- Sat Aug 26, 2017 12:57 pm
#69457
Hello all,
I built a very simple temperature sensor around an ESP8266 and now I am trying to power it with a solar panel. The panel is small, around 2W, and is unable to power the unit directly (through a step-down buck regulator) in anything but Sahara-like conditions.
So I'm trying to get the solar panel to charge up a (super)capacitor, and power the ESP8266 with that.
The challenge is that as the capacitor charges, the voltage will rise fairly slowly from 0V up to whatever voltage I am getting from the panel. Not relevant right now.
The ESP8266 does not like to see its VCC/CH_PD rise "slowly". To put it differently, it does not turn on even when the voltage reaches 3.3V. It sits there in some undefined state, sometimes with the red led dimmed, sometimes with no led light at all.
The unit works fine if I give it a clean 3.3V from my lab power.
SO. What to do. I tried the SparkFun energy harvester based on the LTC3588. The good thing about this unit is its hysteresis which kicks in around 5.5V and turns the load off around 3V (I think) - just what I need. Unfortunately this unit is not able to supply enough current to boot up the ESP8266 - the output voltage on the energy harvester drops drastically as the ESP8266 draws too much current, leaving the unit back in an undefined state.
I have also tried putting a 3.3V step-up regulator (buck) on the capacitor and powering the ESP8266 with the output from that regulator. This does not seem to work for reasons that are not known. The ESP8266 does not turn on. If I quickly disconnect the CH_PD and reconnect it, the unit does power on. I have no explanations for this.
I guess what I need is some kind of low voltage disconnect board, with a high hysteresis: Load on at 5-10V, load off at 3V. Run that though a step-down buck with a 3.3V output and I should be set.
But noone seems to make boards like that, which is surprising.
What can I do?
Just to preempt some suggestions:
1) "Just use a lipo cell and a regular solar charger (pwm/mppt)".
Too complex, too expensive. And I don't want batteries involved because they wear out sooner or later. Also I would have to have an LVD anyway, to avoid over-discharging the cell.
2) "So hook up your panel to a battery with a cheap buck step-down and charger, and the load to the battery."
I don't want to have the load connected to the battery because I have to implement some sort of LVD to avoid over discharge. Batteries suck. Must use the solar panel directly or a super capacitor.
3) "Use a couple of regular batteries"
I guess I could do that, but my generic ESP8266 does not support deep sleep without some serious ninja scalpel-wielding techniques. On 2 AA cells the unit dies after 2 days. Batteries not feasible.
I built a very simple temperature sensor around an ESP8266 and now I am trying to power it with a solar panel. The panel is small, around 2W, and is unable to power the unit directly (through a step-down buck regulator) in anything but Sahara-like conditions.
So I'm trying to get the solar panel to charge up a (super)capacitor, and power the ESP8266 with that.
The challenge is that as the capacitor charges, the voltage will rise fairly slowly from 0V up to whatever voltage I am getting from the panel. Not relevant right now.
The ESP8266 does not like to see its VCC/CH_PD rise "slowly". To put it differently, it does not turn on even when the voltage reaches 3.3V. It sits there in some undefined state, sometimes with the red led dimmed, sometimes with no led light at all.
The unit works fine if I give it a clean 3.3V from my lab power.
SO. What to do. I tried the SparkFun energy harvester based on the LTC3588. The good thing about this unit is its hysteresis which kicks in around 5.5V and turns the load off around 3V (I think) - just what I need. Unfortunately this unit is not able to supply enough current to boot up the ESP8266 - the output voltage on the energy harvester drops drastically as the ESP8266 draws too much current, leaving the unit back in an undefined state.
I have also tried putting a 3.3V step-up regulator (buck) on the capacitor and powering the ESP8266 with the output from that regulator. This does not seem to work for reasons that are not known. The ESP8266 does not turn on. If I quickly disconnect the CH_PD and reconnect it, the unit does power on. I have no explanations for this.
I guess what I need is some kind of low voltage disconnect board, with a high hysteresis: Load on at 5-10V, load off at 3V. Run that though a step-down buck with a 3.3V output and I should be set.
But noone seems to make boards like that, which is surprising.
What can I do?
Just to preempt some suggestions:
1) "Just use a lipo cell and a regular solar charger (pwm/mppt)".
Too complex, too expensive. And I don't want batteries involved because they wear out sooner or later. Also I would have to have an LVD anyway, to avoid over-discharging the cell.
2) "So hook up your panel to a battery with a cheap buck step-down and charger, and the load to the battery."
I don't want to have the load connected to the battery because I have to implement some sort of LVD to avoid over discharge. Batteries suck. Must use the solar panel directly or a super capacitor.
3) "Use a couple of regular batteries"
I guess I could do that, but my generic ESP8266 does not support deep sleep without some serious ninja scalpel-wielding techniques. On 2 AA cells the unit dies after 2 days. Batteries not feasible.
