NRF52DK power consumption with demo UART

RE: NRF52DK power consumption with demo UART

Maxim — Thu, 21 Sep 2017 06:51:36 GMT

Finally I've found a way to reduce power consumption with this program, we must use uart_init(); only when we connect ot the NRF52DK in (BLE_GAP_EVT_CONNECTED) and uninit it when we disconnect (BLE_GAP_EVT_DISCONNECTED) use app_uart_close(). With this mode i get 30µA on advertise, 2.45mA when connected (uart_init()) seems to consume a lot of energy) then 30µA on disconnect and 2µA on_sleep.

RE: NRF52DK power consumption with demo UART

Bjørn Kvaale — Tue, 19 Sep 2017 10:16:28 GMT

It seems that the uart application requires more current consumption to run compared with the ble_app_hrs.

RE: NRF52DK power consumption with demo UART

Maxim — Tue, 19 Sep 2017 08:34:55 GMT

I have try uart demo code provided in the SDK 14 and the result is exactly the same (2.6mA advertise, 2.4mA connected and 2µA sleep). With the ble_app_hrs i get 600µA for advertising, 1µA during sleep and 500µA when connected. In this case i think my multimeter is not prescise enough to measure 112 and 28µA, so i consider your measure with the PPK are true.

RE: NRF52DK power consumption with demo UART

Bjørn Kvaale — Tue, 19 Sep 2017 06:09:51 GMT

I agree with you that the current consumption in this example seems quite high. I just ran the ble_app_hrs example (i.e. heart rate monitor example) in SDK 14.0.0 using the NRF52DK and got much lower current consumption. 112 uA for advertising, 1.49 uA during sleep and 28 uA when connected. So there could be something wrong with the demo code. You could try running the SDK 14 version uart code and seeing if you get lower current consumption.

RE: NRF52DK power consumption with demo UART

Maxim — Mon, 18 Sep 2017 16:35:06 GMT

Thanks a lot for your answer !!! My multimeter is still operational :)

But now i'have a really important question, why on the nRF52 Online Power Profiler avaible Here it's indicate 21µA with 3V and 500ms advertising and our measurements are 100x times higher (2.4mA >> 21µA) ??? Is the System_On (WFE idle) broken on this demo code ?

RE: NRF52DK power consumption with demo UART

Bjørn Kvaale — Mon, 18 Sep 2017 12:50:36 GMT

Hi Maxim. I have just run the code using the power profiling kit and an nrf52 dev kit. I get an average current consumption of 2.66 mA when advertising. During sleep, this value drops to a similar value that you measured. When connected, I get an average current consumption of 2.42 mA. So it seems your multimeter measurements are spot on. Looks like you don't need a PPK after all :) Although if you want graphs and more statistical data, it might be useful to get a PPK kit.

RE: NRF52DK power consumption with demo UART

Maxim — Mon, 18 Sep 2017 09:48:36 GMT

The code i am tring to run is ble_app_uart_pca10040_s132.hex provided in the nRF5_SDK_12.1.0_0d23e2a.zip file. And no i don't have the PPK yet but i 'll try to buy it soon.

RE: NRF52DK power consumption with demo UART

Bjørn Kvaale — Fri, 15 Sep 2017 10:27:50 GMT

I'm guessing you don't happen to have a Nordic Power Profiling Kit (PPK) at your disposal? Which code are you trying to run? Could you try to run the ble_app_hrs example under ble_peripheral from SDK 14 and see what your current consumption is then? I'll try the same thing using the PPK and we can compare results.

RE: NRF52DK power consumption with demo UART

Maxim — Thu, 14 Sep 2017 17:16:21 GMT

Ok, no problem, thanks anyway !

RE: NRF52DK power consumption with demo UART

Matt — Thu, 14 Sep 2017 16:54:45 GMT

I'm more of a firmware dude... you need to consult the Nordic current measurement experts here :(

RE: NRF52DK power consumption with demo UART

Maxim — Thu, 14 Sep 2017 15:23:40 GMT

I have already read these figures, but i think my problem is the prescision of my measurements. I think my problem is my multimeter or my calculations, i'have found this interresting tutorial: Calculating average current from Oscilloscope and Ampere meter measurements avaible Here.

The average current on adv mode seems to be the same i have measure by myself (2.48mA/2.85mA), so the real consumption is much lower according to these calculations (~13µA).

So do you think this issue is due to the low frequency sampling of my multimeter or my method of measurement.

RE: NRF52DK power consumption with demo UART

Matt — Thu, 14 Sep 2017 14:51:00 GMT

Can you cross check the figures in 23.15.2 Radio current consumption (transmitter) and in 23.15.3 Radio current consumption (receiver) in the "nRF52832 Product Specification v1.1"?