By Hesam Moshiri, Anson Bao
A tiny DC to DC buck converter board is useful for many applications, especially if it could deliver currents up to 3A (2A continuously without heatsink). In this article, we will learn to build a small, efficient, and cheap buck converter circuit.
[1]: Circuit Analysis
Figure 1 shows the schematic diagram of the device. The main component is the MP2315 step-down buck converter.
Figure 1
Schematic diagram of the DC to DC buck converter
According to the MP2315 [1] datasheet: “The MP2315 is a high frequency synchronous rectified step-down switch-mode converter with built-in internal power MOSFETs. It offers a very compact solution to achieve 3A continuous output current over a wide input supply range with excellent load and line regulation. The MP2315 has synchronous mode operation for higher efficiency over output current load range. Current mode operation provides a fast transient response and eases loop stabilization. Full protection features include OCP and thermal shut down.” Low RDS(on) allows this chip to handle high currents.
C1 and C2 are used to reduce input voltage noises. R2, R4, and R5 build a feedback path to the chip. R2 is a 200K multiturn potentiometer to adjust the output voltage. L1 and C4 are the essential buck converter elements. L2, C5, and C7 make an additional output LC filter that I added to reduce the noise and ripple. The cut-off frequency of this filter is around 1KHz. R6 limits the current flow to the EN pin. The R1 value has been set according to the datasheet. R3 and C3 are related to the bootstrap circuit and determined according to the datasheet.
Figure 2 shows the efficiency vs output current plot. The highest efficiency for almost all input voltages has been achieved at around 1A.
Figure 2
Efficiency vs output current
[2]: PCB Layout
Figure 3 shows the designed PCB layout. It’s a small (2.1cm*2.6cm) two layers board.
I used the SamacSys component libraries (Schematic symbol and PCB footprint) for the IC1 [2] because these libraries are free and more importantly, they follow the industrial IPC standards. I use the Altium Designer CAD software, so I used the SamacSys Altium plugin to directly install the components libraries [3]. Figure 4 shows the selected components. You can search and install/use the passive components libraries as well.
Figure 3
PCB layout of the DC to DC buck converter
Figure 4
Selected component (IC1) from the SamacSys Altium plugin
This is the last revision of the PCB board. Figure 5 and figure 6 show 3D views of the PCB board, from the top and bottom.
Figure 5
A 3D view of the PCB board (TOP)
Figure 6
A 3D view of the PCB board (Bottom)
[3]: Construction and Test
Figure 7 shows the first prototype (first version) of the board. The PCB board has been fabricated by the PCBWay, which is a high-quality board. I had no problem with soldering whatsoever.
As it is clear in figure 8, I have modified some parts of the circuit to achieve lower noise, so the provided Schematic and PCB are the latest versions.
Figure 7
The first prototype (an older version) of the buck converter
After soldering the components, we are ready to test the circuit. The datasheet says that we can apply a voltage from 4.5V to 24V to the input.
The main differences between the first prototype (my tested board) and the last PCB/Schematic are some modifications in the PCB design and component placement/values. For the first prototype, the output capacitor is only 22uF-35V. So I changed it with two 47uF SMD capacitors (C5 and C7, 1210 packages). I applied the same modifications for the input and replaced the input capacitor with two 35V rated capacitors. Also, I changed the output header’s location.
Since the maximum output voltage is 21V and capacitors are rated at 25V (ceramic), then there should not be a voltage rate problem, however, if you have concerns regarding the capacitors’ rated voltages, simply reduce their capacitance values to 22uF and increase the rated voltages to 35V. You can always compensate this by adding extra output capacitors on your target circuit/load. Even you can add a 470uF or 1000uF capacitor “externally” because there is no enough space on the board to fit any of them. Actually, by adding more capacitors, we decrease the cut-off frequency of the final filter, so it would suppress more noises.
It is better that you use the capacitors parallelly. For instance, use two 470uF in parallel instead of one 1000uF. It helps to reduce the total ESR value (the parallel resistors rule).
Now let’s examine the output ripple and noise by using a low-noise front end oscilloscope such as Siglent SDS1104X-E. It can measure voltages down to 500uV/div, which is a very nice feature.
I soldered the converter board, in accompany with an external 470uF-35V capacitor, on a small piece of DIY prototype board to test the ripple and noise (figure 8)
Figure 8
The converter board on a small piece of DIY prototype board (including a 470uF output capacitor)
When the input voltage is high (24V) and the output voltage is low (5V for instance), the maximum ripple and noise should be generated because the input and output voltage difference is high. So let’s equip the oscilloscope probe with a ground-spring and check the output noise (figure 9). It is essential to use the ground-spring, because the ground wire of the oscilloscope probe can absorb a lot of common-mode noises, especially in such measurements.
Figure 9
Replacing the probe’s ground wire with a ground-spring
Figure 10 shows the output noise when the input is 24V and the output is 5V. It should be mentioned that the converter’s output is free and has not been connected to any load.
Figure 10
Output noise of the DC to DC converter (input =24V, output = 5V)
Now let’s test the output noise under the lowest input/output voltage difference (0.8V). I set the input voltage to 12V and the output to 11.2V (figure 11).
Figure 11
Output noise under the lowest input/output voltage difference (input=12V, output=11.2V)
Please note that by increasing the output current (adding a load), the output noise/ripple increases. This is a true story for all power supplies or converters.
[4] Bill of Materials
Figure 12 shows the project’s bill of materials.
Figure 12
Bill of materials
If you want to build this circuit, you can download the Gerber files or directly order the PCB to be fabricated for you with the highest quality.
References
[1]: https://www.mouser.com/datasheet/2/277/MP2315_r1.01-478439.pdf
[2]: https://componentsearchengine.com/part.php?partID=930350
[3]: https://www.samacsys.com/altium-designer-library-instructions
I see others post an email so mine is ken@knr.net Thank you
Did you meet any problems when use the gerber.cam file?
Hello, I will answer the question here. I will send you the gerbers by Email because I can not upload them here. so your first question has already answered. 2) if you want to get the best of this design for lower voltages than 5V, you must also modify some other component values. The datasheet is straightforward. invest a few minutes and read that. especially you can find best component values for your desired voltages. I designed an "optimum circuit" to cover a voltage range and be variable, but in your case, you can get lower ripple and higher stability if you follow my suggestion. I saw a commercial design and I compared both. This design worked much better. so as I said follow the datasheet, you can either find the resistor values and optimize other parts based on your desired output and then yes you can put fixed resistors instead of that potentiometer
Thank you very much for the files and your answers! I am no way near an expert in terms of electronics, just in need of a relatively simple PCB with some connectors. Instead of taking some external buck converters that connect to this PCB I thought about integrating them directly onto the board (with some jumpers to switch between output voltages/resistors). While researching I stumbled upon your post and this good performing buck converter so I tried to understand and rebuild it. There is no need to go lower than 5V at the moment so it would be okay for me if it isn't perfect. I think I have to learn on how to read those datasheets properly. ;)
After digging more through the datasheet it seems it is not simply done by changing one component (resistor) to change the output but five... or is there a possibility to do that? The potentiometer is the only component that changes in your circuit so it should be possible?!
Something like this, but with your design (which should be much better, like you said, and doesn't burn when used with 3A output: https://ae01.alicdn.com/kf/H28b79b855f6c416cb5cb1421b7a08be7g.jpg https://ae01.alicdn.com/kf/Hf1025b5f03a14b689e49bf4a01730daeO.jpg
in this design, because I tried to cover a range of output voltages, it is easier. you only need to fix the output at 5V by the pot. then remove the potentiometer and measure its resistance. then try to find best resistor to fit in. also if you want, I can design a new project for you, but if you want to learn, you have to invest some time. don't worry
also, please be careful, this circuit can not deliver 3A continuously. Just 2A continuously. 3A with heatsink and short term. if you want a powerful design, you can use my previous circuit or let me know to design it for you personally. contact me by Email in this case
Thank you VERY much! I will try to make it on my own but this is my very first PCB so I hope I can do it. ;) Also I have to order every resistor/capacitor/part because I don't have those SMD laying around. Measuring the potentiometer and selecting the resistors afterwards is a good idea (the PCB can be designed to put different resistors afterwards I think) but I have to wait for the ordered resistors after measuring which can take a while. :)
Yes, first measure the value, then find the closest 1% resistor to it. also you can match that final resistor value with series/parallel resistors
hello, I can't reach the gerber extension, can you send me an e-mail? ilyasdmrts@gmail.com Thanks