Skip to main content

Issue with sleep mode in low voltage situation

2 years ago

Issue with sleep mode in low voltage situation

Posted byjetsonindustries0 points 1 reply
0 upvotes

Hey Team,
We have a device using PXP_reporter project.
When we put the device in extended sleep mode, it advertises for 8 seconds then stops if the voltage is less than 2.90V.
If the voltage is over 2.90V we don't have a problem and the device maintains advertising and operates fine.
While not in sleep mode the device operates fine right down to 2.17V which is our curcuits minimum voltage.

We have tried numerous tests including turning off brown out detection, various different power sources and even different hardware designs, all with the same result of some form of crash or hang-point of advertising while in sleep mode when the voltage drops below 2.90V.

When we try booting a device straight into sleep mode, we see it boot, there is roughly 8 seconds advertising time and then the device stops advertising and gets stuck. When we try to use an interrupt to wake it out of sleep mode, advertising doesn't come back on.

Another experiment we did:
1. Powering system with 2.5V: system advertise and then go to low power mode. At low power mode I measure V14 (this is Radio power, 1.4V) I saw 0.3V
2. Powering system with 3.3V: system advertise and then go to low power mode. At low power mode I measure V14, I saw 1.3V.

Would you have any suggestions on what might be causing this issue?

Best regards,
Ben

accepted answer!

2 years ago

PM_Dialog

Hi jetsonindustries,

Could you please clarify if it is battery powered? If yes can you please check dg_configBATTERY_LOW_LEVEL?

If Vbat drops below this limit, the system enters hibernation mode, waiting either for the battery to the changed or recharged. Also, do you switch off V33 power rail?

In active mode V33 gets the value of LDO_CTRL1_REG[LDO_SUPPLY_VBAT_LEVEL], which can be from 00 = 2.40V till 11 = 360V.

In sleep mode V33 gets the value of LDO_CTRL1_REG[LDO_VBAT_RET_LEVEL], again from 2.4V till 3.6V

Deactivating this rail it’s not an option, as it will cause GPIOs floating, and waking up could be a major risk.

V33 should be powered by the LDO_VBAT_RET, a clean 3V should be measureable in the V33 pin (or anything programmed in the respective register).

If you kill this rail, then GPIOs will not be powered. GPIOs being floating means that

  1. 传感器连接到它会泄漏,由于floating pins (control or power)
  2. Waking up the system will be pretty tricky.

Thanks, PM_Dialog