Christmas Windmill Time

[taylorp035]

So I think I'm going to take Sparweb's suggestion and wrap this with thinner wire so I have more flexibility on how I wires the coils together.  75 turns of 17 gauge, 2 coils in series, 2 sets in parallel for each phase.   70.4 RPM cut in speed.  If I were to wire them all in series, I could eventually make this a 12v machine.   I'm hoping I'll be able to average 15 amps per strand of 17 gauge with a bit of cooling at a max temperature of ~ 200F, but I'm not sure how optimistic that is.   Probably furl at ~12 amps and then burst to 15-18 amps for a few seconds at a time.

I was looking at my blades today and I definitely have them at too shallow of an angle.  The calculations say I won't need much wind to hit ~20-30 amps total per phase and the generator will probably outpace the blades until it furls unless my TSR is lower than 4-4.5.... so I'm thinking of adding another 3 degrees or so to the pitch (~0.2" over 4").

[taylorp035]

I finished up the calculations for the 3 different wire choices.  I'm still planning on doing the 17 gauge for added flexibility if I want to go to 12v.  The wind power assumes a 35% blade efficiency but no eddy current losses.

TSR is assumed to be 4 for the purposes of these graphs.  Since my wind resource is really poor and I want it to start up in really low winds, I'm planning on have a very steep blade pitch.  The airfoils are also extra thick, so it was never going to perform well at high speeds anyways.





The 15.6 mph wind point is 250 RPM.  I'll have to work my way up towards 30 + amps doing 17 gauge with 2 x 2 coils per phase.  I was thinking of having a mini fan inside the nacelle to help blow some air over the stator where I put some extra wraps in a separate rectified circuit.  Or have a set of steep blades on the main shaft in front of the alternator.  Or even have some ducting from outside without getting everything wet.

[taylorp035]

When purchasing magnet wire, should I choose the red 155C rated enamel wire or the 200c rated amber colored polyamideimide wire?


It also looks like I'm going to go with the N52 magnets instead of N48 since the N48's are sold out and were apparently priced lower than they were supposed to be.

I re-ran the magnetic software with the N52's (~gained maybe 3%) but also bumped the number of turns to 86 vs. 75, which yields a theoretical cut in of 29.6 rpm if all 4 coils in a phase were wired in series.  The 200c wire has thicker insulation, but also was available in a 7.5lb spool vs. only a 5 or 10 lb spools of 155c wire I was originally looking at.


And my 3D printed parts for all of this are done.  I'll have to post some pictures once I clean them up a bit.

[SparWeb]

Well... ask yourself what else is going to melt before the 155C enamel?
Sorry I guess that sounds snarky, but it's kind of the truth if you're 3D printing these parts using polycarbonate.
You might want to work up the current slowly, in stages, measuring internal temperature as you go.  Heck, maybe even embed a thermocouple inside the stator a few places.  Do what you can to optimize wire gauge and current, but there's a level where your I2R power lost to heat is going to soften the plastic and the torque will make something "squirm" sideways. 

I'd still like to see how they came out!

[taylorp035]

Well... ask yourself what else is going to melt before the 155C enamel?

- Well, yeah... I know I can't let things get that hot.  At least I made some provisions using the polycarbonate vs. PLA as it lets me have a delta T about 4-5x larger (theoretically).   I guess I was wondering if a particular wire was nicer to work with or the coating was better.  Reading into it, it seems like the 200c coating is better for things that will see moisture and is recommended for generators.

[taylorp035]

Here's the printed parts.  I cleaned them up.  The print job didn't turn out very good, so I'm going to have to super glue a hundred cracks and splits.  The plastic was old and probably full of moisture.  I'm also certain it needed to be hotter surrounding the whole part, so there was a bunch of thermal cracks.

I put a USB drive in there for scale... they are fairly sizeable pieces.  To think it's going to hold 16 magnets and a 6" dia x 4" long spool of copper seems to boggle my mind at the moment.

[SparWeb]

I'm sure you're economizing material, but are you sure you can put any torque on those fingers?

Actually, there are two axes to consider.  Torque on the shaft axis, which needs a reaction from the casing of the stator, and a moment applied across the end of the shaft, which needs a reaction from something like a base flange or mounting holes on the face.  Is there a side of the stator with "feet"?

[taylorp035]

I'm sure you're economizing material, but are you sure you can put any torque on those fingers?

  It's definitely on the thin side, especially for the clearances I have.  I think the fingers will be plenty strong enough, especially if I support them in from the front and back.  They need to react about 6 ft*lbs on average* extra eddy current losses * the cogging torque for a 3 phase design with my magnet layout.

Is there a side of the stator with "feet"?

I was planning on making custom mounts to hold the whole thing from spinning to the base and side walls of the windmill.  So that takes care of the moment about the shaft axis.   The bearing holders have flanges printed in them, so the stator can't slide down the shaft so long as I put shaft clamps on the outside of each bearing.  For any forces radially from the shaft, that's going to come down the the stiffness of my flanges, which will be the marginal part to all of this.  It will be stiffer once I glue all the U-shape sheet metal together on the outside and it's packed full of copper.  I was wondering since I will have 4 coils acting at once, the forces should be fairly well distributed around the stator.




Edit:  -  I purchased the wire and magnets, so there's no going back 

[taylorp035]

Here are some build photos.   I have quite a few hours into fixing up the 3D printed parts.  I would highly recommend anyone else trying this to make sure you print things to the right size vs. having almost everything being undersized.  Today I spent quite a bit of time cleaning the slots out for the rotor so they are big enough to fit the magnets and the glass beads in the glue.

Rotor clearance is looking pretty good.

The magnets are quite strong.  They were spaced about 0.400" apart when I received them and it still took two people to pull one off the end of the stack.  I will be building a wood fixture to guide the magnets into place when I go to glue them.  I'm thinking I'll have to glue them in 4 stages.  This is both because of the magnetic forces, but also so the previous glued magnets reinforce the plastic in the appropriate positions along the way.

[Mary B]

I see you have one of the light duty milling machines, how do you like it? Thinking of adding one to the shop instead of abusing my drill press...

[SparWeb]

I love the cam-less valve tube in the background - "for scale".
Also seems like it serves as a shaft substitute for your machining.

Yeah, and what's it like machining 3D printed parts? 

[taylorp035]

I see you have one of the light duty milling machines, how do you like it? Thinking of adding one to the shop instead of abusing my drill press...

The mini-mill is quite useful.  The 4th axis isn't super useful unless you're into making weird rotary valves like me    Though it's good for when you need to make a very round hole.  If you have 3D drawing software available that's a plus.  You also can greatly expand the usefulness if you have some sort of 3D CAM software for making g-code.  I've had it for 13 years now and have machined 100's of parts on it.  It's definitely a fun toy.   Not any good for machining anything that needs coolant and less than ideal if you want to add any cutting fluid.

Yeah, and what's it like machining 3D printed parts?

The polycarbonate machined nicely as long as you didn't go too deep.  Else it grabbed just like if you've ever tried to drill brass with a regular angle drill bit.  Though I'm sure half the issue was my lack of support on the far side and lack of me pinning the rotor to the 1" steel rotary valve (it was fairly tight from friction).

Right now I'm using a chisel to clean up the ends of the rectangular slots.  Not far off from a piece of oak or maple.

[Mary B]

I do ham radio stuff that needs flat surfaces for heat sinks, need to machine cavities for preamps etc...

[taylorp035]

I went to put the battery in the windmill today and noticed the voltage was rapidly rising in parallel mode when I flipped the switch from 3S to 1S.  This typically means the cells weren't balanced, but this was much more extreme.  I decided to tear the battery apart and found 1/4" of muddy water in the bottom of my container.  Needless so say, the bottom cells didn't appreciate it.  I should be able to fix most of it, but it makes me want to expedite the bigger battery plans.

[Bruce S]

Did you find the offending hole/leak that was responsible for the 1/4" of water?

Be of luck!

Bruce S

[taylorp035]

Did you find the offending hole/leak that was responsible for the 1/4" of water?

I know how the water made it's way in... I never sealed the wood box holding any of the electronics or motor and there were big gaps in the top of it.  My battery also was wide open in the top.  I didn't want to seal things until I had everything straightened out because it would of made it more difficult to service.  It also was too cold and snowy to paint and seal things outside at the time.  For round 2, I will be fixing all of this.


Here's another build update on the alternator.  The magnets are being glued in.  As more magnets go in, the more difficult it gets.  I did the first 4 last week with minimal drama, as they just wanted to repel each other with a few pounds of force.   The next 4 I did earlier today and they were a good bit more exciting.  They wanted to be sucked into the center, so it took about 20-30 pounds of resistive force on my part to make sure the magnets made their way properly into their slots.   Tonight, I just did #9 of 16 and it was even more difficult.   It didn't necessarily want to go in or out of the rotor, but it wanted to rotate axially quite strongly due to both neighboring magnets (NN-SS-NN-SS-NN-SS-NN-SS-) arrangement.  With two people and a pile of clamps, we were able to get it into the slot.

4"x1/2"x1/2"  N52 Neo's.  Most of the reviews online when I bought these suggested they weren't coming apart if two of them came together.

[SparWeb]

You just have to accept that all of your tools will end up magnetized after the project, and move on.
I have a bunch of plastic bar clamps just like those - worked well too.

[Bruce S]

Glad you found the leak!!
Are you able to regain the use of those batteries??

[taylorp035]

Are you able to regain the use of those batteries??

  Only one of the cells was shorting out, so I should be able to save them.  I clipped the positive leads to all the cells that saw water to be safe, but I think I'll be able to re-solder them easily.  I may find another cell to replace the one bad one.

[taylorp035]

The magnet rotor is finished!   The last few magnets required some Channellocks to keep the magnet rotation in control since the wood fixture wasn't able to keep them straight due to the high forces and the fixture wearing out.





I've starting wrapping the wires for the stator.  I discovered I need to make some intermediate fixtures to hold the wires in place while I wrap it, else the wire just falls off the fins.  I've also decided that I'm glad I didn't go any larger than 17 gauge, as the wire is decently stiff to be wrapping in my particular configuration.

[taylorp035]

I've now attempted 4 times to wind the first coil and found it rather challenging to keep my fixturing in place.  After 20 or so turns, the wooden "T" slides upwards as more wires are being jammed into a small space.  I tried adding double sided carpet tape to help keep it in check with improved results, but it keeps moving.  I'm going to un-wind it and add some screws from the ID of the stator to keep it in place while wrapping.

This 3D printed and U-shape lamination idea certainly looked good on paper, but is definitely providing lots of challenges through out that may trip up many people trying the same design.

This is about 40 turns.  I get a bit better each time I re-wrap it at getting the wires to line up nicely.  With fresh, straight wire, it goes on fairly nicely, though getting down below the flange and adjacent fins takes a bit of effort.

[SparWeb]

Thank you for the update.  Some progress though difficult is better than none at all. 
Please let us know if you hit any barriers - it's just too interesting to let you give up!

[taylorp035]

What size bridge rectifiers should I buy for this?  From my reading, it looks like I will need only two of them for a star pattern.
https://www.windstuffnow.com/main/3_phase_basics.htm


The plan is to do dual 17 gauge wires in parallel, so I was thinking of having at least 60A of capability in the rectifiers for a continuous rating if I get a good gust.  Inside the windmill, I have a fairly large aluminum plate (~1/4" x 7"x 10") as a heat sink for everything.   Are there any downsides to using 80A or 100A bridge rectifiers?  The prices on them seems to be only marginally more expensive.

In other news, I brought the rest of the windmill in.  I have re-painted the hubs and re-pitched the blades.  I figure I've gone from about 3.5-4 degrees to about 7-8 degrees.  I think my blades were not at a steep enough angle by quite a bit.    I have also weighed my blades... I was super impressed as they were +/-10 grams, and that's after I have re-carved them twice.   I'm hoping to be able to counteract some of my imbalance in the very heavy steel hubs.

[MagnetJuice]

The plan is to do dual 17 gauge wires in parallel, so I was thinking of having at least 60A of capability in the rectifiers for a continuous rating if I get a good gust.  Inside the windmill, I have a fairly large aluminum plate (~1/4" x 7"x 10") as a heat sink for everything.   Are there any downsides to using 80A or 100A bridge rectifiers?  The prices on them seems to be only marginally more expensive.

Using 80 or 100 amp rated rectifiers is better than using 60 amps rated rectifiers.

If you were planning for 60 amps at peak capacity, increasing the protection to 100 amps is better.

Those 80 and 100 amps higher current rated rectifiers will give you better protection in case of crazy high winds or gusts. That is, as long as they are mounted to a good heat sink.

Ed

[Mary B]

What size bridge rectifiers should I buy for this?  From my reading, it looks like I will need only two of them for a star pattern.
https://www.windstuffnow.com/main/3_phase_basics.htm


The plan is to do dual 17 gauge wires in parallel, so I was thinking of having at least 60A of capability in the rectifiers for a continuous rating if I get a good gust.  Inside the windmill, I have a fairly large aluminum plate (~1/4" x 7"x 10") as a heat sink for everything.   Are there any downsides to using 80A or 100A bridge rectifiers?  The prices on them seems to be only marginally more expensive.

In other news, I brought the rest of the windmill in.  I have re-painted the hubs and re-pitched the blades.  I figure I've gone from about 3.5-4 degrees to about 7-8 degrees.  I think my blades were not at a steep enough angle by quite a bit.    I have also weighed my blades... I was super impressed as they were +/-10 grams, and that's after I have re-carved them twice.   I'm hoping to be able to counteract some of my imbalance in the very heavy steel hubs.

Stay away form eBay rectifiers!!! They are way over rated and a 30 amp might be good for all of 15 amps!

[taylorp035]

I've never used a 3 phase rectifier before, but it looks like I attach the 3 legs of the star pattern to the points in the diagram if I'm figuring this correctly.  I'm liking the looks of this one at 90A continuous.

https://ixapps.ixys.com/DataSheet/VUO86-12NO7.pdf





Tonight I re-drilled my holes to mount the blades since the old ones were at an angle after I changed the pitch.   I also measured the center of gravity from the outer bolt and multiplied it by the blade mass to try and math-out my imbalance a bit.  The difference between blades was about 1.7% for the worst case, so that will be useful when dealing with the un-even hubs.

[SparWeb]

Yes, that looks fine.
Still needs a heat-sink, of course.

[taylorp035]

Unfortunately the 90A was out of stock, so I settled on purchasing the 70A.  I would need to upgrade several other parts of the system to handle much more than that on a continuous basis.

I also have some M19 0.014" thick silicon steel on order.  Hopefully I won't have to use a torch to bend it into the U-shape that I want.


EDIT: -  I finally finished my first coil today after ~4 previous attempts.  I cut my non-straight wire off my spool and started with some fresh wire and then very carefully wrapped it up.  I didn't count, but I bet I'm close to the 70-80 turns I wanted.  It took about 2 hours to do it... I'm sure I'll be faster on subsequent coils and once I improve my T-shape wood fixture.

I'm thinking I'll try to hold off on any potting compound/glue until I get everything wrapped and preferably tested.

[kitestrings]

I've found small wire ties to work pretty well to hold the coils until they are ready to be potted.  Here you can see where I made some slots in the sides of the coil form to accommodate them:




Good luck.

[taylorp035]

That idea worked really well Kitestrings.  Thanks!

I remade my fixture and drilled out the 48 holes I needed to finish this up.  I'm glad I tried to add a few more wraps to the first coils since I think it's a hair on the small side at the moment.  It was hard to judge while I was wrapping it.

[SparWeb]

That looks good and tidy, Taylor.

[taylorp035]

3 coils in.  I've straightened the ends of the fins so the wires don't ride up on the next 9 fins while I wrap.  This should improve the packing factor on a few of the layers.

[taylorp035]

6 coils are now complete.  I received my silicon steel, so I made a shim to make sure the material would not break.  To my surprise, it turned out really nice.  I hope to go use a real shear and pan brake for the ~60-80 shims I need to make.