Idle current consumption

RE: Idle current consumption

Sigurd — Sun, 03 Dec 2017 20:59:11 GMT

Yes, it could. See this post.

Let me know if it does not help to cut SB41.

RE: Idle current consumption

mich_x — Sun, 03 Dec 2017 20:32:20 GMT

‌‌ ‌‌ ‌‌ ‌‌ ‌‌ ‌‌ ‌‌

int main(void)
{
    timer_init();
    power_init();
    buttons_leds_init();
    db_discovery_init();
    ble_stack_init();
    gatt_init();
    nus_c_init();
	app_uart_close();
    *(volatile uint32_t *)0x40002FFC = 0;
    *(volatile uint32_t *)0x40002FFC;
    *(volatile uint32_t *)0x40002FFC = 1;
    for (;;)
    {
		nrf_pwr_mgmt_run();
    }
}

This is my main. One problem is that I only cut SB40 but not SB41. Would that cause the current a little bigger?

RE: Idle current consumption

Sigurd — Sun, 03 Dec 2017 20:30:08 GMT

When I tested the example(without logging, uart, and scanning) I got somewhere betwen 2-4 uA.

How does your main() function look like now ?

RE: Idle current consumption

mich_x — Sun, 03 Dec 2017 19:17:56 GMT

The average current is about 40uA. I want it to be smaller than this.

RE: Idle current consumption

Sigurd — Sun, 03 Dec 2017 13:31:53 GMT

This "wave" is probably the on-chip power regulator working in auto-controlled refresh mode in order to maximize power efficiency. It's completely normal, and lowers the average power consumption. See the quote in this post.

When you measured, what was the average current consumption ?

RE: Idle current consumption

mich_x — Sat, 02 Dec 2017 21:07:37 GMT

I have stopped the uart and the LOG and the scan. I measured again and got a periodic wave. The highest current is 70uA and the lowest is 40uA. And the cycle time is about 50ms. Do you maybe have any idea of what causes this wave? I have commented out everything else. This is my whole program.

   for (;;)
{
	nrf_pwr_mgmt_run();
}

RE: Idle current consumption

mich_x — Fri, 01 Dec 2017 18:37:16 GMT

Really? SB40 and SB41? I only cut SB40 but already got a relatively clean figure.

RE: Idle current consumption

Sigurd — Fri, 01 Dec 2017 11:34:23 GMT

Hi,

If you close the UART(E) with app_uart_close(), and stop the scanning(comment the scan_start();, or call sd_ble_gap_scan_stop()) you should see a couple of µA power consumption.

Note that you on the nRF52840-PDK need to cut both SB40 and SB41 to get accurate current measurements.

Note, you need to power cycle the UARTE peripheral after you call app_uart_close() in order stop HFCLK and DMA bus. See this post

So you need to do it like this when you close the UARTE:

err_code  = app_uart_close();
*(volatile uint32_t *)0x40002FFC = 0;
*(volatile uint32_t *)0x40002FFC;
*(volatile uint32_t *)0x40002FFC = 1;

Also remember to disable the log module in sdk_config.h (set NRF_LOG_BACKEND_RTT_ENABLED and NRF_LOG_ENABLED to 0)

RE: Idle current consumption

mich_x — Thu, 30 Nov 2017 19:18:11 GMT

I am not sure. Basically, I am using the ble_app_uart_c example. And the idle current I measured using power profiler is 0.9mA. Is it somehow too big?

RE: Idle current consumption

Sigurd — Thu, 30 Nov 2017 11:43:17 GMT

Hi,

Are you measuring the System ON, No RAM retention, Wake on any event current ? For the nRF52832 this is 1.2 µA. For the nRF52840 you should see a lower number here.

What did you measure ?