Thanks, Emily. Hi, everyone. My name is Wilson Fu. And thank you for joining and taking time out of your day to join us to learn more about linear and low dropout regulators here at TI.

So here's what I'll be going over today. Go over a brief overview of our LDO portfolio at TI, where and how to find LDO's using TI.com/LDO. Our key investment areas, moving forward, new product highlights. I'll actually go over two new products that were recently released. And then we'll end up with just reviewing some additional resources for y'all, all right. Let's begin.

So here, we can see we've broken down our LDO portfolio here at TI. We offer over 500 plus devices covering a wide range of applications as you see here, from automotive to personal electronics to even industrial equipment and telecommunications. We go into about 80% of all applications. And you know, that makes sense, seeing as how a majority of electronics need power.

Now, given that we have over 500 plus devices, we actually wanted to make it easier for engineers to find the correct device for their applications. You can see we have three pillars broken out by input voltage-- low, mid, and wide voltage, based off of the voltage ranges you see here.

Once the input and output voltage has been decided, we then differentiate depending on the needs of the application. Select low noise for precision applications such as radar, data converters, PLL's, and even heart rate monitors and smartwatches. Low quiescent current for power savings and battery powered equipment, such as portable electronics, or even systems powered off of batteries and cars and drones.

We also have our older robust tried and true solutions and our linear regulators such as our Allan 317's that have withstood the test of time. And also our standard LDO's for powering more generic rails. So as you can see, TI truly has an LDO to cover all your needs.

Now, the easiest way to find our LDO's is through TI.com/LDO, which I've referenced in this QR code. What you'll find is our LDO landing page. As mentioned earlier, you can find the breakout of low, mid, and wide VM, as well as links that take you to the parts that fall under those sections. We continue to also differentiate a bit further by adding negative as well as automotive and space grade LDO's.

You also see our latest and greatest parts. On here we have our latest low IQ LDO and TPS7902 boasting the lowest quiescent current in the market with 25 nanoamps. Now, this would be critical for current and future needs as applications become more portable and battery reliant.

Lastly, you'll find our TI power trends. TI has been at the forefront of developing new process packaging and circuit design technologies to deliver the best power devices for your design. For LDO's, we focus on these three-- power density, so achieving more power in smaller spaces and enhancing system functionality over reduced system costs, low quiescent current, extending battery and shelf life without compromising system performance, and low noise and precision, as I mentioned, enhancing power and signal integrity to improve system level protection and accuracy.

All right, moving forward, these are our focus areas and key investment areas. Now, TI in general is focused on automotive and industrial. And I've listed some technology highlights per market segment. For automotive, we do include our wide input voltage operation to sustain high voltage transients when powering direct from car battery. We also have the highest accuracy across temperature and are also being functional safety compliant.

On the industrial side, we are working towards expanding our low quiescent current family with the TPS7902 and other products. We have leaded and small form factor devices that are common throughout the industry. And we also provide the lowest output noises on some of our devices.

In these market segments, we typically focus on these four attributes, wide input voltage, as I mentioned earlier, increase reliability with the highest working voltage. So in terms of automotive, it's really to make sure we can survive those high input voltage transients. And for industrial, just make sure we can be used off of typical voltage rails such as 12 volts or 24 volts. Low quiescent current, so extending battery and shelf life, low noise, enhancing power and signal integrity and small size, high performance and smaller space.

Now, we'll mention that although the products I'll be going over aren't necessarily wide input voltage, they're design could actually be used for the future wide input voltage products. We'll also go over some of our newer wide voltage parts as a second part to this webinar series at a later date, all right.

So one of our recent releases earlier this year is our TPS7A52-53 and 54 device family. On the top right is a QR code linked to the product folder. This device was actually released earlier this year. This family of devices would actually fall under our power density and low noise power trends, as we're actually dissipating 2 to 4 amps and a very small 2.2 by 2.5 millimeter package while maintaining 4 microvolts for our massive noise and fairly high accuracy.

You can see that this type of device is really meant for high speed analog circuits, imaging, as well as medical in order to take advantage of the performance these devices bring. In the past, we've had the CPS78 family as shown at the bottom, which offered programmable output that we call any out through an internal resistor network.

Up top we actually have the automotive versions of these devices which expands upon the junction temperature range up to 150C. These devices particularly go onto radar platforms where they actually need the ultra low noise performance and small form factor. With the TPS785 family, we decided to make improvements listed here from these devices, so improved dropout, improved PSRR, and similar performance to that of the TPS7A8 and a smaller size.

Now, you can see here a comparison between a traditional QFN versus our HotRod or enhanced QFN package. Now, thermals is one of the most common challenges we face as power engineers. There's generally a trade-off between size and thermal dissipation to the point where it may not make sense to reduce overall system size only to find that it fails multitudes of tests due to the thermal performance suffering.

With the TPS7A5 family, we've introduced this HotRod technology that can be found on some other devices as well at TI in order to improve thermal performance upon the existing family. The biggest differences that I'll point out here are the PassFET location as well as the removal of the silver die attach. So you can see the PassFET has moved from on top of the die to underneath the die on top of these copper pillars, hence HotRod, which replaced the silver die attach and the wire bonds. This actually allows for improved thermal and power dissipation within a smaller package by having it spread out more evenly versus focused in one area.

This improvement in size and thermals is really critical for applications that are being reduced in size. So for instance, an application such as base stations or medical imaging, we've seen the reduction of PCB space while maintaining the same power dissipation for instance, reducing about 4 PCB's to 1 PCB while maintaining the same power dissipation. It's absolutely critical that our devices can survive and maintain signal integrity given these higher power conditions.

Now, in order to prove this, we actually ran an experiment with the TPS7853 versus the previous generation TPS7884. Here's a total solution size comparison between the two devices with the external components included using the same bill of materials. As you can see, there's about a 55% reduction in the size of the device, and about a 27% reduction in overall solution size.

Here is the EVM comparison of the two devices with the components. Now, as you can see here, we used a pretty standard PCB size with four layers. You can also see on these images the locations of the thermal vias which are actually just the small holes that you can see on the screen. These allow for better heat dissipation in the PCP.

Now, since the TPS78 family has that big thermal pad, we actually primarily had the thermal vias underneath it for optimal thermal dissipation. For the 785 family, since it doesn't have that thermal pad, we instead placed the thermal vias strategically on this device. As you can see here, the vias is actually stacked towards one side of the device.

This is ultimately done based off of the input and output pin locations on this device. And you can actually refer to that pin out either on the previous slides or actually on the datasheet. So in this experiment, we assumed a one watt power dissipation evenly distributed on a PassFET area. And we also assumed a 25c degree ambient temperature.

What's interesting to note is that the HotRod packaging actually allows heat to spread more evenly versus the TPS78 family which radiates directly from the PassFET. As I mentioned before, this even spread results in more heat transfer out of the package and thus reducing the overall max die temperature. I will note on the screen, there are actually two scales here. So even though they have similar colors, the TPS7A5 is less than the TPS788.

You can see that the results are almost a total of 3 degrees C difference between the two for the max die temperature while being in a much smaller solution size. This is absolutely critical in applications that have thermals as a concern, such as the base stations and imaging equipment that I mentioned earlier. But this also plays in servers and data centers for enterprise systems where you have very compact stacks of PCBs. And they can actually take advantage of this improved performance and smaller size.

With that being said, I actually wanting to take the time to go over some questions. But it doesn't look like there are too many questions thus far. So I'll just move on to the next device, OK.

So the next device I would like to introduce is actually our new TPS7A20 device. Now, this device actually just released a few weeks ago. But this LDO is what I really like to call a workhorse LDO.

Up top in the top right is a QR code to the product folder. And truly, this device can really do it all. As a combination of very low noise, ultra low IQ, as well as high PSRR and fast trending responses while being in a small solution size.

Now, what some people don't realize is that there are definitely some trade offs when designing a device like this. The biggest one is maintaining low IQ in transient performance with low noise. And what I mean by transient performance is really just a behavior of the LDO when responding to changes in load current for line voltage.

So generally, there are two primary sources of noise in semiconductors. We have thermal and flicker noise, which has these effects on a power versus frequency scale. So you can kind of see that thermal noise being fairly flat across a wide frequency with that roll off due to the amplifier bandwidth. And then flicker noise, which we'll see more at lower frequencies versus higher frequencies.

And also the resultant flicker plus thermal noise is generally the shape of our traditional noise plots they will see in our data sheets, and also on some plots moving forward. For LDO's, we primarily see the bandgap reference as the primary source of noise. This can actually be gained up as a function of V out as well.

So now that we understand some basic sources of noise, there are also some design techniques to reduce this internal noise. We can lower the bandwidth of the amplifier. And we can also increase the bias currents. Now, this in turn will also reduce the size of the resistors and then reduce the thermal noise based off of the Johnson-Nyquist equation on the screen.

However, if you turn these knobs enough, we can actually have the opposite effect of what we actually want, meaning, we'll have a slower transient response and higher quiescent current. So based off of these two simple techniques, you can also see that it's very difficult to maintain a balance between low noise, low IQ, and fast transient response.

And since all these parameters can be fairly dynamic, the easiest way that we can visualize this is with a comparison. So here is an older LDO, low IQ LDO, the TPS706 which was designed for strictly the low IQ market with one microamp of quiescent current. As you can see, it results in 190 microvolts VRMS. So that's between the 10 Hertz to 100 kilohertz bandwidth and has a fairly large undershoot, so almost 400 millivolts in this low transient response which goes from no load to full load, as well as a fairly slow settling time. So I believe on this scale it's around a little bit greater than 100 microseconds, so pretty slow.

On the TPS7820 however, even though we have 6.5 microamps of quiescent current, you can see that we've accomplished a 30 times improvement over something like TPS706 for noise. And I'd given similar conditions for low transients, so no load to full load. We can see about 200 millivolts voltage undershoot with a 5 microsecond recovery time. That's about 100 times faster than the TPS706 with about half the undershoot.

Another important factor to consider for TPS7920 is actually the PSRR and dropout. So PSRR is the Power Supply Rejection Ratio, which is an indication of how well the device can filter out the input ripple. So this is typically from an upstream power supply, such as a buck converter. The most common definition of drop out is the difference between input and output voltages. While this may be true, it's actually more the essence of Ohm's law where voltage is equal to resistance times the current. In this case, it would be the drop out is equal to the RDS on of the passput times the output current.

Another word that could be used here is actually headroom. So LDOs require a certain amount of dropout voltage in order to perform optimally. We refer to this optimal drop out as headroom, and usually can be seen in our test conditions in the sense of the optimal drop out between the input and the output voltage.

But why is this important? You know, as electronics start to become smaller, we're seeing the trend of improving performance in a lower power state whether that be lower voltages or lower output current or both. As you can see, when the input voltage of the LDO starts to decrease, the LDO will lose its ability to filter the input ripple since output voltage will start to track the input voltage. And this is just because the pass fill will be moving from the saturation region to the linear region. The figure on the top right actually depicts this behavior at 2.8 volts with varying input voltages. And you can see as it gets lower, it starts to filter out less.

Since dropout is oftentimes linear with alpha current based off of Ohm's law, it is also important to note that PSRR will increase as output current decreases as shown in the bottom right figure. This is because the output impedance of the pass element is fairly high at low currents and increase the DC gain of the output stage. And this results in PSRR to increase at lower frequencies.

At heavier or full loads, the output impedance of the pass element decreases and thus lowering the gain and lowering PSRR. Again, since these are fairly dynamic parameters, it's actually just a lot easier to visualize what kind of an impact this has on an application.

So typically, we see these low noise and high PSRR requirements in camera applications or pairing image sensors. Critical frequency range is about 101 kilohertz to 1 megahertz. And having insufficient PSRR or higher noise of the power supply baring the image sensor will result in pixelated images. As you can see in the images below, when conducting a black image test where you essentially just cover the camera, you can see that a high noise LDO, or a high noise power supply in general will result in this discolored pixelated image to the bottom left. Whereas the low noise LDO will result in the perfect black image that's right next to it.

These level requirements are really being driven by the power trends I spoke about earlier. So the small size will result in smaller batteries, which means less overall power consumption, which in turn will drive lower output voltages while maintaining and also improving the noise and PSRR performance.

Now, there are also some additional protection features inherent to the design allowing this to also be designed in industrial type of applications as well. Inrush current control is one of these features, upon system startup, the power supply will actually ramp up to the regulated voltage. And as the voltage increases an inrush of current flows into the uncharged capacitors. Inrush current can also be generated when a capacitive load is switched onto a power rail and must be charged to that voltage level.

This inrush current can typically result in the input voltage dropping. To fix this, we actually recommend increasing the input capacitor to help supply the load current. But there's also some downsides to this as well, such as the upstream supply may also go into current limit if the inrush is too high and the addition of size and cost due to the size of the additional capacitor. This also limits the amount of capacitors you can then put on the output.

So as you can see from this plot, this is a 500 milliamp LDO. And that does not have proper inrush control. And the spike goes all the way up to 1.35 amps. On the TPS7820, you can even see that under similar conditions, it does not even hit its own current limit. Which means that you can minimize the need for that input capacitance, or even really eliminate it altogether in some cases.

Another improvement that was done on this device was the ability to handle up to 200 microfarads of capacitance on the output. This helps both the PSRR at higher frequencies, which are dominated by the pull of the output capacitor and helps them maintain regulation during extreme low transients. Having this inverse control feature truly allows this device to loosen requirements in strict industrial systems.

Another additional feature I'd like to point out is the noise reduction capacitor, or the NR cap. Typically, on our older devices as you see on the right with the TPS734, we include this noise reduction capacitor to filter out the bandgap noise and connect it directly to the NR pin. It's really optimal for these lower or older type of LDOs. But with TPS7820 what we've done is we actually use an internal RC filter to remove the need for the external cap.

So instead of filtering the band gap directly, we're actually now filtering after the bandgap reference is gained up, which will minimize the impact of the output voltage on noise and also maintain a very small PCP area size. So now that I've gone over a few of these, I actually wanted to see if there's any questions, No, OK.

Now that we've briefly gone in depth about some of our new products, I actually wanted to share with you some of our other LDO products that can be leveraged for now and in the future. Starting on the left, you can select our latest and greatest LDOs based off of your input voltage and socket needs and then view some of our leading existing solutions and their subsequent next generations, such as our TPS7805 to our TPS7A02, or the TPS7A8 to the TPS7A5 family, and even our LP5907 to our TPS7820. The QR code should also take you to our existing LDO quick reference guide for even more devices to select from.

Hey, Wilson, it looks like we actually have two questions in the chat.

Oh, sure thing. Let me-- if I can open the chat. Sorry, hold on. OK, so the first question is, is it always necessary to have an inboard cap to the LDO, and if there are parts that don't require input caps? So no. It is not always necessary to have an input cap on LDO. We actually have a few devices such as our TLV733, and TLV713 that actually don't require an input or output capacitor.

The second question is, why the critical frequency from 1 kilohertz to 1 megahertz? And that critical frequency is actually more just for image sensors. Ideally, we actually have a 100 kilohertz as the sweet spot for the frequency. That if you have sufficient PSRR, then it will be able to reject all the upstream noise. And so 100k is typically where we would have that metric. But in general since there is a wide range of image sensors in the market, 1 kilohertz to 1 megahertz is the range.

Let's see. How is the inverse current limited for the device of TPS7820. So we actually have a current limiting circuit within the TPS7820 that you can actually find on the dat sheet and goes a little bit more in depth on there. OK. I think that's all the questions I'll answer for now.

Kind of going back to this. As I mentioned before, that QR code will take you to that existing LDO quick reference guide for even more devices to select from. And lastly, if you want to learn more about LDOs, please visit our LDO basic series as referenced in the QR codes. We have our e-book as well as videos that go more in depth about various LDO parameters. As always, if you have any technical questions, I encourage you to leverage our fantastic ET forums, both TI experts to answer your questions.

With that, I will end here. Thank you for taking the time out of your day to learn more about TI and LDOs. As always, please visit www.ti.com/npu. Prefer the new product updates from other product lines coming in the future. Thank you.

Thank you everyone for joining us. And next week's topic will be buck DC/DC switching regulators. Please refer back to TI.com/npu like Wilson said, and you'll find the PDF version of the slot as well as the reporting for the webinar. Thank you for joining.