Inverter | Tech Talks | Lucid Motors

Inverter | Tech Talks | Lucid Motors

Show Video

- Hello everyone. It's me again, Eric from Lucid Motors. Today, I'm excited to take you through our in-house developed inverter as we continue to talk about the major components of the state-of-the-art Lucid electric drive unit. You can discover our drive unit in person by visiting any of our studio locations. Here you can see our Lucid Air drive unit. The second drive unit is shown as an expanded cutaway exhibit, so we can see the parts inside.

And these four components, the planetary reduction gear set, the differential, the electromagnetic motor, and the inverter make up the drive unit. And the drive unit is what turns electricity into motion which is a very important job. In this Tech Talk, I'm going to take you through what the role of an inverter is, how they work and why ours leads the industry today. (mellow music) (mellow music continues) Our inverter is benchmark-setting for a few reasons. For starters, it is extremely powerful reaching power outputs greater than 500 kilowatts.

It is also lightweight weighing in at just nine kilograms. It's a compact package that exceeds 75 horsepower per liter and 70 horsepower per kilogram. It is a modular architecture that elegantly scales for high and lower-power applications. Our inverter is tightly integrated in the drive unit playing a key role in motor thermal management, and the inverter software and hardware work together to deliver a wide efficiency plateau that reaches 99.5% conversion efficiency

which is really, really impressive. Like all our engineering at Lucid, we take a systems approach to everything we create and our inverter is of course no different. As such, with our inverter, we don't sacrifice power for efficiency nor efficiency for power. So what is an inverter? Let's find out. An inverter channels the energy of the battery into the motor, which produces mechanical power.

During acceleration, power flows from the battery into the inverter in the form of DC, direct current. The inverter converts the DC into AC, alternating current, for the motor. And then the motor generates mechanical torque using that alternating current.

During operation, the battery is always connected to the inverter and the inverter is always connected to the motor. This means the inverter is the only smart link between the electrochemistry of the battery and the electromagnetic machine of the motor. It is the inverter's job to balance these two systems in order to provide precisely the torque requested by the driver. The Lucid Air's inverter is a leader in the industry providing smooth, responsive, refined, and efficient torque that no internal combustion engine could ever dream of. So how do we achieve such precise torque control? We do it via an inverter-driven synchronous AC permanent magnet motor.

Peter goes over this in more depth in the motor Tech Talk, but, essentially, this type of motor uses permanent magnets in the rotor to establish a strong magnetic field. A stator surrounds the rotor with coils that form three groups of electromagnets. These electromagnets are fed AC which acts on the rotor's magnetic field to produce torque.

A synchronous motor achieves better efficiency by precisely aligning the stator's magnetic field against the rotors. Now, you might be asking how can we use an AC motor when our battery supplies DC current? Good question. The missing link is the inverter which is also known as a motor controller. Let's do a quick demonstration to illustrate what I am talking about. Here is a simple magnetic compass. It aligns with the Earth's magnetic field, eventually, pointing to Earth's magnetic north pole.

We can influence the direction of the compass needle with local magnetic fields. The compass will align itself to this bar magnet. And now we can turn the needle by moving the bar magnet. We can do this because the compass needle is itself a permanent magnet. The needle has a magnetic field that wants to align itself with surrounding magnetic fields. The permanent magnet motors in the Lucid Air operate exactly on this principle.

The center of the motor is called the rotor and it is permanently magnetized. Like the compass, it wants to align itself with surrounding magnetic fields. By rotating an external magnetic field around the rotor we can cause the rotor to spin.

The rotor spin drives the wheels through a transmission to generate motion. So how else can we generate a rotating magnetic field around the rotor? One method is by using an array of stationary electromagnets called a stator. This device here is an electromagnet. When we flip the switch, current flows within the coil at the top generating a magnetic field. The current strength is indicated by this dial here. Now, let's add a second electromagnet.

As you can see, it uses the same power source. Let's call the first electromagnet A and the second B. As you may have guessed, the compass needle will align itself to the new electromagnet if the first is turned off. By adding a third electromagnet we can spin the compass needle continuously simply by sequentially turning on each coil. As you can see, this form of control is not very smooth. But if we could gradually decrease the field strength of one electromagnet while increasing the strength of the next in sequence, the movement would become much smoother.

Here's a real-life demo. These sliders change the strength of each electromagnet. In this configuration electromagnet B is at full current and electromagnet A at minimum current.

By reducing the current of B and increasing the current within electromagnet A, we can smoothly move the needle back and forth. This can afford a smooth continuous rotation from an array of stationary electromagnets. Now let's plot the ideal current in each electromagnet over time. As you can see, the three electromagnets are provided their own sine wave. The magnitude of each sine wave is identical but they're shifted or phased in time. Naturally, we call this three-phase current.

To maximize performance and efficiency it is critical to perfectly align these phases with the position of the rotor. We have a special sensor called a resolver that informs the inverter of the rotor's exact position and velocity. This greatly increases motor capability at high and low speeds. We choose to repeat these three phases many, many times within our continuous wave-wound stator. This improves the strength and efficiency of the motor but it doesn't change the fundamentals of three-phase power. By applying a precise amount of current at precisely the right time we can generate a smooth torque and this is the job of the inverter.

So now that we understand why we need three-phase AC for the motor, let's talk about how the inverter builds AC from the DC battery. Unlike wall power, the electricity in our battery is DC with essentially a fixed-voltage. AC power changes its voltage over time.

We can plot this change on a graph with voltage on the y-axis and time on the x-axis. The ideal AC waveform is a perfect sine wave for both voltage and current. So how do we build AC from DC? Well, you could connect each phase to the battery in sequence.

This is a decent start, but it doesn't make for a very good sine wave. The positive current pulls on the rotor magnet. But if you could connect the electromagnet to the battery in reverse, just like this, the current flow reverses and it provides a push. To accomplish this an inverter has a circuit capable of instantly reversing the current through the coil. This is called reversing the polarity as you're electrically flipping the north and south poles of the electromagnet. You can then alternate between positive and negative currents in each coil, effectively doubling the amplitude of the signal.

This is where the inverter gets its name. It repeatedly inverts the direct current to generate alternating current. Fascinating, right? This gives us a little bit more power. But this digital square wave here is nowhere near as efficient as a constantly-varying analog sine wave would be. So how do we create an analog sine wave with digital switches? If you can switch each phase on and off multiple times per cycle you can provide an average current that brings us closer to that ideal sine wave. By themselves, digital switches are either one or zero, equals on or off, with no setting in between.

To create an analog signal we need to augment these switches with a digital control system. By adding switching events and varying the duration and polarity of each pulse, we can effectively smooth out the average current. This is a form of control called pulse-width modulation. This method of digital-to-analog conversion is how Class D audio amplifiers can generate smooth sound waves from digital switches.

Just like achieving a smooth tone in a stereo system faster switching can lead to reduced motor and inverter losses. Each phase must complete two full cycles to turn the rotor just one revolution, and that rotor is designed to spin at up to 20,000 revolutions per minute. This means we require switches and control electronics capable of turning on and off tens of thousands of times per second. Our silicon carbide switches and inverter architecture make this possible. Okay, we have one last trick to refine the sine wave. We use the stator as a large induction coil.

Induction coils naturally smooth out pulses of current and with careful tuning of the motor and the inverter we can smooth out the signal of each phase. With very precise and fast switching, we can get very, very close to our target, the ideal sine wave. If we can replicate this across three circuits phased in time, we have successfully generated our desired three-phase inverter output.

This is one crucial piece of the puzzle as to how we achieve our industry-leading efficiency. So now that we know what an inverter actually is, let's talk about why ours are so special. What makes our inverter capable of both ultra-high efficiency and high power? Let's get into it. We'll start with high-voltage. Motors turn AC into torque, inverters turn DC into AC, as we just learned. For the same power, increasing voltage allows you to reduce the current.

As we know from previous Tech Talks doubling the battery voltage can mean quartering the resistive losses due to the formula resistive power losses equal I squared r. Moving to a higher voltage, architecture can enable more power but only if done correctly. One advantage of Lucid's inverter is precision in its torque delivery. An EV with capability for 100 horsepower needs about 30% of its horsepower to cruise on the highway.

So here we've got 100 horsepower, and here are 30 horsepower, roughly 30%. The Lucid Air Grand Touring Performance with 1050 horsepower needs much less. About 3% of its peak power to cruise on the highway. So let's see what that looks like. 1050 horsepower.

And the same 30 horsepower to cruise on the highway. Imperfect control precision leads to small deviations in delivered power. An occasional 1% error in the 100 horsepower EV may not be noticeable as this is only one horsepower.

Due to the much greater power capability of the Lucid Air, that same 1% error would be 10 horsepower which would be very noticeable and bad for efficiency. What this means then is that with greater power comes the need for greater precision. The Lucid Air's inverter is key to delivering the power of a few muscle cars combined with the refinement of a luxury car. High voltage allows us to scale up the peak power of a drive unit in a very cost-effective manner. But how did we achieve both high peak power and buttery-smooth refinement during normal driving? Part of the answer is our use of silicon carbide switches and diodes.

Traditional silicon power electronics are very mature, low-cost, and durable. As such, you will find them used in the majority of EV inverters and chargers today. A recent and exciting development in the world of power electronics has been silicon carbide.

In nature, silicon carbide can be found in the extremely rare mineral moissanite. Technology has advanced to allow the synthetic manufacturing of these crystalline structures resulting in a switching device that can be pushed to higher temperatures, currents, and voltages than traditional silicone. Another promising semiconductor technologies these days is gallium nitride.

These GaN switches are making their way into low-power applications such as phone and tablet chargers due to their low cost and high efficiency. GaN semiconductors are highly efficient, but unlike silicon carbide, they were not the right choice for our inverter due to their voltage and power limitations. Every switching device has its own complicated set of efficiency parameters.

Along with simple resistance losses, heat is generated when changing state. These so-called switching losses are tricky to predict and require accurate modeling of the motor inverter system. What's more, is that the proportion of losses within the switching devices is sensitive to the system voltages and power. We needed 924-volt capable switches that could deliver incredible power density and industry-leading light-load efficiency in the same package. Through testing and complex in-house simulation it was clear that silicon carbide was the right choice for our higher-power, high-voltage, and higher-efficiency application. Now, with the silicon carbides switches that we chose we needed to extract the very best performance from them.

Silicon carbide posed new challenges in the realm of thermal management. Of course, every system has inefficiencies. Generally, power is lost to heat. That heat must be extracted to keep components within their operating temperatures.

Any thermal management system must be tailored to the components it manages. These components vary in loss density and time constants. To be perfectly clear, our system is not 100% efficient.

Losses exist and they vary over operating regime and our mechanical output capability is very high. So in turn, our inverter cooler must be designed for the highest thermal loads within some margin. The cooler is designed to reject nearly five kilowatt from the tiny surface area of the silicon carbide chips within the modules. This heat has similar power to an oven's heating element or a little more than three American tea kettles.

Most of this heat is concentrated in the small chips within the inverter modules and so our cooler must be designed for this concentrated heat. Now let's talk about loss density. Heat losses are measured in watts. These watts are dissipated over a surface area and when the surface area decreases, the loss density increases.

Silicon carbide switches have lower heat losses than silicone, but the chips are also much smaller. This means the heat losses are concentrated leading to a high heat flux. Loss density is why your 60-watt computer chip in your laptop needs a specialized heat sink and fan to keep cool but a heating blanket needs no such complexity to continuously dissipate 400 watts or more. The blanket's surface area is much higher and therefore its loss density is lower.

Because of the concentrated nature of losses the Lucid Air's inverter cooler has special internal features to dissipate heat into the coolant. The other factor we need to consider is a component's time constant. As anyone who has deep fried a turkey will know, given the same input heat, a massive component, or turkey, will take longer to heat up than a small one. In an EV under acceleration the battery heats up slowly due to this thermal mass.

The inverted chips heat up nearly instantly, thus, the cooler must be designed to wick that heat away just as quickly. This means the inverter cooler design is very different from a battery cooler design. And since the area of the silicon carbide chip is small, you need a great cooling strategy to efficiently sink heat into the coolant. Our architecture optimizes this heat transfer with a minimal coolant pressure drop. This lets the coolant pumps run at lower power, improving range. Our in-house simulation experts at the system and component level worked closely together to develop the best cooler design for our application.

We decided on a parallel cooling architecture rather than a series cooling. With a series cooler, the coolant snakes through the chips in series, gradually getting hotter along the way. This usually means the last chip is the hottest, limiting the performance of all the other chips. It is also a restrictive path for the coolant.

With parallel cooling, each module receives the same temperature coolant via an inlet manifold. This means we can maximize the utilization of every silicon carbide chip. It also means less pressure drop due to multiple parallel paths.

So why doesn't everyone use this architecture? Well, it is very hard to design a system that is well-balanced, especially at variable coolant flow rates. This design is possible for us because of Lucid's key know-how and skills within our Engineering and Simulation teams. Keep in mind, we develop in-house the mechanical hardware, power electronics, software, and calibrations of the entire Lucid powertrain. This vertical integration allows us enormous flexibility to solve challenges elegantly. This systems approach extends beyond just the drive unit.

Torque vectoring and traction management is a coordinated affair in the Lucid Air due to mechanically independent drive units on each axle. So in order to achieve our desired handling we must make sure these drive units never fight each other for grip. In addition, we can significantly affect the character of the car by dynamically controlling the front and rear torque bias. Adding a third motor allows for even more exciting control schemes.

With the Lucid Air's control architecture, the driver, central vehicle controller, and the individual drive units all operate in harmony. By taking advantage of the powerful processors and ultra fast speed sensing within the motor and inverter, each drive unit can achieve responsive and precise torque delivery. The end result is traction management that is simply not possible from traction control architectures developed for internal combustion engines. This is key to the Lucid Air's exceptional driving dynamics, truly a part train greater than the sum of its parts. This inverter has been sitting beside me for quite some time now. So what's actually inside this thing? We start with the inverter cooler that we introduced earlier.

We then mount our silicon carbide modules onto this cooling structure. Three modules are then mounted on each side, totaling six in our high-power inverter. The two sides operate together in parallel. One strategic advantage of this architecture is that we can omit three modules to build a half-power inverter.

For example, the 819 horsepower Grand Touring has three silicon carbide modules in the front drive unit's inverter. In the 1050 horsepower Grand Touring Performance you will find six modules in the front inverter among other changes to the powertrain. After adding the module we add our three-phase blind mate connectors. These specialized connectors are a Lucid innovation that eliminates the bolted connection between inverter and motor typical in other drive units. This connection is remarkably short reducing mass and resistance.

Such an elegant solution was only achieved through careful coordinated design of both the Motor and Inverter team working together. Next, we add the capacitor banks. These DC capacitor banks are tuned to stabilize the input DC voltage, allowing for ultra-fast switching frequencies to improve power and inverter efficiency. Now, we get to the circuit boards. This stack of printed circuit boards, or PCBAs, starts with what is known as a gate drive board.

This side. This board interfaces with the modules themselves and delivers precisely timed pulses to open and close those silicon carbide switches within the modules. Each set of three modules comes with its own gate drive board.

One, two, three. Each inverter has a single control board conducting the orchestra of switches through the gate drive boards. This is the control board. This board takes in position and speed information from the motor's resolver. It uses this information, along with commands from the vehicle control unit, to decide the best sequence of switching operations to deliver at any given moment.

We now have assembled the working power electronics of the inverter. This package weighs just six and a half kilograms. If you were to measure the performance of the inverter without its enclosure, you would find the power density to be 100 horsepower per kilogram and 110 horsepower per liter. To complete the package, however, we slide this assembly inverter electronics directly into the enclosure. A few final touches like the DC high-voltage connector here are added to complete the inverter. Done.

You may have noticed that there is no lid. Well, why is that? We designed the drive unit as a holistic system and so we were able to use the top of the motor housing to complete the seal of the inverter. This joint-up thinking extends to the drive thermal management system.

Most EV motors require their own supply of coolant, independent oil and coolant passages, a motor-mounted oil-to-coolant heat exchanger, and sometimes a coolant jacket for the stator. Our solution eliminates this mass and complexity and it's all made possible by the innovative cooling microchannels that you learned about in detail in Peter's Motor Tech Talk. The oil flowing through our stator microchannels is directed up into the inverter housing to a conventional plate-type heat exchanger found on this end of the inverter. So by designing the drive unit as a holistic system we found that we only needed one fluid connection which was coolant to the inverter. This lets us keep coolant out of the motor completely reducing part count, complexity, cost, mass, and even the number of hose connections during assembly.

Those are all great things. You may also be curious about this odd cylindrical structure next to the heat exchanger. In there, you will find one of the Air's most closely-guarded secrets, our thermal control valve.

This Lucid innovation more precisely controls the relationship between coolant temperature and oil temperature to maximize efficiency under light loads and heat extraction during high-power operation. In conclusion, our modular Lucid Air inverter delivers breakthrough performance, power density, and efficiency, and is elegantly designed for high-volume manufacturing. Thank you all for watching this episode of our Tech Talk series. Our incredible inverter would not have been possible without the hard work of our power electronics, software, and mechanical engineers.

So thank you. Stay tuned for our future Tech Talks that will dive into more of our proprietary in-house developed tech. Cheers. Goodbye.

And auf wiedersehen. (bright music)

2023-05-31 11:14

Show Video

Other news