We designed a baseboard for CM4, and it works well. When using CM5, approximately 1% of cold starts fail to boot up, especially when an HDMI device is connected. After inspection, it was found that there is no +3.3V output on the core board, which indicates that the PMIC is not working. I'll call this baseboard Board 1.
On another baseboard, it takes about 1 second for the +5V power supply to stabilize. In this situation, CM4 boots up normally, while CM5 definitely fails to start. I'll call this baseboard Board 2.
When using Board 2, since GLOBAL_EN (PMIC_EN) is used, I can pull down this signal during power - on and release it after 2 seconds. This solves the problem.
When using Board 1, although the failure to boot up is the same as that of Board 2, the oscilloscope recording shows that the power - on waveform is very good. I wonder if the firmware logic of the PMIC is too strict?
Remark: I have updated the EEPROM firmware to the latest version. It has solved the problem that the PMIC cannot be powered off during shutdown, but the issue of failure to start still occurs.
On another baseboard, it takes about 1 second for the +5V power supply to stabilize. In this situation, CM4 boots up normally, while CM5 definitely fails to start. I'll call this baseboard Board 2.
When using Board 2, since GLOBAL_EN (PMIC_EN) is used, I can pull down this signal during power - on and release it after 2 seconds. This solves the problem.
When using Board 1, although the failure to boot up is the same as that of Board 2, the oscilloscope recording shows that the power - on waveform is very good. I wonder if the firmware logic of the PMIC is too strict?
Remark: I have updated the EEPROM firmware to the latest version. It has solved the problem that the PMIC cannot be powered off during shutdown, but the issue of failure to start still occurs.
Statistics: Posted by hiheyoh613 — Wed Mar 19, 2025 6:10 am — Replies 4 — Views 128