Internet Explorer no es un explorador compatible con TI.com. Para disfrutar de una mejor experiencia, utilice otro navegador.

VIDEO SERIES

2021 Seminario web técnico de TI

Más información sobre nitruro de galio (GaN), multiplexores de potencia y fusibles electrónicos, conectividad inalámbrica y accionamientos de motor.

Video Player is loading.
Current Time 0:00
Duration 0:00
Loaded: 0%
Stream Type LIVE
Remaining Time 0:00
 
1x
  • Chapters
  • descriptions off, selected
  • subtitles off, selected

      Presenter(s)

      [MUSIC PLAYING]

      Hello, everyone. Good morning. Thank you to everyone who's joining our webinar for the South Asia region. This will be our first session for this webinar, and we really appreciate your time this morning to be able to join us.

      So I'll be starting off the presentation for this morning. We will be having our topic called GaN in high power density application supply design power supply design targeting EV charging stations and storage-ready inverter applications. So I'm Alfred Logico. I'm the field applications engineer supporting the Philippines.

      And joining me will be Salil Chellappan from the industrial systems team, who will be presenting the second portion of this presentation. And we also have Ricardo Ruflo from the industrial systems team as well, who will be supporting us in our Q&A session later. So without further ado, let's get started.

      So here is the overall agenda that we have for the presentation. So I'll be mostly handling the GaN introduction portion. And for the rest, from the system introduction for the multilevel topologies to ANPC control scheme to optimization, hardware implementation, and results summary, and status of reference design will all be handled by Salil's portion. And we will have the Q&A after the conclusion.

      So to get started, first of all, the question that we would like to answer right now is why GaN. So the main answer to this question is, the main advantage of GaN is having the lowest switching energy loss in the system. So first of all, we would like to answer this question with how can we achieve this lower switching energy loss.

      So in power FET applications, the figure-of-merit is one of the critical parameters to consider. The figure-of-merit is mostly a quantity used to characterize the performance of a device, especially in its application. So for power FETs and GaN, we mostly get this figure of merit through the switching energy multiplied by the RDSON of the FET. And the smaller the figure-of-merit is, the better.

      As you can see here, we have two tables highlighted below. One is for a hard switching figure of merit, which includes turn-on and turn-off losses plus COSS and reverse recovery losses. And we have on the right side the soft-switching figure of merit, which only includes the turn-off losses and ZVS turn-on. This is comparing a TI scan versus available silicon carbide MOSFET in the market today.

      And as you can see for both hard-switching figure of merit, as well as the soft-switching figure of merit, the green graph is what TI's figure-of-merit is. And we are much lower, especially as we go to higher load currents for both hard-switching and soft-switching applications. With this, we are able to achieve 50% lower switching energy losses. And this would help our GaN enable higher operating times to switching frequencies versus silicon carbide in both hard switching and soft switching applications.

      So the next slide, we are mostly focusing on what are the key advantages of GaN and how are you able to achieve this. So we have four main advantages. The first is we have lower gate charge capacitance than silicon carbide or silicon FETs. This enables us to achieve faster turn-on and turn-off and higher switching speeds. We also have lower COSS and QOSS output capacitance and output charge resulting also with higher switching frequencies. We also have lower RDSON versus silicon, or regular silicon FETs, which provides lower conduction losses.

      And a special feature of using GaN is that we would have 0 reverse recovery loss, which means that we do not have a body diode in our GaN FETs. So this helps GaN enable power FETs to achieve very low capacitance between its terminals. It also helps GaN reach switch-on and switch-off very fast with very low power loss. This would also enable higher switching frequencies, as I mentioned earlier, while also maintaining higher efficiency. And because of this, we are able to achieve smaller, lighter, and cheaper solutions in our end equipments.

      So this next slide is where we compare or give an overview of how our GaN competes against silicon carbide in the market today. And as you can see in this graph below, we have highlighted the different advantages of GaN versus silicon carbide, where we are able to achieve a better reliability, better costing, higher switching frequencies, lower QRR loss, and as higher voltage applications as well. Or as you can see, by comparing the reverse recovery loss for silicon carbide, it would usually be around 190 nanocolumns. But for GaN, we would have 0 because of the no body diode.

      And also, an advantage of using TI GaN in your applications where be the supply chain benefits. So TI uses an implementation called the GaN-on-silicon approach, where we use a silicon substrate on our GaN devices, which is more or plentiful from our suppliers, which we could provide more supply. We could also make larger wafer diameters versus regular silicon carbide manufacturing methods. And also, we provide inherently lower cost versus silicon carbide. Also, the GaN and silicon substrate is an investment that TI has been working on for the past years already, and are continually improving on this process, and continue using our existing silicon process technology to be able to improve on our manufacturing of this, and also from transition from larger 150 millimeter diameter wafers to larger wafer diameters, and also utilizing dual-source manufacturing strategies for our devices.

      And in this slide, we are just mostly highlighting the history of GaN with TI. So we've been working on GaN ever since more than 10 years ago already, starting from 2010, and slowly continue developing our process in manufacturing, and also our device features up to now. So for our latest GaN solution, just to give everyone an overview, this is our LMG342x and LMG352x series. These are gen 2 GaN FETs.

      So as you can see, TI's solution is pretty unique. Rather than a regular GaN solution, which is just a discrete FET, for TI, we integrate the GaN FET along with the driver, as well as a bunch of different features in the device. So we integrate not only the drivers, but we also integrate the different protections, like OCP, OTP, UVLO protection.

      We also have active power management and temperature monitoring. We also have fault protection and fault reporting, as well as we also have an integrated high-voltage LDO for powering digital isolators. So this is all in a single package, and you could find out more by looking at these different part numbers or links in our website.

      So to summarize everything, why we used TI GaN is basically to provide design simplicity, enhance robustness and to mitigate risk in our applications. So for design simplicity, because of our integrated driver, we are able to enable higher storage of greater than 100 volts per nanosecond. This results in lower switching losses in hard-switching applications. We also integrate the protections and reporting, like the overcurrent, overtemperature, and undervoltage that I mentioned earlier. This provides redundancy for fault protection in your system.

      In terms of the enhanced robustness, the localized protections in every FET with zero external components. So since we integrate everything, it provides better ease in an application, and also robustness in our application, and also faster response times to fault conditions in our designs. We also mitigate the risk by having lowest parasitic inductance between the FET and the driver since it is integrated.

      And also, we have the adjustable dV/dt, or slew rate. This helps reduce EMI, and also provides EMI versus efficiency optimization. So now, we will be going to the next portion of the presentation, which will be handled by Salil Chellappan, which will discuss the application of our GaN in EV charging designs.

      This presentation is about the GaN or silicon carbide-based bi-directional three-level ANPC inverter or PFC power stage for EV charging stations and storage-ready inverters. So let's take a look at how this presentation is structured. First, we look through some typical applications of the ANPC power staging system in the introduction. Then we'll take a look at some multilevel topologies, and specifically discuss the operation of the ANCP topology.

      Next, the control scheme optimization of this topology will be discussed in detail, with the emphasis on PWM reduction methods and HV failure protection based on CLB. The hardware implementation of the ANPC topology will be explained next with test results and waveforms. The control group implementation with isolated current shunt amplifier will be looked into as well. The results will be summarized next, and the status of the reference is updated. We will conclude the presentation with the list of takeaway points.

      Let's see how EV charging is evolving recently. EV chargers are of two types. They're AC charges and DC chargers. DC chargers do not have a power [INAUDIBLE] stage in them. They use the 1 volt charger inside the EV, and include only the communication and protection interfaces to the grid.

      Every EV will come with a level 1 AC charger that is rated below 3.7 kilowatts to be able to connect to the residential power grid. There are level-two chargers with capacity to 11 kilowatts, but they will still be limited by the capacity of the 1 volt charge. The DC chargers or level-three chargers can bypass the 1 volt charger and directly access the internal battery, and hence can charge much faster. They come with 50 to 150 kilowatt power, and have AC/DC and DC/DC power stages inside.

      They typically use two-level power topology with around 15 to 30 kilowatts power in modular construction. And that is typically built with silicon IGBT devices for unidirectional power flow. They can charge EVs with 400-volt batteries from 20% to 80% charge in about 30 to 50 minutes.

      We are seeing a change to faster charge in all the three levels of EV charging now. The level-two chargers are offering up to 22 kilowatt capability for EV, with more power for the on-board charger that can achieve full charge in about four to eight hours. Home charging is also transforming with the portable DC charger concept.

      Many car manufacturers are now offering an external portable DC charger as an accessory with capacity of up to 22 kilowatts either by removing the on-board charger or limiting its capacity. They typically use two-level or multilevel power topologies in a single stage or parallel model. Being portable, small size is very important here. And hence, the use of faster switching GaN devices is an advantage. They are built specifically for the model of the EV, and can be made to support battery voltages to 800 volts with bi-directional power flow capability.

      Level-three highway charging also is getting upgraded with power capability to 500 kilowatts. They are built for unidirectional or bi-directional power flow with multilevel topologies using mostly silicon carbide devices in modular architecture of around 30 to 60 kilowatts per module. They can handle a range of battery voltages up to 800 volts, and can charge them from 20% to 80% in about 10 to 15 minutes.

      As the previous slide mentioned bi-directional operation, let me elaborate a little bit about the V2G concept that is becoming popular. The V2G system is a special category of EV charger that lets the EV communicate with the grid, and return electricity to the grid or throttle the charging rate based on demand. As at any given time, almost 90% of cars are parked, their batteries can be used as power storage to [INAUDIBLE] grid fluctuation and interruption as in the [INAUDIBLE].

      This can bring estimated savings of up to $4,000 per year per car to the utilities. The main requirement of a vehicle [INAUDIBLE] capable charger is to use bi-directional topology for power conversion. The control unit in the EV, as well as the charger, communicates with each other and decides the direction and rate of power flow.

      View series

      2021 Seminario web técnico de TI