Power Electronics - Buck Converter Design Example - Part 1

and welcome back to

power

electronics

. In this video segment we are going to examine the

design

of a

buck

converter

and divide it into two

part

s. This first video will cover the first

part

and in that we will look at the size of the The components of much of the material we use for the

buck

converter

and

power

stage can be found in this application report by Bridget Hawk of TI and I have the link here and I'll link it to the description as well in this quick overview video. Let's look at the

design

requirements and specifications and this is a somewhat composite design but realistic in the numbers we are using.

Let's estimate the switching duty cycle based on the input and output voltages. We are also going to estimate the output current based on the power requirements that you will see in the requirements specifications for this design and from there we will look at the size of the components and there are four basic components that we will look at. when sizing the inductor, then the capacitor, then the diode and the MOSFET and when you go through a design, you don't have to do it in this order, this is typically the order I do them in, but design is a very iterative process and even as I was making this video I was looking at maybe I should choose this component over that component and I'll describe some of those scenarios when we get there so let's look at the requirements for this design for this is a DC to DC converter and we have 15 volts as input nominal, five volts as nominal output and is a 10 watt power supply or a 10 watt DC to DC converter.

I'm using nominal values ​​and those are more of a requirement and I'm not really providing detailed specifications on these numbers; in a more formalized design we would have firm specifications, so for

example

we would limit the input voltage with a positive or negative voltage rating, likewise with the output, and we would probably have a do not exceed the output power value and Finally, the last requirement is that we would like to see the efficiency be around ninety percent once we put that efficiency at ninety percent, which will bring us into the realm of a switch mode power supply. like the buck converter is what we have been researching for the past week.

I'm going to make a couple of assumptions. I will assume that we will do a continuous driving mode for our buck converter again, we will be on a switch mode power supply and using the 10 watt output I can calculate an average output current equal to 10 watts divided by 5 volts and that will give us a rated output current which is equal to 2 amps, so we will be able to work with that in our design and we can also estimate the duty cycle based on our input and output voltage. Remember that our output voltage V out is equal to the input voltage times the duty cycle, therefore V out divided by VN will equal the duty cycle again. just an estimate, many times we will use some type of controller to automatically regulate and adjust the duty cycle as the input and output voltages fluctuate with the load and using this as an estimate we see that we have 5 volts divided by 15 volts and that gives us a duty cycle of about a third or thirty-three percent is what we can estimate for our duty cycle.

Now let's move on to sizing the inductor based on this. You'll notice that in the requirements we didn't have a specification for it. ripple current through the inductor a standard rule of thumb is our Delta I should not exceed, well let me say Delta I over I out should not exceed 40%, so we don't want to exceed 40% of the output. I will be more. conservative and I'm going to start with a delta I equal to 20% of our output current, so 0.2 times two amps will give us this ripple current through the inductor and that's equal to 0.4 amps. They also didn't give us a switching frequency I'.

I'm going to try to use a smaller space and I'm going to want to increase our switching frequency so without a switching frequency and sometimes the switching frequency will be mandatory for you. I'm going to assume a switching frequency of 500 kilohertz with the switching frequency and the ripple current and our known output voltage and our estimate of our duty cycle, we can calculate the inductor using the following equation L equals V multiplied by 1 - D/FS multiplied by Delta I putting all those numbers in place, that gives us five times two thirds, all divided by 0.5 e6.

I'm writing our frequency for our switching in megahertz, which will provide our numerator in micro farads, so there's a reason I wrote our switching frequency that way. multiplied by Delta I, which is 0.4, we see that our inductance is equal to sixteen point six microhenries, so that will give us a range for our inductor. Before selecting an inductor, we also have to look at the average current that will flow. through the inductor because the value of the inductance itself and the current that flows through the inductor on average are what will help us size that inductor correctly.

Let's move on to sizing the inductor using a search engine like the one found on Digit Key or Mouser. I took a quick look. for an inductor in the 22 micro henry range and we see here that we have one that is 22 micro henries and it was born and I also have to look at the current, that is the other specification that we have to look at and this is the rating. in this device there was 6 amps RMS for current and because we are at almost DC, 6 amps RMS will also exceed our current requirement of 2 amps at DC, the downside of this comes later when we do the power analysis or calculations. efficiency and it is is the equivalent DC resistance due to the windings and I'm going to use the worst case values ​​that we see that we have for the 22 microhenry inductor.

We have 72 miles. Remember that my calculation was originally 16 point 6 microhenries and I. I am thinking of increasing the size but I also assumed that a 20% ripple current if I reduce the size to 15 micro henries it will slightly increase the ripple current, it will not be more than 40% but it will reduce our DC resistance to 50 mil amande, what could help with our efficiency is something we need to keep in mind as we move through the process. Do I want to increase efficiency at the expense of higher ripple current? Let's move on to the sizing of the capacitor.

The single capacitor sizing is equal to delta io divided. times 8 times the switching frequency multiplied by Delta V oh, we were not given a ripple voltage in the requirements or specifications. I'm going to estimate and assume that 5 volts plus or minus 0.01 volts or 10 millivolts should work fine for this design. Again I'm filling in the blanks in the design but that's something you sometimes have to do at least to get a first cut on the values, this gives us a delta V equal to 0.02 volts and we size the capacitor using all this values. see that C is equal to 0.4 divided by 8 again.

I'm going to write our switching frequency as 0.5 megahertz, so 0.5 ee 6 times 0.02 doing these calculations we see that the value of the capacitor is equal to 5 micro farads again, we can go up or we can go down. I'm going to go down slightly, which means our ripple voltage at the output will go up slightly, but I'm going to select a 4.7 microfarad capacitor and size it at 35 volts DC, so that's a typical size, you'll find capacitors with ratings 35 volts, 50 volts, 100 volts, our output will only be 5 volts, our input will only be 15 volts, so this is a pretty good range.

I think I decided on a TD K C Series and again I will link the data sheets in the catalog for this TD k1 that I selected is a multilayer ceramic chip style capacitor. I'm doing all the surface mount components in the selection process here, okay, we have the inductor size, we have the capacitor size, let's look at the diode size because our efficiency spec was 92%. I'm going to go with a Schottky diode which I haven't shown here, but now I'll show it there. so I'm going to go with a Schottky diode. Your average current rating has to be equal to two amps times 1 minus D and that's the average value if you average this current over a full switch frequency, that's the number we would get.

If I design for two amps, I will be safe: our maximum repetitive voltage for the diode must exceed 15 volts and with the Schottky we will see that we will have a very low forward conduction voltage and I hope that this will be less than half a volt again, trying to reduce our conduction loss when the diode is conducting and recovering the diode behaviors when the MOSFET is open during the second half of that switching period, that second half, so again I went to Mouser. Some people like to use digikey.com, you see, and this is a bit, not a bit.

It's pretty misleading that it has an average forward voltage of 30 amps and up here it says five and I'm going to talk a little bit. about that, but it has an ultra-low voltage drop of 0.3 in forward conduction, which is really good. The other good thing about Schottky diodes is that I don't have to worry about reverse recovery, it is a majority carrier device so we don't have hole recombination during the turn-off phase of this diode, so again keep that in mind quickly that there are no reverse recovery losses in Schottky diodes and this one will work quite well, but I want to go back to this 30 amp and 5 amp thing and see what it is Continuing with this device, here's more from the data sheet again, the data sheets are linked below and this graph here is really interesting, this is a surface mount device and it talks about the ambient air temperature and we have talked about 85 degrees C like our industrial.

SPAC and if I go into this device and then follow it, we see what our true average direct current rating is for this device and it appears to be about two amps, which is right where we are for our designs and our average amps will be. slightly below and the other thing to keep in mind is that our instantaneous forward voltage depends on our forward current and we will see that we have a forward voltage which is around 3.3 volts, if we allow the ambient temperature to rise to 125, it will be So. We will be within range for this device.

You have to delve into the data sheets of these devices when you are doing your design. The devil is in the details. We have one last component to size and that is the MOSFET, so let's size the MOSFET. First we will look at the average current through the MOSFET. Here we see that the average current wallet is shorted or conductive. Here we notice that our drain-to-source voltage is close to zero. In fact, the voltage is equal to our drain to source. The current multiplies our drain to source resistance while it's conducting, but we look at that value and it's two amps, it's our average value across the inductor, but the average value that the MOSFET will have will be IO multiplied by D again if we average this out. the total switching period we would get I Oh multiplied by D and that is equal to, for this

example

, two thirds of an ampere or 0.6 seven amperes.

We also have to look at the drain to source reverse breakdown voltage and so we are going to have to select a Mosfet that can have a drain source greater than 15 volts, so those are the two parameters to consider and again I did a search on Mouser and several devices came out and I was looking for a surface mount device and this is the part that I am going to select for this design and you will see that one of its applications is the DC to DC converter, it is a 30 volt device and the current ratings .

There is 20 amps in it, but again, as we saw with the diode, we are going to have to go into the details because for all of these devices I have said this before and I will say it again, it is our junction temperature that will determine whether the device will work or not in its design, if we can keep the junction temperature below the maximum value in the datasheet then the device will be sized appropriately, let's dig a little deeper. I have an expanded version of this. data sheet and I will use it in the next section when we look at the efficiency calculations in the next video.

I'm going to use the rise time and fall time from the data sheet again and I said that sometimes these are not the best ones to use, but if we look at them we have an on and off time of 12 plus 9 nanoseconds, which which gives us a total on and off time of 21 nanoseconds and we're going to use that value in our efficiency calculation, if we had a slightly larger generator for our gate it changes a bit and this was based on a drain current of 10 amps , we're at 2 amps, so it's going to be even less than 21 nanoseconds by how much.

It probably doesn't matter too much in our design, at least for our efficiency calculations, so that's the size of theMOSFET. Here are the on and off calculations. I've shown them once before. One of the things to keep in mind for this device: our threshold voltage. If you look at the data sheet on the last slide, it was about two volts and our Miller plateau voltage will be about two volts, which implies that the gate to source minus the threshold and the gate to source minus the plateau are approximately equal, which would give us ln 1 + ln 1 is zero, so we would have a very slow time: t2, if you remember, was the time it took to go from zero to the threshold and, well, actually t2 is the time it took from the threshold to the Miller Plateau and t3 was the time it took from the Miller Plateau until we got to the linear or ohmic region and then the gate to the source started to rise to its final value, so that's our T for this case the t2 will be close to zero as will our t6 because our plateau and threshold are almost identical they will also be close to zero so most of the time we will decrease and be at the plateau voltage from Miller so that our gate to source voltage, so that's the turning on and off of the MOSFET again.

We're going to assume 21 nanoseconds when we do the efficiency calculations, so let's recap what we did in this video. We use the input and output voltages to estimate the duty cycle we use. output voltage and output power requirements to estimate the output current, that output current I sub 0 and an assumed Delta I was used which I assume is around 20% along with an assumed switching frequency again. I assumed a switching frequency of 500 kilohertz to size the inductor, we sized the capacitor again, we had to make some assumptions and then we sized the diode and used a Schottky for low loss and finally we sized the MOSFET and estimated these on and off times of the datasheet in 21 nanoseconds, there will probably be less for our application, but it is a good starting point for the next part of the design.

In the next video we will see the efficiency and where the losses occur in this circuit. Remember one of the requirements that was presented in the previous slide that we wanted to achieve an efficiency of I think it was greater than 90%, so we want an efficiency greater than 90%, so thank you for watching this and watch the next video to see How do we do it. Let's do the energy losses in this specific design.