I wanted to share some results from modeling different DIY speakers with PSD-Lite. Whatever the results, this is more a reflection on PSD-Lite performance than any of the DIY speakers modeled. I find PSD-Lite a very nice free program because it's relatively easy to use, does not require Excel, and adds baffle step response to the development of the crossover. It is an integrated design tool. A few things one should know about the tool (nothing is perfect): The impedance response shown does not include the effects of the port which are not integrated. If a new set of drivers are to be used, shutdown the program and start from scratch. One cannot save the plots, and one may need to do screen captures. How responses add at the cross over point is a little suspect (may have to play w it more). Edge diffraction is not well modeled (as far as I can tell). But baffle step is good (also as far as I can tell). Only 1, 2 and 3 way. No straight forward support for 2.5 and other stuff. All that said, I find this a fantastic tool. Anyhow here it goes: Paul Carmody's Amiga: I think Paul specs 30 liters instead of 34 liters, but I don't think it makes that much of a difference. Adding some stuffing might help with the small peak at 40 Hz. According to the model, the speaker is tuned with a little punch. Very smooth response. The first stage of the woofer crossover is a notch to deal with the woofer break up. After that, we have a 2nd order filter. Very nice response. Jeff Bagby's Tango: Jeff's Tango deals with the RS180 cone break up in a masterful way. There are two inductors, just like Paul's approach, but it has less capacitors and no resistors. The null takes care of the break up really well. The port is very smooth and takes it down to 42 F3 with no peaking. The Amiga, being a tower does go lower using the same woofer though. The crossover between the tweeter and the woofer is very smooth. John "Zaph" Krutke's ZRT: This is an amazing combo. The woofer needs only two components (the impedance is intrinsic to the inductor). It goes to show why everybody and their mother likes the Revelator. Note how the port extends the response down to an F3 of 31 Hz! This is the default Zaph port configuration. The tweeter cross over is a bit more elaborate though. Paul Carmondy's Overnight Sensation: Well, it's not shit. But it's not as good according to the model as the speakers above it. The HiVi B4N is an $11 bargain. But it's not a very smooth driver. Paul mitigated a major break up with a notch. That said, there seems to be a bit of a hump in the 2 to 3 kHz region. More baffle step compensation may help, but note this is a small speaker with little room inside if using big caps. The extra punch from 200 Hz down to 60 Hz might help mask the slight depression 200 Hz to 1 kHz. Note these issues are still between +/-5 dB. I know Paul's model does not show this issues with the OS. So again, this maybe more of a reflection on the PSD-Lite tool than anything else. All that said, an ND105-4 might be an easier driver to work with on a OS form factor. But maybe $10 more expensive.