[Open Source] SC8703-based Buck-Boost DCDC Adjustable Power Supply with Voltage and Current Regulation, Supporting PD Fast Charging Input

Open-source Buck-Boost DC/DC power module based on SC8703 chip, with input voltage up to 36V, input/output current up to 10A, switching frequency of 600kHz, adjustable output voltage, adjustable constant output current, maximum conversion efficiency of 95%, XT30 input/output connectors, Type-C port, supports PD fast charging input, capable of negotiating up to 28V.


Design Parameters

Parameter Value
Input Voltage Range 2.7V ~ 36V
Output Voltage Range 1.21V ~ 36V
Max Input/Output Current 10A
Adjustable Constant Output Current Range 0.58A ~ 10A
Switching Frequency 600kHz
PCB Size 79.6 x 47 mm

The Buck-Boost power controller chip is SC8703, and the fast-charging protocol chip is CH224Q.


Replication Notes

  1. This project was designed using KiCAD. The project on LCSC’s Open Source Platform was imported from the KiCAD project file, so there might be minor issues. For replication, it is recommended to directly use the Gerber files I provided to order PCBs.
  2. The project is open-sourced on Gitee, LCSC Open Hardware Platform, and Huaqiu Open Hardware Community. Project materials and links to open-source platforms are available at the bottom of this article. Provided materials include: KiCAD project files, Gerber fabrication files, detailed BOM (with component purchase links and prices), combined top and bottom stencil file, silkscreen placement diagrams, interactive BOM (for soldering assistance), datasheets of used chips, PCBA and enclosure 3D models, etc.
  3. The enclosure is a ready-made aluminum alloy case purchased from Taobao. The purchase link is included in the BOM. The front and rear panels of the enclosure were 3D printed by myself; the 3MF model files are included in the provided materials.
  4. The MOSFETs must make thermal contact with the enclosure using thermal silicone pads!
  5. When operating without additional heatsinking, it is recommended to keep the working current below 5A!
  6. The Type-C input and XT30 input connectors are connected in parallel — do not use both simultaneously!

The component cost for small batches is approximately 30 RMB per set (excluding PCB and enclosure). The main cost driver is the SC8703 chip, which costs about 9 RMB per piece.

:play_button: Project video demonstration: https://www.bilibili.com/video/BV1mMfPBzEXH/
:play_button: QQ Group for Electronics/MCU Technical Discussion: 2169025065
:play_button: eeClub - Electronic Engineers Community: https://bbs.eeclub.top/


Physical Photos


Soldering Process

Order a 10x10 cm stainless steel stencil from Taobao using the stencil file I provided (costs about 15 RMB). Then 3D print a solder paste applicator jig to securely align the stencil with the PCB. Apply solder paste, place components according to the interactive BOM, and then place the board on a heating plate or reflow oven. It is recommended to solder the bottom side first, then the top side.






Testing

Power-on test: 12V input, 24V output, no-load power consumption around 0.37W.

Type-C input test: Using a 65W PD charger, successfully negotiated 20V output.

Conversion Efficiency Test

Simple unboxing review of the Ruideng RD6012P digital programmable power supply: https://blog.zeruns.com/archives/740.html


Maximum measured conversion efficiency: 95.45%

Input Voltage (V) Input Current (A) Input Power (W) Output Voltage (V) Output Current (A) Output Power (W) Efficiency (%)
35.999 0.752 27.071 24.130 1.000 24.130 89.14%
35.999 6.699 241.157 23.954 9.600 229.955 95.35%
5.000 9.063 45.315 24.028 1.500 36.042 79.54%
5.000 8.520 42.600 12.054 3.000 36.162 84.89%
12.000 4.382 52.584 5.230 9.000 47.070 89.51%
11.999 8.753 105.027 24.048 4.000 96.192 91.59%
5.000 9.803 49.015 19.990 2.000 39.980 81.57%
5.000 5.063 25.315 21.116 1.000 21.116 83.41%
11.999 6.705 80.453 35.477 2.000 70.954 88.19%
35.999 3.459 124.521 11.998 9.510 114.097 91.63%
35.999 1.839 66.202 12.023 5.000 60.116 90.81%
35.999 9.288 334.359 35.460 9.000 319.141 95.45%

MOS Gate Drive Waveforms

Waveform of high-side MOSFET gate-to-ground on the input side (Buck section): switching frequency is 607kHz, period ~1.65μs, Miller plateau duration ~20ns — normal, clean waveform with no ringing and minimal overshoot.


Waveform of high-side MOSFET gate-to-ground on the output side (Boost section): switching frequency is 607kHz, period ~1.65μs. Overshoot and ringing occur during MOS turn-on, likely due to suboptimal gate drive routing.


Ripple Test

Simple unboxing review of the Rigol DHO914S oscilloscope: https://blog.zeruns.com/archives/764.html

With 12V input and 24V output at no load, measured ripple peak-to-peak is ~60mV.


With 12V input and 24V/2A output, measured ripple peak-to-peak is ~200mV.


With 24V input and 5V output at no load, measured ripple peak-to-peak is ~16mV.


With 24V input and 5V/2A output, measured ripple peak-to-peak is ~35mV.


Thermal Imaging

Without any heatsinking, when operating at 36V input and 24V/9.5A output, MOSFET temperature rises above 100°C within 5 seconds. Bottom-side thermal image shown below.

Front-side thermal image: SC8703 chip temperature exceeds 70°C.

Thermal image with aluminum enclosure installed. Due to reflectivity of aluminum affecting measurement accuracy, a piece of tape was applied to the center. The temperature at the MOSFET contact surface is measured at approximately 53°C.


Download Links

Materials provided include: KiCAD project files, Gerber fabrication files, detailed BOM (with component purchase links and prices), combined top and bottom stencil file, silkscreen placement diagrams, interactive BOM (for soldering assistance), datasheets of used chips, PCBA and enclosure 3D models, etc.


Recommended Open-Source Projects


Recommended Reading

2 Likes

Nice project :+1:

Great and concise project, thank you so much for sharing! I have a question for you: in the schematic, the CH224K chip is configured to negotiate 28V, but during testing it negotiates 20V—why is that?

Maximum spoofing up to 28V, but it’s not guaranteed to achieve 28V—it depends on whether the charger supports it. Generally, 100W or 65W PD chargers have a maximum output voltage of 20V. Additionally, 140W chargers come in two types: one uses a modified PD protocol to achieve 20V/7A, while the other uses the standard PD protocol with 28V/5A.

1 Like

Alright, alright, umm, got it, thank you.

Excuse me,大佬, in the schematic diagram, R11 and R12 are loads. During no-load testing, should these two resistors not be soldered?

This is the current-limiting resistor for the LED; you can leave it un-soldered if you don’t want the power indicator light.

Hmm, okay, I got it. There’s just one thing I don’t quite understand—where exactly should the load be connected? Can the VOUT output be directly connected to the load?

Yes, you can connect a load to the Vout output; you can adapt the XT30 interface to other types.

Alright, got it

Author, why is there almost no current when using the PD charging port for power? The input voltage is normal at 20V, and the output is 24V, not the 12V output voltage shown in your test.

Looking at the picture, there is a current input of 0.04A; this doesn’t mean no current—it’s called no-load current. Your output isn’t connected to any load, and even the output terminals are left empty without soldering!

This design allows adjustable output voltage—adjust it using that potentiometer.

Have you ever studied basic circuit theory? Do you even know what a DCDC power supply is and what power supplies are used for?