I'm not sure if you want to implement passive of that. A high Q dip here. A bump here. Another bump there. A shelf down there. Steep slopes. Basically a lost cause. Better off with DSP.

You can integrate the Rag by doing this:

Buy two of them. Max out volume on both. Then use them as dedicated power amplifiers to accept miniDSP analog outs for the high and lows.

As far as uninteresting, and flat - it's relative. Let's put it this way. Get rid of the mini-DSP shit between the source and the amp, and you get something more interesting. The downside is that you will have to rely on simplier passive designs which may restrict driver choice. Best to use the mini-DSP x-over / preamp as the source though - feeding it directly via USB rather than suffer an AD-DA conversion, which is never good, especially with cheap stuff.

I've done this stuff and variations of this stuff 10 years ago. That's why I have two DCX2496s, three Crest CA-2 amps, and three Teac T-amps.

Very cool thread!

I got into headphones because they are relatively cheaper and support a higher WAF value than my previous obsessions with speakers.  Who would have thought joining a "headphone" forum would lead me right back to speakers again!

Anyone seriously interested in getting into DIY speakers, especially OBs, need to read the work of Martin J. King.  I'm surprised his name/links haven't come up in this thread prior:

http://www.quarter-wave.com/

My current OBs are based on his 18" Goldwood H-Frame model and I like them a lot:  http://www.quarter-wave.com/Project08/Project08.html

I wanted to eventually build a passive crossover but since my choice of upper driver was a bit esoteric I knew that digital would give me the flexibility to tune experimentally to find what I really wanted, so I've been using a MiniDSP to do the XO work for some time.  A Dayton EMM-6 measurement microphone (and some very long cables) let me test room nodes from my PC, usually going with REW as the miniDSP can read it's findings and automagically pick the right EQ settings for compensation.  In theory this is a cool idea because you can readjust every time you move the speakers to a different room or make changes to room placement or treatment.  In reality I've been using no EQ for sometime now though, just to see how good these speakers can behave under less than ideal conditions.

My OBs have the H-Frames on the bottom (3/4" 11 ply baltic birch for everything) and the Audio Nirvana Super 8 full rangers in separate, movable baffles on top with what I call a "damping box" between them which is basically about 40lbs of concrete patio blocks wrapped in the same birch plywood.  The full rangers are set off center (which may have been a mistake) but have been interesting to experiment with by swapping the two sides, moving the full rangers in to center or out, respectively.  Last but not least, separate top baffles allow me to experiment with time alignment (by moving them forward or backward over the H-Frames) and toe-in and would be easy to change out to a different top driver if I want to experiment in that direction too (these Moth & Cicada drivers have piqued my interest!)

I built and tuned them with a target of accurate live music presentation, specifically using the track "Haitian Fight Song" by The Charles Mingus Big Band as a key reference.  They don't disappoint and double well as fronts in my home theater system too (no subwoofer for some time now). 

Tops are powered by a Dayton DTA-100 t amp and the bottoms have seen a few different vintage stereo amps come and go including the Kenwood KM-209 and Yamaha M-65.  I think I'd eventually like to either go passive XOs and a single really good amp (have a Yamaha MX-10000 in the works!) or something like the DEQ2496 to replace the miniDSP and a couple of good pro amps, Crown or something.

Very nice! I'm going to have to check out those drivers.

Unfortunately, the Moth Cicada drivers are no longer available and haven't been for 10 years. The AN drivers you have might be the best bang for buck.

No longer. I found it only marginally effective.

No longer. I found it only marginally effective.

Interesting. Looked up the impedance graph for the Beta 15s (have yet to measurement it) and realized a zobel would be necessary if I wanted to implement a passive x-over instead of the active I'm currently using. I'm curious how the OBs would sound if only the Rag (not the Rag + Crest) powered both the wide-banders and woofer.