi, I'm not sure that the 3/6 dB fall-off with distance really holds at subwoofer frequencies (in a room). Here are some measurements that I took for a different purpose but may illustrate. I'm moving the sub away from the listener rather than the other way around, and all rooms are different so YMMV etc...

This is a monopole (sealed) sub at 1 (green), 2 (orange), and 3.3 (black) meters from the listener (right on the wall):



This is a dipole sub at 1 (green), 2 (orange), and 3.0 (black) meters from the listener (as close to the wall as possible):



I'd not looked at these measurements quite this way before, but this example suggests that the dipole tends to show the 6 dB/distance behavior, but the monopole doesn't. It also confirms that dipoles need to be kept away from the wall, but monopoles are better off right on the wall (if you can't put them next to the listening seat). In either case, room effects can swamp these generalizations (e.g. 50-90 Hz in this room).

I like putting dipole subs close to the LP. It solves so many problems in one fell swoop! From looking at your photo, Fred, if the subs need to be near the speakers I wonder if sealed subs behind them might not be the best way to go.

With regard to the line array of subs, it's been a while since I looked at it but as I recall you only need two drivers to satisfy the minimum spacing criterion i.e. at 25 and 75% of room height. But, you can't satisfy the line length criterion because the effective virtual length (3x ceiling height) isn't long enough anyway.

BTW, nice build Fred, I'm looking forward to hearing more about it

[dBe]

This might be a configuration where you should consider multiple subs. As Duke LeJeune pointed out to me, mixing a line source or any large source with a point source for part of the spectrum will give you a system that can only be correct at one distance. Due to the fact that the sound pressure from a long line source falls of at a linear rate when you are in the nearfield (as you will be in this room) and falls off at a logarithmic rate from a point source, if you balance the outputs of the lines and the sub at one particular distance they will be too loud if you move farther away and too quiet when you move closer.
The only really kosher solution is stereo line array subs, but a reasonable compromise could be multiple cheap subs scattered around the room. This would also carry the benefit of minimizing the dominance of any particular set of room modes, i.e., the bass would sound more or less the same everywhere.

You are in the nearfield of a subwoofer when you are closer than the wavelength of the highest frequency that the sub will cover.  In other words if you are crossing over the sub at 80Hz, then you are in the nearfield at anything closer than 14 feet-ish.  Even at that being a little farther away wouldn't be that bad due to the 60-80Hz area being the realm of muddy bass.  &0Hz just sucks in this regards, so any room with dimensions on multiples of 4' is going to have this to a point.  Good old standard dimensional lumber at work.

Of course you then have room gain that all rooms have at distances related to the longest dimension of the room, boundary gain and other issues like modes, etc.  Some of these things can be mitigated to a point with an open baffle sub not to mention the fact that conventional sub is going to have boxiness and other artifacts that will forever tell you there is a sub at work.  Swarms will help to a point, but the best ones are OB swarm subs.

Open baffle subs rule and are the only thing to have with an OB speaker IME.  Just sayin'.

Dave

[FredT300B]

Hi, I'm not sure that the 3/6 dB fall-off with distance really holds at subwoofer frequencies (in a room).

John R, thanks for the measurement data. Most people who don't have access to measurement equipment tend to underestimate the differential effect of the room on varying frequencies or the level change rate at various distances from the listener. One measurement will usually dispel any notion that the in-room response of any speaker will even remotely resemble the theoretical anechoic response. 

The 3dB versus 6dB argument comes up almost every time I post a message about line arrays, especially where a point source driver will be combined with the array, such as the subwoofer with the DR912ad's or the Sweet 69's with their 12 woofers and one centrally positioned dome tweeter. In every instance I am warned about the dire consequences of combining a line array with a point source. In my case the in-room positions of the subwoofer, the line arrays and the listening chair never change, and I measure the frequency response, dial in the subwoofer level, and do the room correction with the mic set at the listening position. So the comments about the 3dB/6dB thing are theoretically correct but irrelevant in my situation.

Here's some interesting reading for those who might want to learn more about line arrays:  http://www.diy-audio.narod.ru/litr/nflawp.pdf

[FredT300B]

Open baffle subs rule and are the only thing to have with an OB speaker IME.  Just sayin'.

Dave

Based on my subjective listening experience to other open baffle speakers at audio shows, I would agree with you. I believe the ultimate subwoofer system for these speakers would be a couple of open back H-frame 12" servo subwoofer drivers per side with their own servo amp. They could be positioned beside or slightly behind the towers. The servo subs seem to provide the best combination of power handling and control which many open back subwoofers seem to lack.

[dBe]

Based on my subjective listening experience to other open baffle speakers at audio shows, I would agree with you. I believe the ultimate subwoofer system for these speakers would be a couple of open back H-frame 12" servo subwoofer drivers per side with their own servo amp. They could be positioned beside or slightly behind the towers. The servo subs seem to provide the best combination of power handling and control which many open back subwoofers seem to lack.

Yup.  I don't want to get in trouble by recommending one of ACs main contributors and site owners products, but the servo subs that he sells are amazing.

Dave

[matevana]

Short of a servo amp/sub combo, an H frame with a well damped amplifier and a woofer with world class distortion levels and minimal mechanical noise are hard to beat.  Take a look at some power amps with a high damping factor (in the desired range of operation) such as some vintage B&K or Parasound offerings. They seem to exert almost as much motor control as a servo, but have better depth than many of the servo amps I have heard. Regarding a driver's motor noise, I have had great luck with some Usher and SB Acoustics drivers mounted in an H Frame.

Regarding an amps damping factor, it can be a very difficult to measure but the audible results in OB are huge. It has more of an effect on sonic properties such as attack & decay then it does on measuring frequency response aberrations, etc.  Some of the otherwise nice sounding T-amps, for example, have poor motor control and sound really sloppy in OB, IMO. 

[Bob in St. Louis]

Those speakers look like so much fun to build, I'd like to build them just for the sake of building them. 

Looks great though Fred! 

Any idea how much you'll have invested in them when they're finished?

Bob

[FredT300B]

Any idea how much you'll have invested in them when they're finished?

Bob

For the line arrays, the 18 BG Neo8's and 24 Dayton RS180's will set you back about $2,850. Add some mdf, wire, speaker terminals, screws, quick disconnects, paint, etc and you're still under $3K. In the original ported enclosures a subwoofer wasn't absolutely necessary. In the current ob configuration it's an absolute must have, so add the cost of the sub or subs of your choice. You'll also need four channels of amplification and an electronic crossover. I'm using a DEQX ($3K) for crossover duty, but you could use a Behringer DCX 2496 for $350 or a dbx DriveRack PA+ for $500.

If this is too many $$$$, you missed the best opportunity to build an under-$100 pair of open back line arrays when PE was offering their 69 cent Sony boombox buyout drivers. I built the "AudioFred Sony Sweet 69 Frugalarray's" using twelve of those 5-1/4" drivers with a single Vifa XT25 tweeter. This one also requires a sub, and I want all the way, using a Pioneer 8" from Fry's for $111.

  http://fredt300b.smugmug.com/Hobbies/Speakers/132721_kcDVmw#!i=1512196823&k=37hRx64

[Bob in St. Louis]

Nice!
Good info Fred, thank you very much. The others looks like fun as well.
When my workshop stops being "the project", I'll probably find some reason to build one of these monsters.

Good times.   

[versus rider]

Finally nearing completion. I'm waiting for a 100' spool of OFC wire and 100 quick disconnects to arrive from Parts Express. Twenty four woofers and eighteen tweeters take a lot of wire. For those who haven't been following the thread, they stand 8' 9" tall, 16" wide, by 2-1/2" deep. The twelve Dayton Audio RS180 midwoofers and nine BG Neo8 PDR tweeters on each side are bi amplified using a DEQX preamp, equalizer, room correction and active crossover, and four channels of an Emotiva XPR5 amplifier. The subwoofer is a single mono TC Sounds Epic 12 driven by a Datyon Audio 500w plate amp. I'm building a couple of H-frame OB Eminence Alpha 15's to audition in place of the ported Epic 12. The midwoofer to tweeter crossover is a 48dB/octave at 1,200hz. The sub to midwoofer frequency will be determined after testing the midwoofers' rolloff in their ob configuration.   

so your mid woofer is covering most of the critical midrange, seems a strange idea. What is the frequency range of your mid woofers? I see you are using the wonderful DEQX, what model do you have, I use the older PDC 2.6. BTW impressive build.