In this three-part series, Exro’s Chief Technology Officer, Eric Hustedt, helps us explore what a traction inverter is, how inverters work, EV traction inverter development and the latest advancements in inverter technology. This second part of the article discusses EV traction inverter development, including the early history of traction inverters, control algorithms and techniques in power electronics, the difference between linear and PWM regulation, four-quadrant drives and the difference between DC and AC motors.
Error. Please check the form fields and try again.
In "What Is a Traction Inverter?", part 1 of our series, we explored the basics of inverter technology and its primary function in the electric vehicle industry. As we continue with Part 2, we look back at the early days of electric vehicles and the developments that led to the modern traction inverter.
Early History of EV Traction Inverter Development
The first electric vehicles, dating back to the 19th century, utilized direct current (DC) motors and simple lead-acid batteries for propulsion. As this predates semiconductors, the DC motors were often mechanically turned ‘ON’ with no "throttle" control, but thankfully they were generally quite slow.
These early DC motors were not very efficient, and the lead-acid batteries had low energy density compared to today's batteries. This greatly restricted the speed and range of these vehicles, which were soon overtaken by the invention of the internal combustion engine. Brushed DC motor-based electric vehicles remained a niche market, limited to hobbyists and the like, even today.
It wasn't until the early 1980s that several technological advancements converged, creating the modern inverter, and enabling AC motors to cost-effectively move beyond fixed-speed line connected operation and become truly versatile.
As mentioned in part 1, the advent of IGBT technology in the 1980s was a key factor in the development of modern inverters. IGBT, or Insulated Gate Bipolar Transistor, is a type of bipolar transistor that is controlled by a MOSFET. Without getting into too much detail, this made it easier to turn the bipolar power transistor ‘ON’ and ‘OFF’ quickly, which was crucial for efficient inverter operation. At that time, semiconductor manufacturing had advanced to a point where IGBTs could switch hundreds of amperes of current at a reasonable cost, enabling the development of cost-effective AC motor drives.
This section will delve into the various developments that were required to create modern-day inverters. These developments include control algorithms such as Field Oriented Control (FOC) and the micro processors required to perform the real-time calculations.
Control Techniques in Power Electronics
The other two advancements necessary for modern inverters involved controlling the switches to turn ‘ON’ and ‘OFF’ at the correct times to produce the desired AC waveforms. In electronics, there are two main methods of regulating power flow: linear regulation and switching or Pulse Width Modulation (PWM) regulation.
Linear regulators work well but are highly inefficient since they essentially “burn off” the difference between the input and the desired output. Somewhat similar to controlling the speed of a car by applying full throttle and then using the brakes to stay at the speed limit. They are still very popular for low power applications, but highly undesirable for anything more than a few watts of power.
A switching regulator or PWM regulator operates by alternating between fully ‘ON’ and fully ‘OFF’ states, as this is the most efficient mode for a switch. PWM stands for "pulse width modulation," where the ‘ON’ duration of the switch is varied or modulated to achieve the desired output. This method of regulation works based on the principle of averaging – for example by turning the switch ‘ON’ for half of the time, the load will perform only half the work, on average. This type of regulation is more efficient than linear regulation since the two switch states produce very little power loss, when ‘OFF’, the switch experience full voltage but zero current, conversely when the switch is ‘ON’, it has zero (or nearly zero) voltage, but full-load current. Since power loss is a product of voltage and current, if one or the other is zero then the power loss is also zero.
An Analogy: Using Light Switches to Explain the Difference Between Linear and PWM Regulation
The basic idea of PWM regulation is that it works by turning the power source on and off at a sufficient rate to achieve a desired level of output. For example, if you want half the brightness of a light, you can switch it on and off such that half the time is spent ‘OFF’, this will create half the light output on average. To avoid noticeable flickering or flashing, the switch must operate at a frequency faster than the eye can perceive. Generally, this is above about 25Hz, otherwise the load, in this case the eye, “surges” with the switching, i.e., it sees a pulse of full brightness, then complete dark, rather than simply less light. This is critical, the switching action of a PWM regulator must be fast enough so that the load averages or smooths the output. This is how light dimmers work.
Motors can respond very quickly, much faster than the human eye, thus require PWM switching to occur many thousands of times per second. Since each ‘ON’ time must be precisely calculated this presented the second technical challenge beyond the switches themselves. This requirement was fulfilled with the invention of the first microcontroller in 1971, which provided a compact, circuit board-mounted computer to perform the necessary calculations.
Alternatively, you could try to operate this light switch as a linear regulator by holding the switch in a half-‘ON’ position to achieve this half-light output, thereby operating the switch in its linear region. This is far from ideal and a good way to start an electrical fire, needless to say, don’t try this at home!
Control Algorithms in Power Electronics
The final and crucial aspect for modern inverter technology was the development of control algorithms. The challenge was to determine the most effective way to switch the IGBTs ‘ON’ and ‘OFF’, based on the constantly changing angle signal from the motor shaft and alternating currents, to produce a smooth motor operation. Despite a good understanding of the math, early microcontrollers lacked the necessary computing power. This led to the development of a computationally efficient algorithm called vector control or field oriented control (FOC). This technology emerged around the same time as the invention of microcontrollers, but it wasn't until the early 80s, when microcontrollers became commercially available. Thus, inverter-based AC motor control became a cost-competitive alternative to DC motor controllers. With advancements in micro processing power and power silicon, AC motors have now largely replaced brushed DC machines in nearly every application.
Four quadrant operation is a native function of any inverter since the semiconductor switches can operate as both inverters and rectifiers. AC machines are symmetrical, meaning they can produce the same torque whether they are rotating forwards or backwards. This capability of producing positive (accelerating) and negative (decelerating) torque in both rotating directions is referred to as four quadrant operation. There are two quadrants, one in each rotating direction, where the torque and rotation have the same direction and the machine is "motoring." In the other two quadrants, the torque and rotation are in opposite directions, leading to negative power output and the machine is "generating" (the inverter is rectifying).
The ability of AC machines to operate in all four quadrants is a significant advantage over DC motors. This enables electric vehicles to efficiently recover energy while braking through regenerative braking, a process where the machine acts as a generator and eliminates the need for a “reverse” gear set.
How Are DC and AC Motors Different?
All electric motors, including those commonly referred to as "DC motors”, are fundamentally AC machines, the only exception is the homopolar motor, which lacks practical use for various reasons. To achieve operation from a DC source these motors utilize mechanical switching means, such as a brush and commutator, to change the current in the coils.
Much like combustion engines need spark prior to rpm increase, also known as ignition advance, electric motors also need phase advance for the current as their speed increases. This refers to the relationship between the phase of the current flowing in the coils of an electric motor and the phase of the voltage produced by the motor magnetic circuit. The phase advance needs to be continuously adjusted as the speed of the motor increases in order to optimize its efficiency. However, since it is mechanically complex to provide this advance with a brushed system, most DC motors have fixed brushes and therefore fixed phase of the current and motor shaft angle. This greatly limits their efficiency over wide speed ranges that cars and trucks require.
The BLDC (brushless DC) motor, also known as the "electronically commutated" DC motor, was one of the earliest and simplest electronically controlled motors. Today, BLDC motors are widely used, for example the small fans found in computers. However, despite its name, it is actually a type of three-phase permanent magnet synchronous machine, which makes it an AC motor, not a DC motor. Brushed DC motors have become limited to the most basic and low-cost applications, while the inverter is now the standard tool for proper control of electric machines.
In the final segment of this series, we will take a closer look at the cutting-edge advancements in traction inverter technology. Our discussion will cover various aspects, including switches, semiconductor advancements, cooling methods, interconnects, and even a discussion on where Exro fits in this ever growing and changing industry.