Dia. Mag. Alternator

[z.monkey]

New core blocks, 0.5 inch by 0.5 inch and 0.870 +/- 0.01 inch...
The Z Axis tolerances were a lot more difficult.  I botched a couple
of blocks drilling the hole in them.  First get the block square and
plumb, then drill the hole.  Don't try and punch one side part way
and then reposition the part and find the original hole.  It just doesn't
work.  After working out the Z Axis idiosyncrasies, then I managed to
get the core blocks to fit.  Also using countersunk screws in the blocks
so there is no interference in the windings...

[z.monkey]

OK, I got the mechanical parts assembled, with some caveats...

I was banking on the "Auto-Motoring" aspect.  Well, when we add the large blocks of steel between the magnets it tends to focus the lines of force through the blocks.  The auto-motoring still occurs but only at the very end of the 180 degree turn (flipping the poles).  When the magnets are side by side the auto-motoring magnet motors very smoothly (except for a little crossover distortion) with the driver magnet.  So with large cores in there, which are necessary for the coils, we can't rely on auto motoring, and are going to have to fall back to using gears.

A three gear setup will turn the magnets in the same way that they would auto-motor.  While this is a simple change it requires a lot of rework.  The shafts have to be lengthened to accommodate the gears.  I have to find the right size gears, and if not may need to build new end plates to accommodate the kind of gears that I can find.  Chances are that I won't be able to find the right size gear with two different size shaft holes.  So this brings up another topic which is using equal size magnets in all three positions.

Currently I have a 3/4 inch outside diameter driver magnet and the auto-motoring magnets are 5/8 inch.  The shaft on the driver is 1/4 inch, and the shaft on the auto-motoring magnet is 1/8 inch.  This means I need two different shaft sizes and two different bearing sizes.  This is problematic, in that there is a lot of extra re-calculating to design and build the end plates.  So by using equal magnets in all three positions I can streamline the redesign process.

Also, still not happy with my dimensional tolerances.  You don't really see the errors until you assemble it.  Then the lines don't quite go parallel, or its a bit twisted.  Just something that makes it look "off".  With this next set of plates I am have some mechanical assistance in making the measurements, hopefully.  I'm going to get a Bitmoore Vice to help with the positioning.  These things are neat, check it out...

http://www.harborfreight.com/5-inch-drill-press-milling-vise-94276.html

Then, at the end of the assembly process, I managed to drop the assembly on its corner (bottom picture) and bent it.  At the moment I am not worried about correcting it.  This one is never going to get wound.  Its got too many errors.  I need to lengthen the shafts to add gears.  Also need to change the magnet and shaft sizes for the outer (new name instead of auto-motoring) magnets.  This requires that I design and fabricate new end plates.  I can salvage the core blocks and structural hardware because the length of the assembly can stay the same.

Oh, also, forgot to mention the banana end plate.  When I had add the cores and magnets on only one end plate it curled like a banana.  I was able to stretch in back by adding the second end plate.  So, anyway, gonna need thicker end plates.  These magnets have big forces, and are not easy to work with, especially three at once.  The larger blocks helped this process.  The magnets are gonna stick to the largest ferrous mass they can find, in this case the core blocks.  This was easier with the core blocks in close proximity.  Before, on Diamag 2, the screws had wider clearances, which put the magnets further away from the target holes, more difficult to stab...

So, anyway, feeling sorta melancholy about this right now.  It was big work to get here, and it looks like there is bigger work on the horizon.  I'm just forkin' glad its Friday...

[z.monkey]

Mk1, I'm gonna integrate the transformer lamination idea also...

But, although I have plenty of large transformers, there are none that would incorporate my design cleanly.  So what to do?  DIY, duh.  Look at the stack of Iron sheets below.  It's taken two days to cut the laminations so far, and their still solid.  Gonna need lots of holes too...

I have been picking holes in my older designs (Diamag 3).  I think I can still get the automotoring to work, but there needs to be a lot more Iron in the core.  On Diamag 3 the magnets are attracted to the little (1/2 inch) blocks.  So, even with only one magnet in there it is motoring on the blocks, meaning the poles are attracted to the metal, and want to stay oriented that way.  This makes the shaft hard to turn, unnecessary polarization.  The magnets need that attraction all the way around their path, so the core needs to completely surround the magnets.  This adds flux paths between the magnets, and I am hoping that it will allow them the automotor again.

I really do not want to put gears on this thing.  You will be shocked at the pricing.  Tiny little gears, off the shelf, are like $20 each.  Custom gears, made to order, in small quantities are like $100 each.  So, yeah, don't wanna put gears on it if I don't have to...

So, back to the holes.  Got some new toys.  I picked up the Bitmoore Vice and a Drill Press at Harbor Freight.  Also got a drill index with 115 bits.  I am still cutting laminations and bearing plates.  The plan is to stack up all the laminations and bearing plates, in order, and cut the main holes all at once.  This way we can get superior alignment and squareness.  I don't see any rigidity problems with a solid Iron core, but we want to make sure it all comes together nice and square.  So I plan on cutting tooling holes in the stack first.  Then remove the stack from the vice, and bolt it all together with countersinking screws in a piece of wood from the back side.  The area where the tooling holes are will later be removed for the winding slots.  Then after we drill all the pilot holes in the stack we can work on each individual layer to make the large holes for the magnets (1.125 inch) and the bearing plates (0.625 inch), and the winding slots.

The bearing plates will be 1.25 inches square and hold the individual magnets.  In this new design the bearing plates allow each magnet to be removed without disturbing the rest of the assembly.  On the Diamag 3 design once you put the windings on it, there is no more reworking.  You would have to unwind the coils to take it apart, not user friendly.  So on the new one (Diamag 6) each magnet and shaft have their own set of bearing plates.  They also have new bearings.  Before I was using punky little bearings from Mike's Hobbies.  The ball bearings are like 1/32 inch, punky...  The new bearings are a Dayton part, from Grainger, and a load rated at 335 pounds, so no problems there.  New magnets also, moving up to a 1 inch diameter Neo-Dia-Mag, K&J Magnetics part number RX04X0DIA...

So, my previous prediction of bigger work on the horizon is true.  Cutting those steel plates is a lot of work.  I'm sawing straighter lines now anyway...

[Mk1]

@Z

Nice tools ! I like the way you work !

[z.monkey]

Diamag 6
The Stack

I had to cut a few more plates this morning.  Then I was able to assemble "The Stack".  I used clamping on three axis' to hold the stack of plates still while I drilled the tooling holes.  1/4-20 countersunk machine screws hold all the plates together in the Z axis for the precision cuts.  The X axis is held by the machine vice.  I am going to let the combination secure the Y axis, but if it doesn't feel right I'll add the large wood clamp also.

Also, clearance, just barely enough.  I was worried about the Z axis clearance between the drill bit, in the resting position, and the top of the stack.  I have about 3/16 inch.  really close.  My original plan would not have worked.  I realized it was a problem when I assembled the drill press and put the Bitmoore vice on it...

So, you know, spin control...