I struggle to communicate in simple terms what the difference is. Maybe this is a good way to put it. Your twin alternator needs to run at exactly twice the speed of one large single one with the same number of magnets and coils. The slower one would be heavier in steel disks but cost the same for magnets and copper, and have much reduced (half) centrifugal forces. Or you could run a much smaller disk diameter at your 300rpm cut-in speed using much less copper and magnets
Hi Hugh, and I certainly appreciate your insight even though I rarely follow it (that's a private joke between me and Hugh)
I will outline (or try) my method to madness at the risk of dragging the thread out and getting some folks irate about it.
My ferrite generator weighed in at 78.8 lbs with all three rotors and both stators. It has 10.5 lbs of 13 AWG copper in it. It has 30.8 lbs of ferrite magnets. The rest is steel and fiberglass. I came to the conclusion, using the information that you emailed me one day with a big diagram of it drawn on my shop floor with chalk, that if I were to build a direct drive the stator would be almost 1/4th the diameter of the rotor blades!
I had some basic specifications that I wanted. It has to put out 60 amps sustained without damaging it. I did that in two stators, each one carrying 30 amps of the load. Admittedly 13 AWG is a bit light for 30 amps sustained, but it will do it OK in the winter time when we get those kinds of winds here because it has good cooling then. In the summer we get "burst" winds from storms but it is never 60 amps sustained for hours at a time. So even though the winding is a bit "weak" I think it will work OK. The alternative was using 12 AWG wire and I have wound stators with 12 AWG before - it is like trying to wind coils with rebar. It could have been done with two-in-hand 15 AWG, but it would take perfectly wound coils to get them fit. I opted for the 13 AWG because I had it on the spool.
I used 40 2 x 2 x 1" ferrite magnets in my generator with 10 poles in each section. If I were to put 20 of those magnets on one disk I could get twice the voltage out of each coil. However, I am using gearing. Using 10 magnets (poles) I am getting 2.29x the voltage out of each coil that you can get with the same configuration in direct drive (I ended up using .4375 gearing in the final configuration).
The losses in the gearing system are negligible. When I built my first unit I tried to measure the losses and it is less than 10 pound inches of torque at full load. The only loss, really, is in the extra set of bearings on the PTO shaft and those are single row angular contact ball bearings made by SKF in Sweden, which are one of the most free rolling and low friction bearings on earth. The chain drive is used in everything from road grader final drives to bicycles - it is a proven, efficient system used worldwide in many applications. Put it in oil and it doesn't even wear out.
By speeding the thing up I can make better use of my 10 magnets than you can using 20 of them on a direct drive that is much heavier and larger. This also allows me to build a smaller stator with less coils and copper than the big one. Less coils and copper translates to increased efficiency due to lower resistance and heating losses. The efficiency gains in the generator more than outweigh the very slight losses in gearing.
Flying the first one in single phase configuration proved that the 10 pole single phase works. It puts out decent low wind power but is pretty weak when the wind blows harder compared to a polyphase. By adding the second stack to the generator I basically double the output capacity, which while one unit is not enough for 3.8 meter blades, two units is more than enough. I skewed the second phase by 90 degrees electrically, am using IRP (Individually Rectified Phase) with the output parallel'd on the DC side of the rectifier assembly. This gives me the RMS loading of polyphase (not quite as smooth as three phase, but close), double the output capacity, and smooths out the running of the generator because the rotating assembly is loaded at all times. And the rotors are still 11" diameter and the stators 14" OD. The stack is just longer on the PTO shaft, which is easier for me to handle than a huge unit.
Side note: I discovered why the ferrite generator does not display the vibration problems in single phase that a neo unit does. It is because the mass of the rotating assembly is over double. Just like a heavy flywheel on a single cylinder engine, the energy stored in that rotating mass smooths it out.
I will post one photo here of what it looks like and the rest of them are at this link:
https://picasaweb.google.com/christopher.w.olson/TwoPhaseDualStatorGeneratorThis unit weighs 78.8 lbs and meets the specs I wanted. I estimate that if you were to build a direct drive for 3.8 meter blades it would approach 200 lbs in steel, magnets and copper. I would invite somebody to do it so we could all see what it takes to build a direct drive ferrite generator to this specification.
So that is it - it is on the turbine, it is on the tower, and it is running. It was performing beautifully this afternoon when I got the tower up. It is towards evening here now and our wind had died out (I hate the "dog days" of summer). But tomorrow it will go back to work and I will log power output from it on a daily basis to see how it compares to the neo generator that I use as the "benchmark". I fully expect, based on all my testing and experimentation with this project, that it will give the neo unit a good run for its money. It ended up over twice the weight and twice the size of the neo generator. It is merely in a different configuration than going bigger around.
--
Chris