I have a custom board on which a Raspberry Pi Compute Module 5 is mounted.
I have noted multiple occasions when the CM5 fails to start.
What seems to happen is that when I apply 5V or strobe PWR_Button low, then CM5_3.3V and CM5_1.8V will come up for about 12 seconds then drop back to zero.
This only happens "next day" i.e when I have been using the CM5 one day, the sudo shutdown it and remove power. Next morning I apply power and it fails to start. (And if I sudo shutdown on Friday night, then come back on Monday then it does start correctly).
Removing then reapplying power, or strobing PWR_Button doesn't help.
Things that get the CM5 to start are:
Leaving the CM5 with power applied for a few minutes, then strobing PWR_Button
Removing the CM5 and sticking it on a Compute Module 4 I/O board and applying power to that. (I don't have a CM5 I/O board).
My guess is that something is failing to discharge fully and so defeating a step in the power-up sequence of the CM5. But what? My 5V goes down to less than 0.1 V in less than a second so I don't think it is that.
And which step? What happens after about 12 seconds in the normal start-up sequence of a CM5?
I don't have a battery to keep the RTC running (although the board has provision for this).
Due to an oversight, we don't have a working LED hanging off LED_nACT. Is it likely that this pin has useful diagnostic information? If so I shall leave a 'scope hanging off it for tomorrow.
I have noted multiple occasions when the CM5 fails to start.
What seems to happen is that when I apply 5V or strobe PWR_Button low, then CM5_3.3V and CM5_1.8V will come up for about 12 seconds then drop back to zero.
This only happens "next day" i.e when I have been using the CM5 one day, the sudo shutdown it and remove power. Next morning I apply power and it fails to start. (And if I sudo shutdown on Friday night, then come back on Monday then it does start correctly).
Removing then reapplying power, or strobing PWR_Button doesn't help.
Things that get the CM5 to start are:
Leaving the CM5 with power applied for a few minutes, then strobing PWR_Button
Removing the CM5 and sticking it on a Compute Module 4 I/O board and applying power to that. (I don't have a CM5 I/O board).
My guess is that something is failing to discharge fully and so defeating a step in the power-up sequence of the CM5. But what? My 5V goes down to less than 0.1 V in less than a second so I don't think it is that.
And which step? What happens after about 12 seconds in the normal start-up sequence of a CM5?
I don't have a battery to keep the RTC running (although the board has provision for this).
Due to an oversight, we don't have a working LED hanging off LED_nACT. Is it likely that this pin has useful diagnostic information? If so I shall leave a 'scope hanging off it for tomorrow.
Statistics: Posted by ClassicRover — Wed Feb 11, 2026 9:32 am