[brent_s]

First thing, down firing a driver is never really a good thing for the driver itself. The RS390s are particularly bad candidates...the HO has a sag of 4.7% and the HF is 5.4%...that's percentage of Xmax lost to gravity just from sitting there.

22x22x80 gives about 8.4 ft^3 gross...without doing any heavy lifting math I'd say your're looking at no more than 7.5 ft^3 net after driver/port(s)/bracing is subtracted. If the net ends up a little larger somehow, it's gravy as that will just lower your tuning point slightly for a given port length.

Is there a driver budget? The Tempest-X only has about 2.7% sag...with 27mm of Xmax, there's still plenty to work with. The DVC15 on the budget end of the scale, has 3.9% sag, but has been used numerous times in down firing designs.

Both look pretty good in 7.5 ft^3 tuned to 17hz. With 350 watts, output is virtually identical down to 15hz. That's about where the DVC15 will exceed Xmax for good if you don't use a rumble filter. The Tempest-X will keep on chugging down to almost 12hz before hitting Xmax without a rumble filter.

Where do you plan to set your rumble filter?

With a more realistic 7.5 ft^3 net, your dual 4" ports will need to be 28"...plus another 1" if you use PrecisionPort type flares. Many will probably disagree with me, but I'd just use one 4"x13.6" PSP including flares to preserve your net volume. Dual 4"x28" thin wall port, such as PSPs, will consume .46 ft^3 of your gross cabinet...standard PVC will bump that to ..52 ft^3. FWIW, I practice what I preach using a single 4" PSP with a 5.9 ft^3 DVC15...haven't noticed any port noise so far.

-Brent