CCS/CC2640R2F: Reducing Power Consumption

Hi Christian,

With IDLE, I assume you mean Standby. The current consumption for the IC in Standby is 1 uA. A measured current consumption of 50 uA is indicative of current leakage somewhere in the design, for example through external pull-ups or -downs with incorrectly configured IOs. 

What is your HW setup and what SW are you running?

Have you read through this application note: Measuring CC13xx and CC26xx current consumption (Rev. D)?

Regards,
Fredrik

Hi Fredrik,

Thank you for the link. 

My application is quite basic. I have one analog input for an sensor and one digitial output for a LED. The main CPU is most of the time in standby. The sensor controller has an scheduled event each 1s, where it increments a variable. In addition it has an analog comparator event for the external analog sensor, which activates the main CPU for some calculations. The sensor triggers the device 2 times a day. So device is most of the time in standby. 

I was doing the current measurement according to the link and it looks like the picture below. The 100ms pulses are quite higher (~1mA) in comparison to the description. I guess because of the sensor controller. So in average there is a 50uA consumption. 

Christian

Hi Christian,

How are you measuring the current consumption? Can you verify that you see 1 uA with the PinStandby example?

Regards,
Fredrik

Hi Fredrik,

I am using this device (pic, 1mV=1uA), which I have connected to my scope.  

For the software, I use a modified version of the BLEstack tutorial code. Measuring my code and the original BLEstack code show the same current consumption (pic2).

Christian

Ah, Dave Jones :-)

I am not sure you are able to capture the Standby recharge spikes properly. Did you try to measure the pinStandby example?

Hi,

I did the pinStandby example on my device and measured the current. I was also doing the same measurement on the LAUNCHXL-2640R2 board. Both show the same result. There is not a lot of consumption, only each 400ms there is a short 1.4mA peak.

Christian

Hi Christian,

That looks as expected. The average should come in around 1 uA.

This means that the 100 ms pulses you see are from either the SensorController or the main MCU waking up. You need to check your code to see why that happens more often than expected.

Regards,
Fredrik

Hi Fredrik,

It is not coming from the Sensor Controller. It shows the same result, when I remove the code of the sensor controller. It comes from the main CPU. 

I use a modified version of the ProjectZero code,because I need the Bluetooth interface. I guess I should have it mentioned before, that I am also using the BLE interface. Reading through the upper link, it says the main CPU is advertising each 100ms, so there should be peak.

Christian

Hi,

Is there a way to change the advertising time from 100ms to 200ms? Would reduce the energy consumption a bit.

Thank you

Christian

Hi Fredrik,

Yeah, I was thinking that is clear that I am using BLE. Learned something new!!!

I was already figuring out, how to change the advertisement interval. With that I was able to reduce the total consumption a bit.

Are there any other things I can do for further reduction? If not, I guess we can close this thread!

Thank you

Christian

Hi Fredrik,

Yes I was playing with the radio and I guess I have found a good compromise. 

Thank you for your help. 

Christian