nRF52 current consumption

RE: nRF52 current consumption

Stian Røed Hafskjold — Mon, 06 Feb 2017 12:10:54 GMT

No, you can do that, as long as you have control on how the application interrupts are handled and you know when power_manage() is returning to main loop. Or, you can use the app_timer library to set up a timed event, or use the scheduler to schedule application tasks.

RE: nRF52 current consumption

sfujimoto — Fri, 03 Feb 2017 10:00:44 GMT

thanks, Although there is power_manage () in the main loop, we describe other processing, is it a rule violation?

RE: nRF52 current consumption

Stian Røed Hafskjold — Thu, 02 Feb 2017 14:09:26 GMT

You can try to adjust the two parameters above to 40ms. I think that should be enough for the amount of data you transmit.

There is a low power parameter for the ADC in the sdk_config.h file. Look for SAADC_CONFIG_LP_MODE.

As long as you call the power_manage() function in the main loop, the CPU should be sleeping between the events.

RE: nRF52 current consumption

sfujimoto — Thu, 02 Feb 2017 13:22:13 GMT

thanks,

Is the connection time below? MIN_CONN_INTERVAL MSEC_TO_UNITS (20, UNIT_ 1 _ 25 MS) MAX_CONN_INTERVAL MSEC_TO_UNITS (20, UNIT 1 _ 25 _MS) Where do I set the parameters? Just send 433 bye of transmission data consecutively every second. Other than that, BLE is off for more than half. Sensor is many, it is about several mA. Is M4 consuming it? Every 1 ms, the ADC is converting with a timer interrupt. Is it it?

RE: nRF52 current consumption

Stian Røed Hafskjold — Thu, 02 Feb 2017 11:46:19 GMT

I guess the reason why it is not reduced when you connect is that the connection interval is higher than the advertising interval, and that you are transmitting more data when in connection than when advertising. What is the advertising interval? It is not like connection will always consume less current than advertising. It all depends on the BLE parameters you use, like BLE interval and payload. Different projects can have different parameters.

You can check out this power calculator to see the expected current consumption with different parameters: devzone.nordicsemi.com/.../

From what I can read from your plot you have a constant current draw of 20mA. This must be because of some external components on your board. You should figure out what is consuming so much current.

RE: nRF52 current consumption

sfujimoto — Thu, 02 Feb 2017 11:29:42 GMT

I will attach a screenshot.

RE: nRF52 current consumption

sfujimoto — Thu, 02 Feb 2017 11:15:35 GMT

thanks, The reference is ble_app_uart. I transmit 433 bytes per second. 20 bytes is 23 packets. It is custom hardware. However, I refer to PCA 10040. Besides NRF 52832, there are acceleration sensors and others. For other sample projects, the current consumption is reduced when connected. However, since this has not changed at all, I think that it is probable. I do not know the reason for the two valleys. It is also another sample project. However, it disappears when it connects. What is the cause?

RE: nRF52 current consumption

Stian Røed Hafskjold — Thu, 02 Feb 2017 09:21:19 GMT

What is the connection interval you are using? If this is a fast connection it might use a lot more than when advertising. It is also possible to advertise while in connection. I depends on what kind of code you are running. Is your code based on one of the SDK examples? If so, it would be helpful to know which one, then I can give you some numbers on the expected current consumption.

Also, it looks like something is consistently using a lot of current in your plot. If nothing else than BLE is running on the chip the current is supposed to drop down to about 2 µA between each BLE event.

Is this custom hardware, or are you using one of the DKs? It looks like something is wrong based on your plot. Can you explain more about the test setup you have? Do you know what the two drops in current in your plot is?