We are developing a custom board based on the ESP32-P4 and are experiencing a consistent issue where the chip fails to boot on initial power-up. The CHIP_PU pin stabilizes at approximately 1.2V instead of the expected 3.3V, and the USB Serial/JTAG interface is never recognized by the host PC. A single press of the reset button (which momentarily pulls CHIP_PU to GND) always resolves the issue, and the system boots and runs normally afterward.
We followed the recommended circuit from the ESP32-P4 Hardware Design Guidelines for the CHIP_PU pin (10 kΩ pull-up to 3.3V + 100 nF capacitor to GND, with a pushbutton to GND for manual reset).
Our 3.3V rail is stable and correct. The external DCDC (TLV62569) is connected as per the reference schematic, with EN_DCDC and FB_DCDC controlled by the ESP32-P4.
Observed behavior
On cold power-up (board fails to boot):
- 3.3V rail comes up correctly and is stable at 3.3V
- CHIP_PU stabilizes at ~1.2V (not 3.3V as expected)
- EN_DCDC remains at 0V — the external DCDC (TLV62569) never activates
- VDD_HP (1.2V digital supply) does not come up
- USB Serial/JTAG is not enumerated by the PC
- The chip appears to be stuck in a partial/failed boot state
- CHIP_PU is pulled to GND momentarily by the button
- Upon release, the chip boots correctly
- EN_DCDC goes high, TLV62569 activates, VDD_HP comes up
- USB Serial/JTAG is recognized immediately
- System runs normally
- However, CHIP_PU still reads ~1.2V (appears to be clamped internally by the chip)
- The board boots correctly on the first attempt, every time
- This confirms that the issue is related to the power-on sequence, not a hardware defect
- Test 1 (reference design): R = 10 kΩ, C = 100 nF, τ = 1 ms → CHIP_PU ~1.2V → Fails to boot
- Test 2: R = 10 kΩ, C = 1 µF, τ = 10 ms → CHIP_PU ~0.6V → Worse, voltage dropped further
- Test 3: R = 10 kΩ, C = 4.7 µF, τ = 47 ms → CHIP_PU ~0.6V → Worse, very slow ramp, low average voltage
- Test 4: R = 1 kΩ, C = 100 nF, τ = 100 µs → CHIP_PU ~1.32V → Slightly higher voltage, still fails to boot
When we increased the capacitor from 100 nF to 1 µF (Test 2), the steady-state voltage on CHIP_PU dropped from 1.2V to 0.6V. This is counterintuitive for a simple RC circuit — in DC steady state, the capacitor should not affect the final voltage.
Reducing R to 1 kΩ (Test 4) increased the available current to ~2.1 mA and raised CHIP_PU to 1.32V, but this was still not enough to achieve a successful boot on power-up
Our analysis
The root cause appears to be a power-on sequencing issue. During cold boot, CHIP_PU rises simultaneously with the 3.3V supply through the RC circuit. The chip begins its boot sequence before the analog/LP power domains are fully stable, fails to complete the DCDC enable sequence, and gets stuck.
When the reset button is pressed (or held during power-up and then released), CHIP_PU transitions cleanly from 0V to its operating level after all 3.3V rails are already stable. This allows the boot ROM to execute correctly and enable the external DCDC.
The ~1.2V steady-state reading on CHIP_PU appears to be normal behavior for the ESP32-P4 — the pin seems to have an internal clamp. This is different from other ESP32 variants where CHIP_PU reaches VDD.
Questions:
1) Is the ~1.2V clamp on CHIP_PU expected behavior for the ESP32-P4? The datasheet and hardware design guidelines do not explicitly mention this internal clamp behavior. On other ESP32 variants, CHIP_PU reaches VDD (3.3V).
2) The ESP32-P4-Function-EV-Board reference design does not include a reset supervisor. Given the behavior we are observing, how does the official development board handle the power-on sequencing? Does it rely on specific power supply characteristics (fast rise time) that may not be present in all custom designs?
3) Are other users experiencing similar issues with ESP32-P4 custom boards? We would like to understand if this is a known limitation or if there is something specific to our design causing this behavior.
Environment
Chip: ESP32-P4 (32MB)
External DCDC: TLV62569 (from Espressif's verified list)
3.3V supply: Stable, verified at 3.3V
CHIP_PU circuit: Pull-up resistor to 3.3V + capacitor to GND + pushbutton to GND
ESP-IDF version: v5.5.x
Thank you in advance for any guidance.
