CrowPanel ESP32 5.79inch E-paper Display – USB-C Charging and USB Serial Not Working When Battery
Hi everyone,
I'm using the CrowPanel ESP32 5.79inch E-paper HMI Display and found two issues that only happen when a Li-Po battery is connected.
I wanted to share my findings here in case others experience similar behavior, and to ask Elecrow engineers whether this is expected or a design oversight.
1) Charging does not start if the battery is connected first
When a Li-Po battery is attached to the BAT connector before USB-C:
- Plugging in a USB-C PD power source does not raise VBUS to 5V
- Charging never starts
Charging works if:
- USB-C PD is connected first, or
- A USB-A → USB-C (non-PD) cable is used
So the issue only happens when the connection order is:
Battery → USB-C PD
2) USB serial port never appears on macOS when a battery is connected
If the battery is attached:
- My Mac never recognizes the ESP32 as a USB serial device
- No
/dev/cu.*port appears at all
If I disconnect the battery and power the board with USB-C only:
- The serial port appears immediately
- USB works normally every time
This suggests the USB enumeration failure may share the same root cause as the charging issue.
3) Measurement results (VCC of the charger IC)
I measured the VCC pin of the DFN8_4054A charger IC, which is directly tied to USB-C VBUS:
| Condition | Measured VBUS/VCC |
|---|---|
| Battery only | ~2.0 V |
| USB-C PD only | 5.0 V |
| Battery first → USB-C PD later | ~1.6–2.0 V (never rises to 5V) |
So VBUS sits around 2V even with no USB cable connected.
4) Possible explanation (not an expert)
I’ve heard many USB-C/PD power sources behave like this:
- They only enable 5V if VBUS is near 0V at connection time
- If VBUS already has some voltage, they assume the device might be sourcing power
- As a safety precaution, they refuse to enable 5V
If this is true, then the ~2V leak on VBUS (likely coming from the charger IC) could explain:
- Why USB-C PD never starts charging
- Why macOS never detects the USB serial device when a battery is connected (USB enumeration fails)
USB-A sources do not check VBUS, so they always work.
5) Schematic observation
From the 5.79-inch CrowPanel schematic:
- USB-C VBUS is directly connected to the charger IC’s VCC
- No ideal diode or reverse-blocking FET is used
This may allow BAT → VBUS leakage, which creates the unexpected ~2V on VBUS.
(Just my interpretation — I may be wrong.)
If Elecrow engineers are reading this, could you please confirm:
- Is this behavior expected?
- Should VBUS remain at 0V when no USB is connected?
- Is there any recommended workaround?