We used ble_adv project on homemade minimal system board (SDK v1.0.6.968) to investigate low energy function.
we have settings as below:
pm_set_sleep_mode(pm_mode_extended_sleep);
advertisement interval is default 687.5ms.
During the test, we have two readings which are significant with official documents:
1, active current is 2ma from start to 8s.
2, current drop down to 600 ua after 8s later.
My question is:
Is 2ma current normal for reading #1?
As noted in AN-B-037 DA1468x Power measurements_v1.3, the avg. current for 1000ms interval advertisement is about 20.3713 ua, but my reading #2 is near 600ua. What reason might be in your opinion? (we are using a minimal system, no other peripheral attached.)
Device:
Hi auzzz,
For the first 8 seconds the device is active all the time (no sleeping between advertising intervals) since that 8 seconds is the maximum time for the XTAL32 to stabilize, so a quite larger power cosnumption is expected.
Regarding your 2nd measurement, i can verify approximatelly 20uA on my board with the ble_adv project with modified advertising interval in 1000ms and around 35uA in 687ms advertising interval with the Smart Snippets tool. Regarding your measurement and 600uA that you see i am not able to replicate that value, are you certain that the interval used is 1000ms, i can see approximatelly 600uA when the device is advertising every 20 to 30 ms. Other than that this offset might be due to some current leakage from the board that you are using. I dont think there is anykind of sw configuration that can cause this amount of current consumption when sleep is enabled.
Thanks MT_dialog
Hi MT_dialog,
Leakage current confirmed.
Thank you!
我们使用一个I2C外围4.7K pull-up resistor, which caused leakage current in sleep mode.
Do I have to reconfigure I/O setting to floating before entering sleep mode? If so, in which part should I modify in BLE_ADV project ?
Thanks
Hi auzzz,
When going to sleep the pins are latched to their current configuration (the configuration that you had before go into sleep mode) and when you wake up you reconfigure them in the periph_init() function (the function is used as a parameter in the pm_system_init() function). So what is the configuration on that specific pins that are used for the I2C interface ?
Thanks MT_dialog
We didn't modify SW settings
/* Init hardware */
pm_system_init(NULL);
默认的洞穴ings is pull-down according to document while I2C peripheral we used requires pull-up which makes a loop. After configuring GPIO pin function, leakage current disappeared.
Thanks for your support and guidance.
I configure timer0 and timer2 in PWM mode.
Must I configure timers in the periph_init()?
Hi Ruslan,
It depends when you would like to activate them, if you would like the timers to run on each wake up, yes you can configure them in that function, since the periph_init() is invoked in each wake up of the device.
Thanks MT_dialog