Power management differences between SDK 15 and SDK 11

RE: Power management differences between SDK 15 and SDK 11

GeneralSpontz — Tue, 14 May 2019 14:32:38 GMT

Ok, thanks, that's what I thought.

RE: Power management differences between SDK 15 and SDK 11

Simonr — Tue, 14 May 2019 07:06:01 GMT

Hi Spontz

After some digging it seems you are right. You have to make your own loop for deinitializing and reinitializing your peripherals when using SDK15. I have not been able to find an equivalent function to the power_manage function in SDK11.

Best regards,

Simon

RE: Power management differences between SDK 15 and SDK 11

GeneralSpontz — Mon, 13 May 2019 16:17:35 GMT

Hi Simon,

I start with the UART example (BLE_peripheral /BLE UART) because I needed basic data communication on BLE. I use SPI to communicate with a MEMS sensor on my board.

In the SDK11, I just called PowerManage() and it took care of putting the SPI peripheral into sleep.

In the SDK15, the idle_state_handle() function does not do that. I have to explicitly call de-init the SPI before sleep, then call init again everytime I wake up. I wake up on a timer event every 50ms.

I use the legacy SPI driver. (identical in SDK11 and 15)

So my question is related to SDK15:

- is what I do (de-init / re-init) the right way to manage a peripheral power when using sleep/wake ?

- is there a higher level function that can desactivate all peripherals before entering sleep? It seems to exist in SDK11.

Regards

Spontz

RE: Power management differences between SDK 15 and SDK 11

Simonr — Mon, 13 May 2019 07:08:40 GMT

Hi Spontz

1. Why would the SPI be initialized at all in the UART example? The idle state handle should do the same as the power_manage function after handling any pending operations. So I am guessing the extra current draw is for the application to finish up any pending operations before going to sleep.

2. The pins not used in the SEGGER example, should all be disconnected, so yes, you will have to initialize the other IOs. Struggling a bit to understand what is the problem in this question.

How high of a current draw are you experiencing compared to the SDK11 example? It shouldn't be too hard to achieve the same results if you are using the same kind of chip.

Best regards,

Simon