Self Spinning Gyroscope

This project was designed to be interesting and beautiful, demonstrating the physics of a gyroscope in a form factor that you can easily play with. It spins constantly, due to the integration of a single phase brushless motor into the rim of the gyroscope. By adding pivots on the sides of the gyroscope, that central rotor is free to rotate out of plane in response to torque on the outer frame. The frame of the gyro conducts current through the pivots and the electronics surrounding the rotor, avoiding issues with twisting and binding wires.

Tools Needed:

• Laser Cutter
• CNC mill, At least 3" by 3" by .5" cutting envelope.
• Lathe
• Sander
• Soldering iron

The components used in each step are listed with the step, along with the CAD files and links to relevant documentation.

The first part to fabricate and test is the rotor of the gyroscope. Before moving on to machining more expensive brass, the geometry and drive circuitry needs to be verified. The motor I designed is a sensored single phase permanent magnet brushless motor. It is sensored because it needs active sensors to control the drive coil, single phase because there is only one drive coil, brushless because there are no mechanical switches controlling the direction of current through the coil, and uses rare earth magnets to provide magnetic fields embedded in the rotor of the gyroscope.

The magnets I used were:

https://www.digikey.com/product-detail/en/radial-m...

Permanent magnets in a motor need to alternate the pole that is facing towards the drive coil, which means that there must be an even number of magnets to insure there are not any magnets of the same orientation next to each other.

By choosing a gap between magnets, and a number of magnets, I set the diameter of the rotor. 0.020 inches between magnets is enough to guarantee that the inside and outside of the rotor are well attached to each other, and that the magnets aren't too far apart. With the dimensions set, I laser cut the design from a piece of scrap, and press the magnets in to the outside of the rotor. I also pressed a skateboard bearing into that center of the prototype rotor to allow it to spin freely without friction.

The Fusion360 sketch attached is a parametric CAD file for designing the laser cut part and properly generating vectors to laser cut.

The stator coil is hand wound from magnet wire. The magnet wire can be found on Sparkfun in a bundle of 3 different gauges, 22, 26 and 30 AWG:

https://www.sparkfun.com/products/11363

Most coils are wound around a ferrous core, which provides helps direct and focus the magnetic fields generated by the coil. Because the gyroscope motor is meant to coast freely, the coil was wound around nonferrous core. Having a piece of steel close to the magnets would have caused cogging when the magnets got close to the steel and did not want to move away. Spinning rapidly with these uneven magnetic forces causes vibration and energy loss, which would interfered this the smooth coasting of the rotor.

The coil is wound from the 30AWG wire in an effort to get as much linear length (and from that resistance) as possible for an given volume. The inner diameter of the coil is 0.2 inches, and the outer diameter is about 0.5 inches, which was a diameter that fit the design of the frame in an orientation that coupled with magnetic fields of the rotor. Ultimately, the wound coil with 30AWG wire ends up with about a resistance of 1.4 ohms.

There are 4 main components to the electronic system of the motor:

• The latching hall effect to provide the commutation waveform:

  https://www.jameco.com/Jameco/electronic-products/...

• The high current PNP drive transistor:

  https://www.digikey.com/product-detail/en/on-semic...

• 75 Ohm Resistor

• Drive coil from previous step.

• 2 AA Batteries

In order to spin the motor, current has to run through the drive coil and generate a magnetic field when it would push the rotor in the correct direction. This is mediated by the latching hall effect sensor, which turns on or of the transistor, depending on which magnetic pole is currently below the sensor. By offsetting the coil from the sensor and playing with that offset, when a magnet opposing the electromagnet is lined up properly the coil is turned on and pushes on the magnet in the rotor. As that magnet moves away, the transistor turns off, stopping current flow through the coil and allowing the rotor to spin freely. Once another opposing magnet is lined up again, the coil turns back on and pushes on the rotor, spinning it up more and more. This switching on and off in time with magnets is called commutation, and can be read about more here:

https://www.monolithicpower.com/Portals/0/Document...

The base resistor controls the current to the coil, and from that the power dissipated in the coil. Too much power and the coil begins to overheat. Too little, and the motor won't spin very fast. I wanted to run about 1 amp through the coil, and found that for my transistor, a base resistor of 75 Ohm created about a 1.1A current when the base was connected to ground. When the base is grounded by the hall effect sensor, the base to emitter current is ~44mA. This transistor has a gain of about 25, resulting in a 1.1A current out of the collector.

Because the transistor can only drive current through the coil in one direction, the motor does not necessarily self start because it can only push on half of the magnets. This problem can be solved with more components, but I wanted a minimal number of electronic components as to not detract from the look of the brass and steel gyroscope.

With the geometry and electronics tested the finalized components of the rotor can now be machined. I used a HAAS Office mill, but any reasonably accurate CNC that is capable of cutting brass should work fine. Machinable 360 series brass is a joy to work with, and I used a 0.25" sheet from McMaster Carr:

https://www.mcmaster.com/#8954k102/=16ex81a

The holes for the magnets are bored with an endmill .002 inches larger than the diameter of the magnets and then seated with retaining Loctite:

https://www.amazon.com/Loctite-231097-603-Retainin...

This ensures that the magnets cannot come be removed from the rotor, and will not fall out. Pay special attention to the orientation of the magnets, as removing Loctite is very difficult. and involves a blowtorch.

The center of the rotor is drilled with a 0.120 drill, and then reamed with a reamer to 0.125 inches. If you don't have a reamer, a 0.125 drill used carefully should be accurate enough. The attached CAD file was toolpathed with HSMWorks, which can be used for free in Fusion360. Tutorials on toolpathing can be found at:

http://www.hsmworks.com/docs/hsmworks/2013/en/#Tut...

and suggestions for feed and speed can be found using HSM Advisor:

http://hsmadvisor.com/?page=HSMAdvisor

To clean up the edge of the rotor, put it on a lathe with a live center holding it against an arbor and use a finishing to remove the machining chatter from the outside of the rotor, and add a chamfer to make the edge more comfortable to touch. With that done, use a collet to hold a stainless steel pin which will be the axle of the rotor:

https://www.mcmaster.com/#90145a487/=16exe0i

Turn a point in each end, which will be the cone part of the cup and cone bearings that allow the rotor to pivot. This axle should slide into the center hole in the rotor, and will eventually be attached together with the same retaining compound used for the magnets.

The frame of the gyroscope is cut from the same piece of brass the rotor was cut from. Each part is a simple 2D machining operation, profiling the parts out of the brass using a 3/16" carbide endmill. The inner frame is 2 separate parts, to allow a positive voltage to be applied to one side of the frame and the other side to be grounded. The central frame pivots on spring pins, which allows the frame to rotate and conduct current.

There are 2 types of screws in the gyro, set screws for bearings and machine screws to hold the split frame together with acrylic pieces. The brass set screws have a cone shaped dent in the tip, into which the conical tip of the gyro axle sits:

https://www.mcmaster.com/#92991a106/=16exies

The machine screws are simply fastening parts together, and are countersunk to reduce their profile:

https://www.mcmaster.com/#92451a106/=16exj6y

there are 10 holes that cannot be made during the first cut on the CNC, because they are in the sides of the parts. Put these parts in the vise, and drill the holes in the correct position, as seen in the CAD.These holes are for seating the bearing set screws in the sides of the frame and mounting the frame to a base. The holes on sides of the outer framed are not tapped, but the holes on the inner frame and bottom of the outer frame are all tapped with a 4-40 tap.

With all of the parts machined, spring pins can be soldered into the sides of the outer frame. These pins are both a pivot and electrical connection to the inner frame:

https://www.digikey.com/product-detail/en/mill-max...

Due to the size of the brass frame, the soldering iron needs to be quite hot to get the frame ready to solder. In addition, flux should be used to promote a good solder joint. With those pins soldered in, all the parts can be test fit together. The inner frame is held together by clear laser cut 0.125" acrylic, a power supply can be attached to the bottom to verify voltage is making it through to the inner frame. It also allows measurement of the exact distance between the mounting holes on the bottom of the outer frame.

The base of the gyro is made from a solid block of walnut. A template for the position of the holes ensures that when mounting holes are drilled and the outside of the walnut block is rounded everything is centered nicely. Cut the block to a close shape with a band saw, and then use a sander to true up the circle precisely. Use a drill bit and drill clearance holes for the 4-40 mounting screws through the top of the block. Flip the base over and use a forsner bit to make enough room for 2 AA batteries underneath the base of gyroscope.

A battery holder for 2 AA batteries is mounted in the bottom, and is connected to a switch which is put through a hole and panel mounted through the base to make it stick out the top.

Battery Holder:

https://www.digikey.com/product-detail/en/keystone...

Switch:

http://www.jameco.com/webapp/wcs/stores/servlet/Pr...

The switch is put in line with the +V line and controls whether or not it is connected to the outer frame. Each side of the batter holder is soldered to a washer so that it can be slipped on to the screws that hold the frame to the wooden base making the connection from the battery to the frame through the screw. The battery holder is then glued into the bottom of the base so it won't fall out when picked up.

Using wet dry sand paper, sand all of the brass from 320 grit to 1500 grit. once at 1500 grit, use polishing compound to hand polish all of the parts:

https://www.amazon.com/Blue-Magic-400-Polish-Cream...

The electronics should be soldered as closely as possible together, with leads from the power and ground sides of the circuit bent so that they are touching the separate parts of the frame. this should power up the circuit, provided the switch is switched on.

With a coat of boiled linseed oil put on the walnut, all of the bearings seated with retaining compound so that vibrations won't remove them, and the rotor glued to the shaft, the gyro should spin very quickly when turned on. If it does not spin immediately, a gently nudge should be enough to get the rotor whirring. If there are issues with the motor not spinning well, try slightly bending the hall effect sensor to change the offset from the drive coil. This adjusts the commutation timing of the motor.