DIY High Efficiency 5V Output Buck Converter!

I wanted an efficient way of stepping down higher voltages from LiPo packs (and other sources) to 5V for electronics projects. In the past I have used generic buck modules from eBay, but the questionable quality control and no name electrolytic capacitors did not fill me with confidence.

So, I decided that I would make my own step down converter to not only challenge myself but to make something useful also!

What I ended up with is a buck converter that has a very wide input voltage range (6V up to 50V input) and outputs 5V at up to 1A load current all in a small form factor. The peak efficiency I measured was 94% so not only is this circuit small but it stays cool too.

While you can certainly make a buck converter with a handful of op-amps and other supporting components, you will get better performance and certainly save a lot of PCB area if you instead pick a dedicated buck converter IC.

You can use the search and filtering functions on sites like DigiKey, Mouser, and Farnell to find a suitable IC for your needs. In the above picture you can see a daunting 16,453 parts get narrowed down to 12 options in just a few clicks!

I went with the MAX17502F in a tiny 3mm x 2mm package, but a slightly bigger package would probably be better if you plan on hand soldering the components. This IC has a lot of features, most notable of which is the large input range of up to 60V* and the internal power FETs that mean no external MOSFET or Diode is needed.

*Note that in the intro I stated it was 50V input yet the part can handle 60V? This is due to the input capacitors and if you need 60V input the circuit can be modified to suit.

More often than not, there will be what's called a "Typical Application Circuit" shown in the datasheet which will be very similar to what you are trying to achieve. This was true for my case and although one could just copy the component values and call it done, I would recommend following the design procedure (if provided).

Here is the datasheet of the MAX17502F: https://datasheets.maximintegrated.com/en/ds/MAX17502.pdf

Starting on page 12 there are about a dozen very simple equations that can help you choose more suitable component values and it also helps to provide details about some of the threshold values - such as minimum inductance value.

Wait I thought we already did this part? Well, the previous part was to find the ideal component values, but in the real world we have to settle for non-ideal components and the caveats that come with.

As an example, Multi-Layered Ceramic Capacitors (MLCCs) are used for the input and output capacitors. MLCCs have many benefits over electrolytic capacitors - especially in DC/DC converters - but they are subject to something called DC Bias.

When a DC voltage is applied to a MLCC, the capacitance rating can drop by up to 60%! This means your 10µF capacitor is now just 4µF at a certain DC voltage. Don't believe me? Have a look at the TDK website and scroll down for characteristic data for this 10µF capacitor.

An easy fix for this type of issue is simple, just use more MLCC in parallel. This also helps to reduce voltage ripple as the ESR is reduced and is very common to see in commercial products that need to meet stringent voltage regulation specs.

In the above images there is a schematic and corresponding Bill of Materials (BOM) from the MAX17502F Evaluation Kit, so if you can't seem to find a good component choice then go with the tried and tested example :)

With your actual components chosen it is time to create a schematic that captures these components, for this I chose EasyEDA as I have used it before with positive results. Simply add your components in, making sure they have the right size footprint and connect the components together just like the typical application circuit previously.

Once that is complete, click on the "Convert to PCB" button and you will be brought to the PCB Layout section of the tool. Don't worry if you aren't sure on something as there are many tutorials online about EasyEDA.

PCB layout is very important and it can make the difference between the circuit working or not. I would strongly advise to follow all of the layout advice in the datasheet of the IC where available. Analog Devices has a great application note on the topic of PCB Layout if anyone is interested: https://www.analog.com/media/en/technical-documentation/application-notes/an136f.pdf

I'm sure most of you at this point have seen the promotional messages in youtube videos for JLCPCB and PCBway, so it shouldn't come as a surprise that I used one of these promotional offers too. I ordered my PCBs from JLCPCB and they arrived just over 2 weeks later, so just from a monetary standpoint they are quite good.

As for the quality of the PCBs I have absolutely no complaints, but you can be the judge of that :)

I hand soldered all of the components onto the blank PCB which was quite fiddly even with the extra room I left between the components, but there are assembly services by JLCPCB and other PCB vendors which would eliminate the need for this step.

Hooking up power to the input terminals and measuring the output, I was greeted by 5.02V as seen by the DMM. Once I verified the 5V output across the entire voltage range, I connected an electronic load across the output which was adjusted to 1A current draw.

The Buck started straight up with this 1A load current and when I measured the output voltage (at the board) it was at 5.01V, so the load regulation was very good. I set the input voltage to 12V as this was one of the use cases I had in mind for this board and I measured the input current as 0.476A. This gives an efficiency of roughly 87.7% but ideally you would want a four DMM testing approach for efficiency measurements.

At 1A load current I did notice the efficiency was a bit lower than expected, I believe this is due to (I^2 * R) losses in the inductor and in the IC itself. To confirm this, I set the load current to half and repeated the above measurement to get an efficiency of 94%. This means that by halving the output current the power losses were reduced from ~615mW down to ~300mW. Some losses will be unavoidable, such as switching losses inside the IC as well as quiescent current, so I am still very happy with this result.

Now you have a stable 5V 1A supply that can be powered from a 2S to 11S lithium battery pack, or any other source between 6V and 50V, there's no need to worry about how to power your own electronics projects. Be it microcontroller based or purely analog circuitry, this little buck converter can do it all!

I hope you enjoyed this journey and if you have made it this far, thank you very much for reading!