http://www.fieldlines.com/story/2007/3/10/23942/1876

And here's a link to get you started, one of the many places supplying DC motors like mentioned.
http://www.surpluscenter.com/sort.asp?UID=2008091215563525&catname=electric&keyword=MDDD

My first thought about the 90vdc motor from Grainger was to use a varible speed controller, but then most of those have a low voltage cut out of around 20 volts.
I geuss the only true solution would be to add more PV modules.
To think that I could find a 12vdc motor that would run such a large whole house fan effectively on less than 10 amps dc was a pipe dream.

For what you are talking about I would get something much bigger than just a treadmill motor.

You may have to weld up a new steel mount for the big motor since it would weigh so much.

Your fan blades are large at almost 4' in diameter, so my wild guess is you may want it to go 400-600rpm??  
So for a 12v system like yours, a would first try a big motor with ratings around 130vdc/ 4000rpm / 10-15A .  I have never tried anything that big though.  Sounds like fun though!

Hooked to 12v batts or direct 12v solar panel , it might only use around 60-100watts of power to run and put out quite a lot of wind. But I am quessing somewhat, it is difficult to know for sure.  
It is difficult to estimate how much CMF you really need and thus how much watts that will need to be used in a complex application like yours.

If you could somehow have a relay that would hook the fan DIRECTLY to one of the solar panels, you could get even more for less power by skipping the batteries when the sun is out.  I have hooked my 31 inch floor fan directly to my 45Watt (actually 25W output) HF solar panel, and it really runs like a jet.

You probably won't get the kind of CFM you mentioned, but it should be quite good I suspect especially if the fan has nothing resisting the air flow into the attic.

-Hopefully others can share some more experiences.   -Hope it works good.

One thing I have neglected to do is actually run the fan with the existing motor and take an amp reading. Something I should do with the fan installed to get a more accurate amp reading but, I am reluctant to do since there is so much building involved first and, I wanted to know that I can actually use it first. I know from experience that nameplate amps are not always what the motor draws.

Even after crunching numbers and getting second opinions, sometimes it just takes, a leap of faith.

Motor controller - http://www.grainger.com/Grainger/items/2LHT9

And last - 1,000 more watts of PV array.

Mounting it in a cupola is an interesting idea.  I just wonder if you are going to pull so much attic air through the roof vents and other gaps and cracks that you won't get enough flow up through the ceiling grill?

One nice thing about the cupola mount is that it would be quieter.  They can be on the noisy side at high speed -- two speeds is very helpful to cut the noise down on low speed and still get a lot of ventilation.

Even if you mount the fan at ceiling level, it will cool the attic by blowing the house air through it, which is likely to be a lot cooler than attic air.

Anyway, having a 43" blade diameter (housing is 48") and an overall height of almost 24" and given the fact that a weight bearing wall runs through the house, underneath of and parallel to the roof ridge, which by the way only has 3' of pitch in 10', you can see that something different had to done from conventional whole house fan installations. Not to mention the possibility of noise and vibration (remember the original installation?) Not having sufficient negative pressure at the interior of the home to create a cooling effect shouldn't be a problem, but if it is, it should be easy to overcome and still allow natural aspiration to happen. I was actually more worried about to much negative pressure in the attic space.

As for the speed controller, its just a way of controlling amperage use. If, after all is done, I don't need to run the fan at FLA to make the home feel cooler I can turn it down. I was also kicking around the idea of a timer and or thermostat, to turn it off automatically

All over time of course, not all at once. First thing is to build the "squared" copula saddling the ridge and setting the fan and ceiling grate in place and taking an amp reading with the original 1959 Hoover motor. Which is in perfect condition by the way. I want you to know that I looked at new motors, thinking that they would take less amps, mainly because of progressing technologies, this motor is 40 years old, but that exact size motor today still takes 6.6 amps. Something that puzzles me is that the 12vdc motor I am looking at uses 27 amps.
Same 1/3 HP
RPM's 1750 ac, 1800 dc
It would seem that the dc motor is way more efficient.

115. X 6.6 = 759
759. / 12 = 63.25

I have used a similar motor, not over years.
For the price may be worth it to see how things go.
Also have speed controllers.

Should look to replacing the blade with an air foil blade.
Move same air with 30% less power on a celling fan project.