CC1310 WOR and Sniffer Mode Code Example

TI-RTOS containing WOR examples are now possible to download from software-dl.ti.com/.../index.html

Hi,

I could measure current consumption for active RX phase, but I wonder how big is current consumption between RX ON states and average volume for rfWakeOnRadioRx sample application?

I can not use my DMM because of small time between RX activity (just 500 ms).

Here is CC1310 rfWakeOnRadioRx consumption measured on 10 Ohm resistor in power line (XDS100V3 disconnected)

Is there any information with regards to estimated current usage for different WOR wakeups per second?

I'd just like to get an idea on how much lower than the standard 5.5mA receive current? 

Hi

The chip enters stand-by in between wake-ups so the the average current consumption between wake-ups when using the sniff command is less than 0.6 uA

Siri

Hi, Siri,

Am I understand right - average current for this sample application is 52.472 uA and I should use this figure to calculate battery life time? Of course we need to add real RX activity - data receiving and faulse activity due to interference.

What is measurement settings for above screenshot (measurement device type and method)?

BR,

Oleg

Hi Oleg

The plot was meant to show the average current consumption when the chip is in standby. This average current consumption is 542 nA in my case.

In the test I did, the radio enters RX every 500 ms (it is in RX just before cursor 1 and just after cursor 2). The peak in the middle of the plot is the recharge current in standby (please see http://www.ti.com/lit/swra486) for more details.

BR

Siri

Hi Scot

The average current consumption (when taking RX into account) was in this case below 16 uA.

I've bumped the CPU clock back up to 48MHz since it doesn't seem to affect average power much.

The last time I measured with a meter, the RX lasted only one sample (less than ~68ms). Out of 200k samples collected, 15k were at the higher current, the rest were lower than 10uA. Attached is a plot of my latest testing, where I reduced symbol rate to 2.4kBaud, and deviation to 1kHz which seemed to help peak power - the frequency of high current events has also been reduced since I'm now doing a WOR wakeup every 5 seconds, but the time it spends high is the same.

I've attached a slightly zoomed graph, the peak only shows up as one sample so I'm also including a table that shows one of the peak values. To measure the current, I disconnect all jumpers on the Launchpad, then connect the GND jumper and the USB cable. I am using a Keithley DMM7510 to then make a current measurement by connecting the hi-probe to the 3.3V on the XDS side and the lo-probe to the 3.3V on the CC1310 side. In this case the current consumption is without receiving a packet, only the periodic wake up; however, the peak current doesn't change even if I am transmitting packets.

I changed ccfg.c to make SET_CCFG_MODE_CONF_DCDC_RECHARGE 0x01, then turned on local optimizations with -mf 2 in the project properties. The reported average current is now 20uA, probably slightly higher than default (<16uA) because of antenna mismatch at 915MHz, but much better than my initial results.

Hi TER,

Can you explain, why reducing the CPU clock does not affects the power consumption ?

Thanks TER.

To make more clear,
CC1310 Active Mode current consumption is 1.2 mA + 25.5 uA/MHz.

@ 48 MHz = 2.5 mA
@ 24 MHz = 1.9 mA

If we have 2 mS(@ 48 MHz) task for every 1 second period. The average current consumption is;

i_avg = i_active_mode x (t_active_mode/period)
i_avg_48MHz = (2.5/1000) x ((2/1000)/1) = 5 uA
i_avg_24MHz = (1.9/1000) x ((4/1000)/1) = 7.6 uA

So it seems, faster clock speed is more energy efficient as you told.
Thanks.