SCK2.2 - SCK2.3 reset loop

[oscgonfer]

The SCK2.2 and SCK2.3 have an issue on boot while powered only via USB. This does not happen consistently, but is fairly easy to reproduce with the normal USB charger provided and with the SEN5X and the urban board, but without a battery:

This is due to the higher cumulative energy consumption by several systems, but more importantly, because the SEN5X has a peak consumption at boot that seems higher that the PMS5003, even if the sensor consumes less power on normal operation.

smartcitizen-kit-2x/sam/src/SckUrban.cpp

Lines 2103 to 2107 in 49b152f

	// The SCK needs battery to survive the sensor startup current draw, with only usb power it normally does not work
	// If battery is not present it can enter a reset loop, (not always)
	pinMode(pinPM_ENABLE, OUTPUT);
	digitalWrite(pinPM_ENABLE, HIGH);
	delay(25);

I will add more graphs here showing the consumption during boot. Posibly related to fixes in #97

[pylonsoft]

The bug is not related to the AP fixes for the ESP. The reset happens before the ESP is activated as soon as the SEN5X is powered. The over-current must be very short as my USB power meter doesn't notice a spike or anything similar.

After a few reboots the boot completes and if the power is interrupted shortly after that it will boot correctly again. That leads to the assumption that the power needed to heat the gas sensors may be at least one reason for the reboots. I cannot explain yet, why that doesn't show up in the USB power consumption. Especially as the power consumption is much higher later on when the ESP is started.
I had to disconnect power for more than 10 minutes to see a (rather short) reboot loop again.

My experience of stopped reboots after a while may be caused by my configuration as my SEN5X (I have a SEN55) is powered constantly. But deactivating VOC and NOx doesn't changed anything.

[pylonsoft]

I added some code to put out the reset reason at startup. During the reboot loop about 2/3 of the reset reasons was the brown out detector on the 3.3V voltage the rest were Power On Reset. That means the 3.3V drops 1.75V (BOD33) or even 1V (POR). As the USB voltage doesn't drop and there is (visible) spike on the USB current consumption, my current guess is a special behavior of the battery charger chip. I will study the datasheet for anything that leads in that direction.