Close

Log Episode VI: Return of the Manufacture

A project log for Open Source Spherical Motor

Development of a an open-source, Arduino-based spherical motor to promote omnidirectional drive systems within the open-source community.

alex-dunnettAlex Dunnett 03/30/2020 at 18:060 Comments

After much journeying through the murky world of commercial laser cutting and 3D Printing I have been forced to abandon my BLDC Spherical motor as it would cost far too much. As it stands I have already managed to spend my budget twice over by cannibalising the budgets of other students who have been forced to abandon their manufacture. 

But at least the one of my designs can still get made :).

Progress

Turns out that, thanks to covid-19, and with little else to do but go for runs and watch films on Netflix with my housemates, lots of progress has been made. Weird that.

Power Supply

Done! Finally and after such a very long long time. This aspect of the project I started working in October! 

What I now have is a fully functional bit-bashing 3 Phase AC Current generator powered from batteries. Although in practice I am limited to 12V and 4A by my choice of Arduino motor shields, it would not take too much more work to use my motor shields as relays for a much potent supply of power.

Though I have yet to conduct a formal test of its capabilities, by looking at the LEDs on the circuit the system appears to be capable of supplying power at anywhere from between 1Hz to 40Hz. However I may be able to widen this range by changing the prescaler value on the system clocks.

Hopefully I will post a video tutorial on how to do this soon if anyone is interested :)

Manufacture

After spending a significant amount of time over the last week trying to contact and then, in one instance, haggle with various metal cutting, laser cutting and 3D-printers I have managed to arrange for the majority of my components to be manufactured external. 

So that's good, all that's left now is to assemble my pieces together. Though that could be easier said than done considering that I really do not have access to many tools in my student house. Perhaps I'll go home to do that stuff.


Numerical Simulation

As I mentioned in my last log I have moved to put the burners back onto this aspect of the design. After some tinkering I finally overcame a small issue which had been stopping any of my previous simulations running, a consequence of my version of the software being more recent than that of the youtuber I was learning from.

Not that it matters anyway as my last simulations were done in January and my design has changed significantly since that point in time.

My current plan is create 3 or maybe 4 simulations:

Three Phase Induction Replica

A replica of the three-phase induction motor I saved from the scrap-heap so I can validate my power supply against what its supposed to do.

Linear Induction Machine

A representative linear-induction machine. In order to prove the concept behind my method of operation.

Circular Linear Induction Machine

Using the same parameters as the linear induction machine to describe a linear induction machine in the shape of semi-circle. Through this exercise I should be able to show that the assumptions about applying linear induction logic to a rotating machine are justified.

BONUS ANALYSIS: My Model

Strictly speaking not necessary as the points I am trying to prove should be proved in the previous simulations. Additionally because my model was designed with ease of manufacturing and assembling in mind the design parameters, such as slot design and stator thicknesses, are not consistent with the other designs.

Additionally I do not believe it is possible to create spherical geometry in Ansys Maxwell (Though I could be wrong).

Could still be quite cool though.

^^So far all simulations are being run in 2D as the simulation time is much, much, much faster and time is pressing. If I am confident in the validity of my 2D sims I may run some in 3D in order to see what impact this has (end-effects etc.),


Testing Rig

Some very unexpected and accidental success here. Upon realising that my super - cheapskates hall effect sensors weren't giving results that made any sense (like at all, they would go haywire as soon as magnets went near them, though they would then remain haywire after that regardless of magnet presence).

To remedy this I purchased, for six quid, a board intended for use on a drone. It contained an IMU, Gyroscope and also a magnetometer. Ostensibly this was only supposed to be for reading the Earth's magnetic field and should have been too sensitive to accurately record the magnetic flux that my neodymium magnets were producing. 

Fortunately they worked just fine and had the added bonus of being able to give the exact global co-ordinates of the IMU at the time of the magnetic reading!

I then realised, as may seem obvious to the reader, that this would allow me to essentially paint a map of the flux densities around my stator in both time as well as space. Fantastic! 

Also the board has a temperature sensor. Not useful for my project though it was quite useful for checking for fevers. #2020 problems.

Discussions