ESP32 Forum

I use the ESP32-WROOM-32 module on the D0WDQ6 chip. When sending data to the network via Wi-Fi via TCP at a speed of 10 kB/s (1 kb every 100 ms), I use the Power Save mode instead of PS_NONE to reduce power consumption.
In active mode (if PS_NONE), the current consumption is stable at about 144 mA. However, with PS_MIN and PS_MAX, an average high power consumption is observed at first, which has the ability to switch independently to the desired value of 20-40 mA after a while (I attach an image of the moment of arbitrary switching of the current level). However, there is also a chance that this automatic switching will not happen.
I am interested in the mechanism of automatic switching of the module between states with different current consumption. I would like to know how to disable this mechanism so that in PS_MIN or PS_MAX mode there is a stable low level of current consumption (any within 20-40 mA)?

Apparently, if/when/for how long the RF modem can go to sleep depends among other things on the amount of (broadcast) traffic in the network, which could be a reason for unpredictable (lack of) switches to/from modem sleep.

MicroController wrote: ↑

Wed Aug 02, 2023 10:42 pm

Apparently, if/when/for how long the RF modem can go to sleep depends among other things on the amount of (broadcast) traffic in the network, which could be a reason for unpredictable (lack of) switches to/from modem sleep.

I agree that the level of traffic congestion will affect the ability of the ESP 32 to go into mode with the Wi-Fi transmitter disabled.
But during the tests, I feel that the picture of the current consumed also depends on the moment the module is connected to the Wi-Fi network (rather, from the moment the data transfer begins). If you connect at some "right" moment, then automatic switching does not happen afterwards, or they are very rare even for a long time (during measurements lasting up to an hour at an initially low current value - it remains low, practically without switching).
I would like to understand what exactly is the "right" moment? And how would it be tracked using ESP32 in order to start data transmission at this moment?
Perhaps this is due to the fact at which specific points in time the beacon frames come from the router and how does this combine with the moments of the attempt to send data (now data is sent once every 100 ms)?