Low Barrier of Entry PMSM Design Guide - 001_Introduction

     This post marks the beginning of my motor design guide!  I have been wanting to build a functioning motor for a while now, but only now do I feel I have enough knowledge (related to motors and fabrication) to complete the task.  It was a year ago when I first started really reading about motors and understanding them comprehensively.  This initial interest led me to an internship that I have only just completed that revolved heavily around brushless motors.  I certainly wasn't the one designing them, but I got to talk with and work alongside the engineers who did and we're going to find out just how much they rubbed off on me.

    The goal of this guide is to lead someone with little understanding of motors, step-by-step, through the design process of an outrunner, sinusoidal, radial flux, PMSM motor with concentrated windings of their specifications.  This guide will even held them identify their desired specifications if they don't have everything worked out yet.  The resulting motor will certainly not be amazingly optimized, but it will be 'good enough' and the guide will help the user understand how 'bad' their motor is in the end.

    This first post is not going to be a good reference for people looking for a comprehensive design guide.  It is more to begin the process and capture my unrefined thoughts on how this thing should go.  I will begin with the knowledge that I have at the time of writing this post and, when I reach difficult design decisions, I may reference existing simplified models and tactics.  I believe that understanding what is going on is very important, but I also acknowledge that some people do not care to understand the physics and operating principles of brushless motors and only want one to appear in their hand.  Because of these dueling ideas, I will try to provide visual aides and anecdotes for the crowd interested in understanding, but they will be optional in the design development process.  As a part of the visual aides, I will do some simulation in FEMM for the purpose of demonstration, but none will be required of the user.

    So, there are a lot of parameters that constrain a brushless motor and I need to be able to address any number of them, so long as they do not contradict one-another.  Ideally I need to list every conceivable parameter, but I will start with what I deem the most common or sought-after parameters and then add more as I go.  This guide would work best as a type of flowchart (my chemistry professor used to say that engineers love flowcharts) where the user is presented with starting blocks that represent different parameters, those blocks flow into others and so on until the users initial choices have cascaded through the entire chart and they are left with a complete motor design.  The first two starting blocks that I will focus on are maximum, continuous torque output and motor package because these are the two things that even the mostly uninformed user should have an idea about.  And, even before determining any of the other parameters, these two initial values constrain each other and the user will immediately know if they will need a gear reduction or need to accommodate a larger motor.  "Suggestions" will be made for all of the other necessary parameters for chart completion.

    This is the flow process that I have so far:

• Desired torque -> Estimate a reasonable current for that torque -> Calculate required kT

• Estimate range of plausible number of slots based off of package and fabrication constraints -> Find a slot to pole ratio that is sinusoidal in that range -> With that ratio determined, calculate the number of turns of conductor per winding that is needed to achieve the kT value previously determined

• Determine what gauge of wire is needed from the current and how much space around the teeth is required for winding -> Create the profile of the stator

• Do analysis in FEMM to size the stator axially in order to reach magnetic saturation at the determined torque output -> Size the magnets based off of the diameter of the airgap, the number of poles and the axial length of the stator

• Determine supply voltage from either desired max speed or desired torque at a certain non-zero speed -> Spec MOSFETs and power electronics

    Ok so the obvious weak point right now in the flowchart is determining the current earlier in the process.  It's kind of just guessing and then revisiting it if things don't work out down the line.  If I had a nice system of equations, then all of the parameters could be solved for concurrently and this issue would not exist.  Something else I could do to get an estimation of the current would be to take a desired supply voltage (or suggest one) and either a maximum unloaded speed (or max speed at a certain load) or maximum continuous torque at a certain speed and output a current value.  I'll work that into the next iteration.

    I'm going to end this post here since this was just meant to be the very beginning of a series of posts on motor design and not exhaustive.