Expected power consumption for Mesh node

[choehyunho]

Hi teams,

We made BLE  mesh custom boards based on nRF52833.

These nodes are configured to send broadcast messages for every 10 seconds.

Currently, these boards uses 12~13mA in provisioned state.

Is this normal and expected or possible to reduce currents.

TIA

[Mttrinh]

Hi,

You can try with the DC/DC regulator, this should reduce the current consumption.

[Mttrinh]

Hi,

Can you check if you have the necessary components to use the DCDC:

For the nRF52833 you need the L3, LF4 and C15.

If you have these components, the PMU will control which regulator that will be used. The DCDC will be enabled in the SDK.

Link to the reference circuitry for nRF52833.

[choehyunho]

sd_power_dcdc_mode_set() makes no difference whether I put it before/after mesh_init().

Is there any chance to reduce currents?

[Mttrinh]

It is weird that the current aren't reduced when the DCDC is enabled. Do you get any return codes when calling that function?

[choehyunho]

Hi, Mttrinh.

I am using Mesh SDK 4.1.0 and the result is following.

1. Call sd_power_dcdc_mode_set(NRF_POWER_DCDC_ENABLE) before ble_stack_init()

    Return 2 (= NRF_ERROR_SOFTDEVICE_NOT_ENABLED)

2. Call that function after ble_stack_init()

    Return 0 (= NRF_SUCCESS), but still there is no change in power current.

[Mttrinh]

It might be that the DCDC isn't enabled at all even though it return success. Can you try with a nRF5 example and see if you can see a difference in consumption when having it turned on and off?

[choehyunho]

Hi, Mttrinh.

I tested two examples.

The one is ble_app_uart in nRF SDK. If DC/DC enabled, the current is changed from 4.4~5.3mA to 4.4~4.9mA. The difference is quite small, but definitely different from each other.

But the other example, which is in nRF SDK for Mesh, beaconing shows no difference whether DC/DC enabled or not.

Mittrinh, can you check the reason why sd_power_dcdc_mode_set() not affecting in Mesh SDK?

[Mttrinh]

That is quite a small difference. Do you have a DK available you can test on? To eliminate that it could be an hardware issue.

You could also have a look at the LPN example for reference, the example enables the DCDC by default.

[choehyunho]

I compared my board with DK, and DK shows current reduced when DC/DC enabled.

After my team reviewed current board design comparing with Reference Circuitry, it follows Configuration #2.

Now we are trying to change it to Configuration #4, which can be used with DC/DC enabled.

I'll leave comments after we changed to.

[choehyunho]

I checked the schematics with my team.

We designed based on configuration #4 already, but some pull-up resistance for other chip caused leakages.

Now we removed pull-up resistence & current consumption is now around 7mA.

Thank you.