I'm a tiny bit worried that the driver seems out of spec, but a sealed box is not that critical and so our results can be off a bit and we will still build the right sealed box. Vented boxes are far more critical and I might try breaking the woofer in for a few days and retest if I were using a vented enclosure.

The next part of the process is to test in a sealed box (or with added mass). This allows SW to approximate the acoustic compliance of the driver (V_AS).

So, I then tested the woofer on a sealed box. I had one lying around from a prior NHT project so I lined the hole with a closed-cell foam (insulating foam tape from home depot) and set the woofer on the hole with weight to keep it stable. Here's the picture (click for a bigger one):

For a box you want to use something with an internal volume of about 1/2 the rated acoustic volume. This will bump the Q substantially and provide a good second curve for the V_AS estimation. Make sure you use an empty box (no stuffing, yet) and really really make sure the box is totally sealed and the gasket is sealing well. If the box has air leaks or is not sealed well you will see a second (smaller) impedance hump near the resonance of the free air driver. Reseal and retest, if so.

After doing the impedance test (Measure / Impedance in Sealed Box) I had the following two curves (sealed and free air). The curve in black is the free air dataset, the curve in red is the sealed dataset. Note that the resonance peak has moved higher in frequency and lower in amplitude. Both are appropriate for this test and so things look good. I don't like the impedance hump near 7Hz for the sealed test - this may indicate a tiny air leak, but it's adequate for our purposes.

Now I evaluate the results and start prototyping the enclosure. 

See here to see using Added Mass rather than Sealed Box to evaluate V_AS.