- Tue Jan 13, 2015 4:11 pm
#7225
You are right, this is important point to keep in mind. My plans are to use non-rechargable batteries, here is my list of preffered ones:
- CR123A
- lithium AA variants (15-20 years shelf life, > 2500mAh)
- SAFT batteries (if I convince myself about the price)
- and why not "regular" alkaline - like energizer max, with 10 years shelf life?
Rechargable batteries that have low enough self discharge are rare. Currently I don't see a good use case of them.
Another thing to keep in mind is the filtering capacitor - electrolitic (Al) might get higher leakage than the ESP idle current (at some extreme temperatures).
But I logged in to give some results from my tests of the current consumption. My setup uses 5V/3.3V switching DC/DC supply (chinese one with MP1584 chip) to feed the ESP module from Olimex.
In line with the 3.3V I put parallel group of one resistor (1K) and one shottky diode. The purpose of the 1K is to allow me to measure the ESP idle consumption, it gives me easy way to read out values down to 1uA with generic multimeter. The purpose of the shottky diode is to bypass large current spikes that appear while ESP is active.
This seems to work fine for deep sleep measurement, but it is not suitable for active currents ( > 1mA, 70-150mA typical).
The results: first trials were made with 1000uF capacitor on the supply pins of the ESP (after the measurement group). This way I got 150-160mV readings that corresponds to 150-160uA. I think that slow ramp down of the supply voltage tricks the ESP to behave bad and leads to this high readings.
Update: too good to be true ... when I disconnected the usb-uart converter my consumption climbed to 40uA, steady. Not happy.
.
- CR123A
- lithium AA variants (15-20 years shelf life, > 2500mAh)
- SAFT batteries (if I convince myself about the price)
- and why not "regular" alkaline - like energizer max, with 10 years shelf life?
Rechargable batteries that have low enough self discharge are rare. Currently I don't see a good use case of them.
Another thing to keep in mind is the filtering capacitor - electrolitic (Al) might get higher leakage than the ESP idle current (at some extreme temperatures).
But I logged in to give some results from my tests of the current consumption. My setup uses 5V/3.3V switching DC/DC supply (chinese one with MP1584 chip) to feed the ESP module from Olimex.
In line with the 3.3V I put parallel group of one resistor (1K) and one shottky diode. The purpose of the 1K is to allow me to measure the ESP idle consumption, it gives me easy way to read out values down to 1uA with generic multimeter. The purpose of the shottky diode is to bypass large current spikes that appear while ESP is active.
This seems to work fine for deep sleep measurement, but it is not suitable for active currents ( > 1mA, 70-150mA typical).
The results: first trials were made with 1000uF capacitor on the supply pins of the ESP (after the measurement group). This way I got 150-160mV readings that corresponds to 150-160uA. I think that slow ramp down of the supply voltage tricks the ESP to behave bad and leads to this high readings.
Update: too good to be true ... when I disconnected the usb-uart converter my consumption climbed to 40uA, steady. Not happy.
.
