Dear Seeed Studio Team,
please provide a circuit, which allows to connect external power supply 5V DC to the 5V Pin for the XIAO ESP32-S3 and XIAO ESP32-C3 boards.
The mentioned on the wiki pages suggestion to connect the 5V to the board via “…some sort of diode (schottky, signal, power) between your external power source and this pin…” does not work.
The direct connection from 5V external supply to the 5V pin gives the same results like for the solution with the diode.
I have tested above solution on the two XIAO ESP32-S3 boards and three XIAO ESP32-C3 boards.
Boards were connected to 3-7 Analog/digital sensors (INPUT), SPI (device), WiFi, PSRAM (on S3).
I used exact the same external power supply to power the boards in all tests: USB, direct 5V pin, via schottky diode to 5V pin.
Observed board behaviors:
- when comes to boot process on wifi with setting IP address the board makes restarts in the loop
- board stops booting when setting WiFi address to the board
- from time to time board has started but read data at the startup process from PSRAM where not correct
All above disappears at when the board is powered via the USB connector - all boards worked stable.
Solving the above problem opens for me the possibility to use XIAO boards in wider range of applications.
Thank you in advanced.
What kind of external power supply are you using?
When using wireless, a peak current of 350 mA or more is drawn. Can it be supplied by an external power supply?
1 Like
sounds like a brownout issue to me… maybe add a power capacitor
we really need a XIAO power management expansion board
Hi there,
SO can You provide a picture of this Apparatus? I would like to verify your assumptions, and if non proprietary please post up the code you are trying using the code tags above “</>”
also as @msfujino inquires , the power supply parameters you are trying
TIA
GL 
PJ 
it does seem that this is a common complaint that some do not “Accept” power thru the 5V pin… also the question is how low is the threshold voltage to the regulator… how low can you go? I assume it is not a boost converter
I guess one could use an arduino Uno as a 5v power supply for testing
There is a relevant description in the datasheet. I think the input voltage needs to be 3.3V plus the voltage drop of the high-side FET.
100% Duty Cycle Operation Mode
When the voltage of the input source, such as battery, falls and its value is close to the output voltage, the PWM duty cycle (D = VOUT/VIN) increases to near 100%. Eventually the high-side switch remains continuously on to keep the output regulated. Even when the input voltage falls below the output, the high-side switch is turned on to minimize the error.
1 Like
Hey SEEED Studio! I have this problem as well on an Xiao ESP-32 C6. The issue is that the power ramps up on the 5v line as a turbine generator spins up. The capacitor / resistor circuit on the EN circuit isn’t tuned optimally for this. It works great on USB power that is an instant, perfectly regulated 5v, but that is not the case in my application. It would be nice to see your boards incorporate something like an mcp130 as they are tiny and quite inexpensive. I lost several hours trying to understand what was going on as the chip was powered, 5v was stable, 3v3 was solid and the charge light even came on and the chip was warm…
Hi there,
And Welcome here…
So, On XIAO boards, the EN pin typically uses a weak pull-up (100 kΩ–1 MΩ) with a small C (~0.1–1 µF). When VBUS ramps slowly (e.g., from a spinning turbine), this:
- Causes partial EN activation,
- Powers the chip partially (warm to touch),
- But fails to start properly.
Unfortunetly you can’t have it both ways as built.
Use a Soft Power-Up Supply (e.g. Supercap + Fast Diode + LDO)
Design your power circuit so that:
- A Schottky diode charges a supercapacitor or large ceramic cap from the turbine.
- Then feed that into a low dropout regulator (LDO) with a Power-Good (PG) pin.
- Use PG to gate the LDO or control a switch to 3.3 V rail or USB port.
This ensures the LDO doesn’t output until it sees steady voltage is one way to do it that I know works.
I have used a small power bank in between the Charge cable and powered it with a solar panel power mc4 input, for example.
HTH
GL PJ
When you insert a USB power bank or regulated USB device between your turbine and the XIAO, it:
- Buffers the input voltage from the turbine’s ramp-up (slow/dirty 5V).
- Provides a clean, regulated 5V USB output to the XIAO.
- Holds enough internal charge to avoid brownouts on minor dips.
- Effectively delays VBUS rise seen by the Xiao until it’s stable.
This “virtualizes” the power-up curve and mimics the stable power expected by the ESP32’s internal EN circuitry.
Wow thank you for the well thought out response sir.
Indeed I do see that circuit design in the schematic tied into the reset button. If everything wasn’t super tiny surface mount I’d play around with the values. The lDO on the Xiao board works fine for supplying power but there is no supervision on the power state. That means the LDO can throw voltage that is lower than functional at the MCU. The purpose of the MCP130 is to perform a reset once when stable power is reached. If I can’t solve it with a capacitor on the supply I think I’ll go that route. The thing is I can’t use a battery, the power has to come straight from the generator and the loss of power from the generator is important for the operation of the system as crazy as that sounds. Basically the ESP running is a tattletale that the generator is active. There are a lot of ways I could have approached this but this way is cheap, simple and no batteries to go bad.
All that said I don’t have direct experience with the MCP130 or variants. They may be on some of the other boards I have though I’m not sure. Any advice there?
1 Like