Hello. And welcome to the fifth and final installment of Texas Instruments Power Density video series. In this presentation, I will explain how TI's integration technologies improve power density in addition to system performance and ease of design. Increasing power density can be challenging, but TI's ability to integrate functions and components can help engineers solve this common power supply design challenge.

In the simplest sense, integration enables the following benefits. You can reduce your board area and BOM. It improves performance by reducing parasitics. And you can reduce design time by eliminating design tasks and component selection, while at the same time, simplifying the PCB layout.

In this presentation, we will dig into three of TI's many integration technologies and explain how integration works. These three examples will bring to life the benefits of improving power density by shrinking board size, increasing performance, and reducing your development time, BOM comp and cost.

The first example of integration is TI's LMG3410R070. This device is a 600-fold GaN power FET with integrated FET driver and protection. The figure on the lower right shows the block diagram. Typical power devices are discrete three-terminal devices. By integrating the driver and protection, the device switching performance can be maximized and the protection optimized. Integrating these functions reduces the BOM comp and simplifies the PCB design.

When integrating the driver in the package with the GaN FET, the switching slow rate can be increased to reduce switching losses. The integrated driver minimizes the gate loop parasitic inductances and eliminates the common source inductance. Combining the integrated driver with an 8 millimeter by 8 millimeter low inductance QFN package, TI GAN is able to switch at slow rates of 100 volts per nanosecond with almost no overshoot or ringing on the switch node.

This figure on the right is switch node waveform from TI's GaN AVN. The integrated driver also eliminates ringing on the gate GaN FET for improved reliability. TI's LMG3410R070 integrates the over current and over temperature prediction for a robust power switch solution.

Now over current circuit senses the GaN drain current, eliminating the need for discrete current sense element and associated comparator circuits. TI's over current prediction response is less than 100 nanoseconds to provide true short circuit protection, which is difficult to design in discrete circuits.

A discrete current sense element such as a shunt resistor or current transformer will add resistance and/or inductance to the power loop. This will increase conduction losses and increase drain-to-source ringing. The integrated thermal protection provides for power supply protection in the event of an overload or thermal system fail and again eliminates the need for discrete external circuitry.

TI's LMG3410R070 daughter card half-bridge GaN FET AVM is shown on the left. This AVM is a complete half-bridge power state solution that requires only the input bus voltage, 12-volt unregulated bias supply, and the low-end high side logic PWM signals. This AVM can be used for a power stage in either hard-switching or soft-switching applications.

As discussed, TI's GaN FET with integrated driver and protection reduces the number of discrete circuits that have to be designed. The circuit on the right shows all eliminated components needed for discrete GaN design, which provides you with the simplified design and bill of material.

When compared to a discrete GaN half-bridge power stage, TI's power stage is only on one side and 64% smaller. It also provides over current production and thermal protection that is not in the discrete design. TI's integration of the gate driver and protection reduces board space, improves switching performance, provides protection with a reduced BOM to make the power supply designer's job easier.

Let's switch over to our second example of how TI's integration technology can help solve your power density challenges. Our MicroSiP and MicroSiL families are integrated DC-DC converter modules, which reduce board area by embedding a Picostar die into the substrate with converter filter components on top.

The pictures on the right show both module families. Both are similarly constructed with a Picostar die embedded inside the PCB substrate with the passive components, chip inductors, and ceramic capacitors soldered on top. The MicroSiP includes the power module capacitors and inductor for a complete power supply solution. And the MicroSiL is for high-power applications and only improves the power module inductor.

The MicroSiL devices have BGA solder bumps on the bottom, while MicroSiL uses QFN type of packages with thermal pad. The thermal pad is useful to get the heat out of these higher-current MicroSiL devices.

Looking at a MicroSiP in detail, we can see the PCB substrate in green, with the silicon die and a Picostar package embedded inside. In addition, we have the inductor on top of the substrate and the BGA solder bump on bottom, giving the overall thickness of the MicroSiL. MicroSiL devices are typically taller in height due to the taller inductor used for the higher currents. This integrated module technology leads to significant space savings on a designer's PCB.

Here, we are looking at how much board space can be saved by comparing the discrete TPS62085 design with the TPS82085 power module design. Although the discrete device already has quite a small solution size of only 62 millimeters squared, the MicroSiP module, with its 3D construction, only requires 35 millimeters squared, which is about 40% smaller. In addition, this technology allows an optimized printout to simplify the layout on the designer's PCB, thus making it even simpler to integrate on your design.

So why use a MicroSiL? Well, the first reason is for smaller solution size. When you set the components vertically, as we do here, as opposed to side by side on the PCB, the x and y size is reduced. The height, of course, increases a little bit when you stack the components, but this is not so critical for many applications. Specifically, we can achieve around about a 40% smaller xy size with a MicroSiP compared to an equivalent discrete solution.

The second reason to use a MicroSiP power module is ease of use. With zero, or few external components required, the power supply design is very much simplified. And the PCB layout is very much simplified as well. Your procurement team also just needs to take care of one component instead of many. The Darnell Group did a survey amongst power design engineers, and according to this survey, it takes about 45% less time to design a module compared to the discrete version.

The third reason is performance. Integrating all the components into one design gives a repeatable and expected EMI and noise performance, versus doing your own PCB layout. And this is true on different systems and different projects. You can use the same MicroSiP on many different projects in many different systems and get the same EMI and noise performance.

If you tried to lay out the same design on different PCBs, there will be small differences in the layout that can affect those parameters. The passive components have been selected to optimally work together with the embedded eyes, leading to a higher conversion efficiency. Additionally, TI also qualifies these passive components to our high quality standards.

Our third example highlights significant power density improvements in isolated bias supplies. The UCC12050 utilized TI's breakthrough integrated transformer technology to miniaturize isolated power transfer in an IC-sized package. The UCC 12050 provides 500 milliwatts of isolated power in an SOIC16 package that has a solution size that is more than 60% smaller than discrete or powered modules solutions.

The integrated transformer technology is very efficient and has low EMI, allowing for a very simple PCB layout. TI's UCC12050 is a fully reinforced 5 kV isolated bias supply. The integrated transformer technology has a 1.2 kilovolt working voltage with ultra low primary to secondary capacitance, making it an ideal solution for gate driver and many other applications.

In comparing the UCC12050 solution size to existing solutions, the 2.65 millimeter SOIC package is much thinner than existing modules. In many applications, the bias supply is the tallest component on the PCB. By reducing the height, the power density of the power supply can be increased. The integrated UCC12050 is smaller than discrete board designs, with the surface-mount transformer. The UCC12050 solution size is at 32% smaller in area.

TI's isolated bias supply have higher efficiency and better package thermal performance than competing Aircore bias designs. As you can see from the images above, the UCC12050 runs considerably cooler, with only a 15 degrees centigrade rise above ambient. This makes the PCB thermal design much simpler and can reduce the PCB area needed for cooling.

TI's integrated bias supply has also the industry's lowest EMI. As can be seen above, the EMI is much lower than competing Aircore designs. The UCC12050 EMI performance eliminated the need for discrete EMI filters that would increase PCB area and the bill of material.

In this video, we discussed three examples of TI's many integration technologies. Each of these devices enables the power supply designer to improve power density and performance and, at the same time, reduce the BOM and design time. TI's LMG3410R070 GaN with integrated driver provides high-efficiency switching with protection to provide a robust power supply design.

TI's MicroSiP and MicroSiL power modules provide the smallest module solutions with 45% less design time. The UCC12050 is the smallest isolated bias supply, with improved efficiency and low EMI. For more information about TI's power products, please visit www.ti.com/power.