I worked on this machine over the weekend a bit, so yesterday was really a pretty light day... just assembly for the most part. Overall though, there are about 4 days work into this project.

I finished up the mould. I've got the mould layed out so that we know exactly where each coil must fit. The two circles in the middle are the path of the magnets, so the coils must be centered of that space, and the 12 radial lines from the center show the maximum width of the coils - any wider than that and things wont fit together. When I wind coils, I place each one in the mould to make sure it's a good fit. Ideally, it should occupy all its alloted space but not more.

I've decided to make this a 48 volt machine, and hopefully soon I'll upgrade my whole system so that I can install it! Last Thursday I tested a single coil made from #13 wire with 80 turns on it and estimated the cutin speed would be around 70 rpm if I stuck with that. I started worrying about heat - and a cutin of 70 rpm is a bit slow, so I wound all 12 coils with 2 strands of #15 (equiv to #12 wire) and 68 turns. I was shooting for 70 turns, but they just wouldnt fit, and half way through I changed my plan - then went back and removed 2 turns from each coil. Now they fit perfectly. Each coil weighs about 1.25 pounds, so I figure we've got about 16 pounds of copper in this machine.

One all the coils are wound, I stack them in their 'phases' and cut the leads to the right length. Then I burn off the insulation on the leads with a torch and sand the ends clean.

I connect the coils together (and solder the connections) for one phase at a time. I like to make the connections around the inner diameter, it looks nice, and it makes sure we have lots of clear space on the outside for drilling holes. I suppose it lowers resistance slightly too - but I doubt thats an issue.

Kind of ugly... but it works! There are surely better ways to make sure that all the coils are in exactly the right spot, but this is my preferred method for now. After all the connections are made between coils, I use duct tape, and tape the coils to the mould so that all 12 are exactly in the right spot.

I cut a bunch of little squares from fiberglass cloth, and glue them over the legs of the coils with super glue, attaching each coil to it's neighbor. This holds things together nicely so I can pull the whole assembly out of the mould.

Theres a picture of it after the duct tape has been removed. I've also made the 'star' connection here. At this point, it's pretty tough - easy to pick up and nothing will break or bend.

I grease up the mould, this time with 'johnsons wood wax' (because I'd run out of turtle wax), put some resin in the bottom - then put in a 'doughnut' of fiberglass cloth.. a bit more resin, then the coils, then more resin - another piece of cloth, then more resin... then clamp the top down. This time I mixed the resin according to the instructions. I didnt have any talc or anything else around for filler (wished I had) so I just poured it straight. It took almost 3/4 of a gallon. I was worried about such a large volume getting too hot and possibly cracking. Sure enough - there were a couple of minor cracks when it came out... but nothing too bad, nothing a little epoxy couldn't fix. Talc would've helped to prevent that - possibly a bit less hardener would've helped too. My experience, when I cast these with 'bondo' brand resin - it takes 1 - 2 hours before its hard. I watch things carefully, and when it just starts getting hard (But still somewhat soft) I find it's easy to cleanup the top of the mould (and whatever ran onto the floor). After it seems hard, I usually pull the top of the mould so it can cool faster. Once it's cooled down to room temp I'll take it out. It doesnt usually take that long.

I pour resin in the magnet rotors at the same time so my shop only smells bad once! Each magnet rotor here has about 1 quart of resin. Overall - I used almost a gallon and a half on this project.

I usually drill three small holes (3/16" or so) in the stator right in the center of 3 of the coils, and screw the stator down to the wooden template that I used to locate the stator brackets. (I also used this template as a lid for the mould) Then I drill right through the same holes that I had the stator brackets bolted to when I welded them on the machine. This assures the stator will fit the brackets perfectly and it assures the stator will be centered on the machine.

Here the stator is finished and all the mounting hardware is on. I used 3/8" diameter brass bolts for terminals.

Here we've got the back magnet rotor mounted to the hub.

One rotor on the machine! The magnet rotors are about 45 pounds, the hub is about 8... so this is getting fairly heavy.

Here we're getting the stator adjusted with a nice airgap. Green and yellow... I'm certainly not a football fan, but it was pointed out to me that these are almost exactly the right shades of the right colors for the CSU (Colorado State University) Rams. Woops.... too late now. I went to CSU for one semester...

Using the 3 jacking screws to lower the front rotor on.

There it is, all finished up except for some wood working! It's amazingly stiff when shorted out - it makes the last 'big' wind turbines we made seem somewhat small. Overall it's got 25 pounds of magnets and 16 pounds of copper in the alternator. We hit 50 volts at about exactly 80 rpm now which seems reasonable. Im not sure if Ill have the ambition to truck test this alternator - I'd like to, but I dont really have all the stuff to do it right yet. We may... or I may just stick a 17' diameter prop on it and see what happens. It'll be a while, I still have a 12 volt system here. Lots of sorting out with batteries, solar panels and inverters before I can really fly this one.
Again, this all took about 4 days. Here are links to the rest of the project.
On Day 1 I pretty much got most of the metal parts cut out and some of the welding done.
On Day 2 we got the rest of the welding finished, the magnet template made, and the magnets put on the rotors.
On Day 3 We built the coil winder, tested a single coil - and painted all the metal parts. Lots of fun...