nRF52840 VBUS state when on battery power

[ilia]

Hello,

I have custom board with nRF52840. Most of the time It works on USB power, but it also has a battery. It has some peripherals that powered via USB too (LED w2812b array). When USB is disconnected I want it to go to power save mode, so I need to shut down LEDs. I can do it via LED control pin. But I see that on the LED power pin I still have 2.5V (maybe battery is not fresh).

Is it possible that battery current goes through VBUS into USB circuit?

[Stian Røed Hafskjold]

I can't see anything obviously wrong with the regulator design. The diode D1 should prevent any leakage back from 3.3V to 5V.

How do you turn off the LEDs? Are you pulling the GPIOs high?

Can you post the rest of the schematics, including the nRF52 and the GPIOs. Thanks

[ilia]

I have control over the LEDs signal input pin, and not over the power pin. To turn off the LEDs I just send "black color" signal. But it should not be necessary if there is no power on the LEDs power pin.

[Stian Røed Hafskjold]

I think you should try to remove one component after another to see where the leakage current goes.

[Stian Røed Hafskjold]

Can you describe the LEDs you are using? Is it just normal LEDs you control by pulling the GPIOs high or low, or is it a more fancy RGB LED you control with commands? Like your "black color" command..? If the data bus between the nRF chip and the LEDs are still active there's a good chance that these are pulling up the 5V power line to the LEDs.

[ilia]

I use an array of w2812b RGB LEDs. They are controlled via commands. I use SPIM to deliver commands. LED changes its color after receiving a command, and stays in this state until it receives command with new color. Particular command is basically a sequence of bits representing RGB color bytes.

I don't think that leakage occurs in the LEDs, because even if all LEDs are physically disconnected, I have current around 2.5V on the power pin for those LEDs.

[Stian Røed Hafskjold]

Ok, then I don't know what the problem is. But in general if you remove the power to an external device, you must also disconnect the data lines between the nRF52 chip and the external device, otherwise you might power the external peripheral through the data lines.

What you can do is to set all GPIOs on the nRF52 to inputs, then they will all be disconnected, and you can be sure that no leakage current is going out from the GPIOs. If you still see 2.5V on the 5V line there must be something external to the nRF52 that is leaking.