Monday, February 21, 2011

p0w3r Drive: Three Cheers for the Arbor Press

Expecting to do some research this afternoon; did not exactly pan out. So I elected to say on west campus and do some wheel work.

The other night I tried boring one of the 5" 30A durometer wheels..

... but then it exploded. Synopsis tells me there are air pockets in the core material, and my feed rate was too high, which caused the bit to grab and toss the wheel. 300 fine for aluminum, not so much for plastic

Today, I went for round two. Turned the lathe into high gear, 455 rpms and lots of non-cinnamon (unfortunately) Tap-Magic. Like magic, we now has a properly bore'd wheel.

Posing with the stator for comparison.

Speaking of which, I put the stator hub back on the lathe and reduced part of the center mass. This would make internal wire organization easier by allowing more internal wiggle room.

Decided to tackle the flats next. Unable to find the binding screw for the large Vertrax mill, I went over to the ME2110 labs to barrow their cute mini mills.

OH GOD THE STORIES ARE TRUE! These mills really suck. I was making passes through aluminum like a pro, when I realized the mill axis started drifting. I felt like a inverted pendulum balancing robot, constantly adjusting to changes in my environment.

Either way, I finished the task, and with a half decent finish as well. Next chance I get, I'm asking my fellow studio ULI's how to tighten the collets on the Vertrax.

While I added a roll pin groove using a hacksaw, I finished preparing the waterjet, post-dredging. I want to say it would take about an hour to fill the tank manually, but since I filled it over the span of two days, nobody remembers. Similar is to be said about the number of licks it takes to get to the center of a Tootsie Pop.

Waterjet filled, cutting commences. I cut the magnet ring laminations from a 1/4" plate of A-36 mild steel, end caps from 1/4" 6061 Aluminum, and the spacers from 1/8" polycarbonate.

Ironically, the material for the spacers was the sheet of polycarbonate that was stolen from me last semester. Hmmm, how circular.

There was drilling, tapping, countersinking, and a big hand from the enormous arbor press. I declare the hubmotor mechanically complete. That leaves us in this current state:

Gathering inspiration from Iron man 2, I proceeded to wind the motor. 23.5 turns of wire per tooth, dLRK style for space savings.

Lesson learned: If you wind with anything 20ga or lower, wear gloves or wrap cloth tape around your fingers. Your writing hand will thank me later.

Completed stator winding, terminated with a Wye connection. I'll wait to see if this motor actually works before adding the sensors. This means the possible run date is the 23rd of Feb, seeing as how UPS managed to butcher the shipping of the box of magnets (routing delays).

Until then, cheers.

Saturday, February 19, 2011

p0w3r Drive: Turning

Late Friday night, looking to build. Waterjet is under its scheduled maintenance, so I elected to roll on over to the lathe for some quality turning time.

Feeling too frugal to purchase a large piece of Aluminum round, I pulled out a 1.25" dia round of 6061 from my scraps. Slightly less than 32mm. Shouldn't affect the system too much. I may add some anti-static tape to fill the gaps if necessary.

Post-Turn part. Drilled through with a 7/32" and tapped to 1/4-20. Chamfers with whatever degree I could find were made on outer edges. I was surprised to find a beautiful mirror-like sheen for the finish. Too bad that won't last long.

As illustrated in the model in a previous post, the ball bearing goes over that extrusion.

Radial picturage.

Next time when I finish prepping the waterjet, Ill cut the endcaps and the magnet can.

Speaking of magnet can, I elected to make a small design change in favor of using more (than one) of the 25ct bag of 2mm roll pins I bought. So here we have 7 additional 2mm holes through the magnet can rings as alignment devices.

I will probably partially press the rings together, add a gratuitous amount of superglue and press the entire assembly together. Should be solid. This part should NEVER BE SEPARATED!

Already I am beginning to find some slight design oversights. A 1/2" shaft with a 1/4-20 tap was a poor decision because that leaves very little room for wires to exit. Think of it as a 1/2" tube with 1/8" wall. My options are to make flats on three sides of the shaft or increase wire gauge. I am favoring the latter because cable organization would be far easier. I should be within reasonable current ratings in normal scenarios.

Friday, February 18, 2011

Hububububble Motor

The next logical step for any piece of technology is to make it lighter, more powerful, and smaller. These goals I hope to find in the first actual hub motor I will be building.

Don;t get me wrong, I have built my own brushless motors before. In fact, if anyone recalls the first Attrition, it had a custom built 40mm motor to internally spin the drum. Now it is time to build yet another motor, slimmer, larger in diameter, and torquier (new word or sp?).

First round will be a test platform. I have a bunch of ideas on improving the sea of hub motor designs but those will be applied once I can prove to myself that I can produce one basic device.

Borrowing the techniques of Shane Colton and Charles Guan, this motor will have the following characteristics:
  1. Flux ring is composed of waterjet stacked laminations. It will feature internal grooves to make magnet housing easier.
  2. Stator hub is a single piece turned aluminum round. It will have milled flats on both ends to facilitate the exit of motor and sensor wires.
  3. The motor will have three hall effect sensors to interface with the controller
  4. Motor will be wound 30-turns per tooth, dLRK for space savings
  5. 5" OD, 1.5" wide
  6. 70mm x 16mm Hitachi Stator
Where did I get these numbers from? Charles wrote up a neat instructable with will teach you motor building basics and some sample calculations for rough estimates. Very useful and highly recommended.

Instead of featuring threads to grip a tire, I will be using a "keyway" integrated into the flux ring.

In actuality, it is an extension of the through holes used to bolt the entire motor together. I sunk in the material where there were no screws, and left the extrusions as mechanical "grips" to grooves I will be adding to the inner diameter of the cored out 5" wheel. Forces are in the correct direction, I dont expect anything pushing the motor horizontally, this should work. The number of screws might change depending on how unsafe I feel about the sheer forces on 6-32 screws.

The pile of parts has already begun. McMaster-Carr arrived today, so I received two 30A durometer 5" wheels, three 1/2" bore ball bearings, and 2mm roll pins. I already salvaged the stator from an earlier grab, and the magnet wire recycled from previous orders. The bag of 6-32 flat screws came from Cake, where I ordered a bag of 50 only to use 2. Gotta love bulk buys.

The remainder of the parts should be coming in throughout the week/weekend. Speedy Metals is expected to arrive tomorrow, which means I can begin waterjet cutting the flux rings granted the machine is not broken (oh no my baby!). I suppose then it is possible to have a rotating wheel by Saturday morning. The Applied Magnetics order is projected to arrive Monday, but gluing magnets was never a time consuming issue since it is just adding glue. I don't need to be watching the grass to know that is growing.

For the time being, project name sits at "Power Drive", or "p0w3r Drive" if you are feeling creative.

Monday, February 7, 2011

Survey: Self-Balancing Electric Skateboard

Please help my team win the Inventure Prize by filling out this short survey. I promise it won't hurt!

The next prototype. Notice no hand controls, unexposed electronics, and simple design.

-22" long, 12" wide, 6" tall (wheel height)
-Weight reduced all metal frame for strength and durability while light enough to carry.
-Pneumatic wheels allow the user to silently and gently cruise over bumps and obstacles.
-Lithium Polymer batteries allow over an hour of run time with minimal weight
-Runs about 9 mph
-Rider Sense safety features prevent accidental injuries to self and others
-Customize your ride with a variety of interchangeable decks!

If you're completely unsure about prices, here are some related products:

$550- EMAD Electric Skateboard

$160- Razor E200 Electric Scooter

$190- Sector 9 Longboard

$6200- Segway i2 Personal Transporter

$300- Altered Wombat Electric Skateboard

$760- Altered Drop Electric Skateboard