It is solved. Supplying it through battery is OK, supplying it from +5V Vusb pin causes 4mA consumption.
I am already working on my application.
Good for you.
The on-board battery charge controller consumes a lot of current when powered from the 5V pin or USB connector.
Looks like, yes, still, canât fathom how that chip managed to consume 4mA. Yet here we are, Thanks for idea!
According to the datasheet, the current consumption of the charge controller BQ25100 is 0.75mA.
There may be other causes for the 4 mA.
edit
When the battery is not connected, the charging terminal falls below 4.2V, so the BQ25100 attempts to charge the battery until it reaches 4.2V. The operating current at this time would be about 4 mA.
1 Like
Heyy guys
facing an issue observe 5 spikes every 100ms from 2uA to 23uA unable to understand whats the issue
i tried multiple ways , those didnt worked i cimmented them
here is following snipet
//----------------------------------------------------------------------------------------------
// BSP : Seeed nRF52 Borads 1.1.1
// Board : Seeed nRF52 Boards / Seeed XIAO nRF52840 Sense
// important : need for Serial.print() <Arduino.h> <Adafruit_TinyUSB.h>
// \Arduino15\packages\Seeeduino\hardware\nrf52\1.1.1\libraries\Bluefruit52Lib\src
//----------------------------------------------------------------------------------------------
// 2023/07/07
// Central sketch : nRF52_XIAO_LightSleepAdvScan_central.ino
// Need to be deleted /Documents/Arduino/libraries/SdFat
#include <nrf_rtc.h>
#include <Arduino.h>
//#include <Adafruit_TinyUSB.h> // For Serial.print()
#include <bluefruit.h> // \Arduino15\packages\Seeeduino\hardware\nrf52\1.1.1\libraries\Bluefruit52Lib\src
#include <Adafruit_SPIFlash.h> // Need to be deleted /Documents/Arduino/libraries/SdFat
#define SLEEP_TIME 30 // Sleep time [sec]
//#define SLEEP_TIME 30 // Sleep time [sec]
Adafruit_FlashTransport_QSPI flashTransport;
//uint8_t timeout = 5;
void setup() {
// Serial port initialization
//Serial.begin(115200);
//while ((!Serial) && (timeout)){
// delay(1000);
// timeoutâ;
// Enable DC-DC converter
NRF_POWER->DCDCEN = 1; // Enable DC/DC converter for REG1 stage
// Sleep QSPI flash
flashTransport.begin();
flashTransport.runCommand(0xB9); // enter deep power-down mode
flashTransport.end();
//NRF_UART0->ENABLE = 0; // Disable UART
//NRF_SPI0->ENABLE = 0; // Disable SPI
//NRF_TWI0->ENABLE = 0; // Disable I2C
//Bluefruit.begin();
//Bluefruit.Advertising.start();
//Bluefruit.Advertising.stop();
//Bluefruit.end();
//RTC initialization
initRTC(32768 * SLEEP_TIME); // SLEEP_MTIME [sec]
SysTick->CTRL &= ~SysTick_CTRL_ENABLE_Msk;
// Initialization of Bruefruit class
// Bluefruit.begin();
//Bluefruit.setTxPower(4); // Check bluefruit.h:151 for supported values
//Bluefruit.autoConnLed(false); // turn off LED to save power Sleep Current 4uA
// Bluefruit.setConnLedInterval(50); // Should not be used, Sleep current 300uA
suspendLoop();
// Serial.println(âEnd of initializationâ);
}
//*************************************************************************************************************
void loop() {
// Bluefruit.Advertising.stop();
//delay(5000);
//digitalWrite(LED_BLUE, LOW); // turn the LED on (HIGH is the voltage level)
//delay(5000); //Led on // light for 5 seconds, but the rtc in the background keeps running. â light off only for 25 seconds
//digitalWrite(LED_BLUE, HIGH);
//enterLowPowerState();
// delay(SLEEP_TIME * 1000);
// Sleep
__WFI();
__SEV();
__WFI();
// System ON Sleep by FreeRTOS
}
//**************************************************************************************************************
// RTC initialization
void initRTC(unsigned long count30)
{
// See â6.22 RTC - Real-time counter 6.22.10 Registersâ
NRF_CLOCK->TASKS_LFCLKSTOP = 1;
NRF_CLOCK->LFCLKSRC = 1; // select LFXO
NRF_CLOCK->TASKS_LFCLKSTART = 1;
while(NRF_CLOCK->LFCLKSTAT != 0x10001)
NRF_RTC2->TASKS_STOP = 1; // stop counter
NRF_RTC2->TASKS_CLEAR = 1; // clear counter
NRF_RTC2->PRESCALER = 0; // set counter prescaler, fout=32768/(1+PRESCALER) 32768Hz
NRF_RTC2->CC[0] = count30; // set value for TRC compare register 0
NRF_RTC2->INTENSET = 0x10000; // enable interrupt CC[0] compare match event
NRF_RTC2->EVTENCLR = 0x10000; // clear counter when CC[0] event
NVIC_SetPriority(RTC2_IRQn,3); // Higher priority than SoftDeviceAPI
NVIC_EnableIRQ(RTC2_IRQn); // enable interrupt
NRF_RTC2->TASKS_START = 1; // start Timer
}
//**************************************************************************************************************
// RTC interrupt handler
extern âCâ void RTC2_IRQHandler(void)
{
if ((NRF_RTC2->EVENTS_COMPARE[0] != 0) && ((NRF_RTC2->INTENSET & 0x10000) != 0)) {
NRF_RTC2->EVENTS_COMPARE[0] = 0; // clear compare register 0 event
NRF_RTC2->TASKS_CLEAR = 1; // clear counter
//intFlag = true;
//digitalWrite(LED_RED, LOW);
//delayMicroseconds(10000); // for LED
//digitalWrite(LED_RED, HIGH);
}
}
following is ouput observed on PPK
@PJ_Glasso @iggzzzz @daCoder @hobbya @1melc @msfujino
I think the DC/DC converter may be working.
Please use </> when you post your code.
For example, if you want XIAO to wake up every 30 seconds, you should write delay(30000); instead of using RTC, which will reduce the sleep current.
1 Like
Hi there,
As @msfujino indicates use the delay or the Sleep is the Zyphur using the RTC in System on sleep.
My .02 after scanning some old threads, in anticipation of the Xiao that uses battery like a âCamel in the Desert.â 
When it Drops, Ninjaâs will be getâtin âSturdyâ 
Xiao4Ever 
GL 
PJ 
1 Like
Hi there, I am a bit late to the party but I am at my wits end because my measurements look nothing like the things I see here.
As a demo I used this sketch:
because @msfujino provided his measurements.
My measurements look like this 100k samples/sec, 3.6V to battery pads:
Every 30ms i get heavy spikes. These spikes are about 400uA while in light sleep (delay) and up to 1mA during LED on. While spikes of a few dozen uA seem to be normal, these are waaay of the chart.
Even in deep sleep (adding NRF_POWER->SYSTEMOFF = 1; to the end of setup) the 400uA spikes are there but only every 200ms
While the average of the sleep period is about 4uA (matching msfujinoÂŽs value) the âLED onâ period averages at about 320uA (about double of msfujinoÂŽs value).
My original setup was using PlatformIO but in my desperation of narrowing this down i also tried ArduinoIDE with Board 1.1.10
The only thing that helps a bit is providing 3.3v directly to the 3.3v Pin instead of using the battery pads. This reduces the light-sleep spikes to more normal 40uA. But the 1mA active-spikes are still there:
I also tried:
- Another Xiao nrf52840
- Using the PPK2 in ampere meter mode
- Changing Breadboard and wires
- Double checking my PPK2 against other devices
Has anyone an idea how to bring these spikes down?
Hi there,
SO there is a Ton of Nuance to Sleep we all have learned, even the BSP, the Process , where you connect what , ALL of itâŠ
Keep digging youâll get there. The Rtos does some stuff too. Check this thread I think youâll find a similar post about the spikes and what the user did to mitigate them. YMMV 
Read all of the Power ninjaâs 
posts too @msfujino. heâs got the most receipts and testing under his cap.
HTH
GL PJ
ThePradox,
Let us know your ArduinoIDE sketch and BSP version and connections.
We will try to measure it.
EDIT
Current consumption varies depending on the LEDs to be lit. My data was for green LED.
Red:720uA, Green:120uA, Blue:320uA.
1 Like
ArduinoIDE: 2.3.6, BSP: 1.1.10, nothing attached just PPK2 3.6V to the Battery Pads
Using this sketch:
#include <Arduino.h>
#include <Adafruit_TinyUSB.h> // For Serial.print()
#include <Adafruit_SPIFlash.h> // Need to be deleted /Documents/Arduino/libraries/SdFat
#define SLEEP_TIME 30 // Sleep time [sec]
Adafruit_FlashTransport_QSPI flashTransport;
void setup() {
delay(2000);
// Enable DC-DC converter
NRF_POWER->DCDCEN = 1; // Enable DC/DC converter for REG1 stage
// Sleep QSPI flash
flashTransport.begin();
flashTransport.runCommand(0xB9); // enter deep power-down mode
flashTransport.end();
//NRF_POWER->SYSTEMOFF = 1;
}
//*************************************************************************************************************
void loop() {
digitalWrite(LED_BLUE, LOW); // turn the LED on (HIGH is the voltage level)
delay(5000); // light for 5 seconds, but the rtc in the background keeps running. --> light off only for 25 seconds
digitalWrite(LED_BLUE, HIGH);
delay(SLEEP_TIME * 1000); // System ON Sleep by FreeRTOS
}
Do you get the same 400 uA spikes during light-sleep?
A 300uA spike is visible with a period of 22mS.
Thank you, so this is normal. I was just very irritated because the screenshots in this thread never showed this
This has been a very helpful thread, and Iâd like to thank everyone for their contributions. I am using the Xiao nRF52840 in an anemometer to record mean wind-speed and gusts, sending the data back to a hub using LoRa. It will be battery powered and mounted on my roof, so I am keen to get low power consumption for long battery life.
I am using PlatformIO in VSCode with the Arduino framework. I had trouble getting the MCU to go in to POWER ON sleep until I found a suggestion by daCoder to use semaphores, which worked very well. Using unmodified libraries, these are the current consumption figures I have measured with the board powered through the battery terminals (I am using 3 lithium AA batteries, so max. voltage is about 5V):
- unconfigured board 13.5 ”A
- enable DC-DC convertor 13.5 ”A (no change)
- power down flash 4.1 ”A (- 9.4 ”A)
- initialise low frequency clock 4.1 ”A (no change)
- attach external interrupt 11.1 ”A (+ 7.0 ”A)
- enable voltage divider 13.9 ”A (+ 2.8 ”A)
PlatformIO ini
[env:seeed_xiao_nrf52840]
platform = https://github.com/Seeed-Studio/platform-seeedboards.git
board = seeed-xiao-afruitnrf52-nrf52840
framework = arduino
lib_deps =
adafruit/Adafruit SPIFlash@^5.1.1
monitor_speed = 115200
upload_speed = 115200
main.zip (1.7 KB)
1 Like
Caution is required when connecting batteries other than LiPo batteries rated at 3.7V (max 4.2V) to the battery pad. If USB is connected, the battery will be charged at 4.2V.
1 Like
Yes, I avoid having the USB and battery both connected at the same time. I have a Schottky diode in series to protect the battery pack from the charge circuit just in case. Iâll remove the diode when complete to get a bit more of the battery capacity as the voltage drop is about 300 mV.
Yes, I have seen that but it also needs a LoRaWAN gateway, further adding to the cost. Also, it is a bit of an eyesore to stick on my chimney! I have a more discrete anemometer for the chimney and the rest goes in the garden. I already have a few sensors and donât really need LoRaWAN. I am using an ESP32 with SX1262 as a hub and uploading data to a weather service via wifi.
1 Like