During BLE operation, can it suddenly sustain 20mA?

[pkr2258]

HI

A phenomenon occurs where the current remains at 20mA intermittently.

Actually, the BLE specification states that RF TX/RX is less than 10mA....

I'd like to know if there are any other factors that could cause a BLE SOC to consume 20mA.

I'm trying to narrow down whether it's a problem with the BLE SOC or with another circuit.

Thank you

[Susheel Nuguru]

Hi,

Can you please be a bit more descriptive on what exactly you are testing and what is the exact moment where you noticed 20mA current? Do you have the power profiler snapshots? 

The power of Radio in TX/RX is the power of that peripheral only. You can write an application that can keep CPU, other power hungry peripherals like UART/SPI and RADIO on at the same time. If your application is in debugger mode, then that can add to the power consumption. So in short, yes it is possible that the SoC can consume more power based on how many peripherals your application kept active at the time of measurement.

[pkr2258]

Hi,

Susheel Nuguru said:

Can you please be a bit more descriptive on what exactly you are testing and what is the exact moment where you noticed 20mA current? Do you have the power profiler snapshots? 

* The test configuration maintains one peripheral and one central unit connected to each other.

This occurs when a button is pressed on the central unit, transmitting data to the periphral unit.

In some cases, current exceeding 20 mA occurs, rarely exceeding 1%, making analysis extremely difficult.

.

* Current consumption is tested by connecting the periphral to a current meter.

  This current meter was used for verification.

.

Susheel Nuguru said:

The power of Radio in TX/RX is the power of that peripheral only. You can write an application that can keep CPU, other power hungry peripherals like UART/SPI and RADIO on at the same time. If your application is in debugger mode, then that can add to the power consumption. So in short, yes it is possible that the SoC can consume more power based on how many peripherals your application kept active at the time of measurement.

.

* The SOC I use primarily functions as a BLE radio, CPU, and I2C.

I don't use any other power-hungry UART or SPI.

.

* To reiterate the situation where the current consumption exceeds 20mA and the device does not enter Sleep:

1) A key input occurs in the Central and data is transmitted to the periphral.

2) The periphral consumes (maintains) more than 20mA of current in standby mode.

3) If no external buttons are pressed on the periphral, ​​the current consumption remains above 20mA.

-> Pressing an external button returns the device to normal operation.

.

Is there any part that I should review? 

[Susheel Nuguru]

The TX for Radio at 0dBm should take around 7mA but if the CPU is not sleeping at all with I2C active and if the SoC is in debug mode, then it is possible that there are windows where the whole SoC consumes a bit more. If you are saying that the chip is staying longer times in 20mA.

If this is the average power you see, please help me reproduce this on the nRF52832 DK so that I can debug this for you.

If this is not the average current and you see this spikes at data transfer, then you need to use Power Profiler  using the documentation and record some measurements which you can send to me. I can then do a closer analysis on where and why these power spikes happen.

[Turbo J]

Too much for either CPU core or radio, and even both have trouble reaching that together. Softdevice is typically very good at turning the radio off after use.

You might be looking for a shorted output pin - that would roughly fit a 20mA current consumtion scenario.