The Death of Pablo and The Hardware Hacker
Pablo, the motor that I designed to be built from scratch is dead (or at least postponed for the foreseeable future) and I learned a lot from the process I will discuss the downfall of my child here, alongside a book that I finished at the beginning of the summer, The Hardware Hacker, that relates very well to the reasons why I cannot have my way this time around.
(side note: The Death of Pablo is the name of a project one of my friends did a while ago)
So, I did manage to get around to finishing my design before totally scrapping it, so no one here can name laziness as the downfall of this project. The design files can be found on my GrabCAD page. Here are some nice renders (I upgraded to Solidworks visualize form photoview360 for these):
Since the last post, I've been working on finalizing the design and trying to work out where I would be getting all of the materials from. Components made out of aluminum or general fasteners came easily, but more awkward and custom components like the tire and steel laminations in particular were hard to nail down. The main characteristics I wanted from the tire were for it to not increase the diameter of the wheel too much (limit the effective moment arm of the ground) and for it to be solid rubber and not a shell with an inner tube. Finding a cheap rubber wheel of decent dimensions online proved difficult, but I happened to luck upon this video in which the restorer makes their own wheels which I had not considered as an option. I also knew that the ME department had a vacuum former that I could borrow to cast resin properly.
Getting steel for the laminations proved much harder.
Some quick context on these motor laminations: The direction that flux flows through the stator changes a lot. Every time a tooth rotates from alignment with one permanent magnet to the next, the polarity is reversed and the flux lines invert accordingly. This happens # of poles times per revolution and if the motor in question is spinning quite fast, the magnetic field is oscillating at an accordingly high frequency. This oscillation likes to cause what is known as eddy currents, or a circular flow of electrons in the steel itself. Any movement of electrons through a material requires energy (I^2*R) and thus these eddy currents can be thought of as eating directly out of a motor's efficiency. The main way to drastically reduce the effects of eddy currents is to be clever with the construction of the stator in terms of geometry and material. Because of the way that flux lines flow through the stator steel, eddy currents tend to have a lot of axial movement. However, if the stator steel is broken down into electrically isolated laminations, stacked axially, the space that eddy currents have to form is much reduced. If you add to that a special alloy of steel that has low conductivity, eddy currents are no longer a threat, even going into higher frequencies. This is what silicon (electrical) steel is, a mixture of silicon and iron that retains great magnetic properties but also boasts high resistivity. The only downside when moving to laminations is the space lost to whatever you have between layers insulating them from each other, and the increased complexity and cost added to production. There are other specifications for silicon steel such as grain orientation and insulation and there are even newer materials that seek to replace laminated steel in high frequency operation, but that is outside the scope of this post.
So, all I needed were some sheets of silicon steel that I could throw on the waterjet and cut some motor lams out of. This stuff has been used for decades and is present in the vast majority of electric motors worldwide so it shouldn't be hard for me to get my hands on some, right? Wrong! This is where my lack of understanding of the silicon steel supply chain and supply chains in general bit me in the ass. The good news is that it was a great learning experience. Before that, let me do another introduction.
-
I found The Hardware Hacker on the Invention Studio's bookshelf. Most of the literature that you can find on that haggard structure is just reference material. You have your workholding guide, your McMaster Carr catalogue, your heavy duty sewing book, your robotics and electronics handbooks, but someone in the past also took it upon themselves to purchase literature pertaining to making/hacking and make a little section out of it. I picked out my new conquest during one of the sporadic free days that came from covid's destruction of a proper spring break and read enough to become invested. I quickly got busy again, so I wasn't actually able to finish it until the spring semester ended. But boy am I glad that I finished it.
The Hardware Hacker is an edited compilation of blog posts by prolific blogger and hacker, Andrew "bunnie" Huang. I would say that the book is divided into two sections. The first is all about the author's experiences and insights into the electronic hardware production ecosystem in China, specifically Shenzhen. There is heavy emphasis on manufacturing, supply chains, and intellectual property, each item fascinatingly different from its western counterpart. The latter section is a compilation of hacks bunnie has encountered or discovered over his lifetime. They are all individually clever and insightful, giving the reader a brief look into his mind.
Both sections are fascinating and I highly recommend the book, but what I want to focus on is an aspect of the first part and how it applies to the situation that I am in with my motor build. When bunnie first beings the process of coercing a factory in Shenzhen into production, he was doing it for a small company that was going to have a small production run compared to American giants like Apple, HP, and Google. It took him a while to find a willing factory and his company had to assume more risk than normal (paying for all parts made, including defective ones) in order for it to happen at all. This process that he went through mirrors some of my efforts with some American and Chinese manufacturers to get my laminations made. The difference is that I was at an even greater disadvantage, being a single, unbacked consumer/maker rather than a company that had raised a lot in rounds of funding.
-
When I was considering the feasibility of being able to build my own motor, a couple of things were in my head that skewed my perception:
One, I had just finished an internship with an electric motor startup that raised millions in funding, not so different from the one bunnie worked for. At that startup, there was a person dedicated to setting up the company's supply chains because we were trying to transition into production really quickly. He was an old hand at this job and knew how to get all kinds of things seemingly very easily. Because of this he was quite often the person that was asked to handle odd and typically difficult orders (like tons of custom, tiny magnets or several prototypes of a hub we were trying to build from a Chinese firm for free or even very particularly shaped pieces of electrical steel cough cough). So, from my perspective, the materials I worked and built with were seemed to be procured with very little effort. I was blissfully unaware of the work he put in to keep things running smoothly.
Two, like I mentioned earlier, the electric motor industry is old, universal, and absolutely massive. Why would I not be able to get my hands on some measly electrical steel? Here were the avenues that I tried:
I'm trying to build cheap, so the first place I go is Alibaba. Unfortunately, the retailers there will only sell in huge bulk, hundreds the times I need for my single motor. I then try to go through some American suppliers. I have better dialogue here with a few different places, but the tooling costs are absolutely prohibitive to my small scale project. There is no market for selling small quantities of electrical steel to individual customers. Since acquiring stock for the purpose of processing it myself had fallen through, I looked at firms who would cut the laminations for me. These places specifically do small runs for prototyping and research (one of them had even worked with GT faculty before). I got quotes back from 5 different places and they all came around to the same value and it was more than I was willing to invest. My last option was to look for pre-built stators, all the laminations cut and bonded together. This would have taken out a lot of the leg work for me had anything fit the bill, but not too surprisingly, nothing did.
Herein lies my predicament: I am unique. No one is out here trying to buy small quantities of electrical steel or build mediocre, low-cost, and fully custom hub motors like me. In short, there is no consideration towards makers in this market (not that there is in very many markets).
-
There are a few ways forward here. I could stumble upon a viable way to acquire steel laminations and continue the project. I could also scavenge an existing motor and use its stator or buy one of the pre-built stators that I mentioned above; both of these routes would require a redesign of the motor and then the project would cease to be Pablo. As much as I dislike this stopping point, I feel I must face reality and focus on other work as continuing here doesn't seem like it would be worthwhile.
Comments
Post a Comment