The Baldor was a 3 HP, 3-phase, 4-pole, motor. It also was wired for dual-voltages and Star/Delta, so it was really perfect. I converted it with an interesting cluster of alternating magnet sizes to fit as many on as possible, minimize cogging, and keep the clearance down. Link:
http://www.sparweb.ca/3_Gen_MoCo/Baldy.html
The Baldor was converted in winter of 2010, and went up in August of the same year. I was not happy with the tail hinge, or the noise carrying through the tower, so I built a new tower mount in 2012. It runs very quietly now. While I was at it, I also made the tower 10 feet taller in 2012, and gave it a tilt lift mechanism with more mechanical advantage.
I replaced the bearings when I had it down in 2012 and in 2015. I guess it's due for another set, but no complaints have been heard in a long time. I remember being a bit fussy over the choice of bearings in 2015 since the previous set had only lasted 3 years. I won't know if that was worthwhile until I take it apart next year when I take it down for maintenance. But I'll probably replace both bearings, good or bad, just so that they don't force a replacement at some inconvenient time.
The wooden blades are really worn out, and I will have to make a new set next summer. They went through a horrible hail-storm and have been noisy ever since (but not bad enough to replace them yet). The current blades are 8-feet diameter, but now that I have a data-logger, I can tell that it's really not running at full potential. My next set of blades will be 10-feet diameter, but have the same pitch. I'd like to make a new hub at the same time, too.
I often shut it down in high winds, but I know I don't need to. I've watched it using a datalogger in some severe storms and learned a lot about the behaviour. There's a self-induction reaction in motor conversions that limits the power they will put out, and this works to my advantage and may help you out too. The brief peaks can be pretty high, but during those peaks the efficiency drops so much that the blades slow down immediately. The tail furls at the same time, so either mechanism would do, but both happen at the same time. So again, really robust combination that I can trust to run a long time.
For me, shut-down in storms has more to do with not eroding the surfaces of the blades any more than necessary, than any concern about the generator. If you watch yours run and if you have any way to estimate the RPM's, you may be able to see the same thing happening. If not, you can always judge by noise and vibration etc. and after a few years watching it, you may decide the same.
If the tower is easy to tilt down, then checks that everything stays tight of a fairly regular basis, and trial adjustments to the furling, are a good way to get it into commission and make sure everything is set up right.
I once tied a ribbon to the tail to see how "true" it ran to the wind. The ribbon didn't last long but it certainly was informative to compare the ribbon's path to the tail's as it furled.
