I made a weather station to continuously monitor daylight and temperature. Since the project is placed away from the house, it is more convenient to use a battery to supply power, rather than a long extension cord.
During the day, the weather station should run from a solar panel as much as possible to avoid consuming the battery. And, rather than a person occasionally checking on the battery levels and replacing the batteries, the circuit should recharge the battery pack from excess solar energy from the solar panel.
This article describes the solar and battery backup circuit, along with the recharging results so far.
Schematic of a simple solar-panel based circuit.
A solar panel is on the left side of the schematic. A diode (D1) protects the target device against a negative voltage in case the solar panel is installed in reverse. The diode is a Schottky type (such as 1N5817) so that very little voltage is lost going through it.
The device itself contains the usual power supply circuitry, such as capacitors to stabilize the power source and a voltage regulator to establish a voltage level usable by all of the chips and other components.
The problem with running a device from a solar panel is that it won’t receive power at night. And, in fact, unless the device has a fairly large capacitor, it may turn off during the day when a cloud passes over the solar panel.
To keep the device running, a battery can be attached in parallel as a backup power source.
Schematic of a simple solar-panel based circuit with battery backup.
Diode D2 serves the same basic purpose as D1, in that the device won’t be harmed if the battery is installed backwards. However, the diode prevents the battery from receiving power from the solar panel because a diode is a one-way valve.
A solar circuit that can’t recharge the battery pack is beneficial if you want to install standard non-rechargeable batteries as the device’s backup power source. For example, you could use this circuit with alkaline AA cells. Alkalines are cheaper than rechargeable batteries and last longer for a single use.
When the solar panel receives enough light, the device runs entirely from the solar panel. Let’s say the solar panel provides 5 V, less 0.1 V for the diode drop -- meaning TP3 will read 4.9 V. If the alkaline battery pack consists of three fresh AA batteries, it may be as high as 1.6 V × 3 = 4.8 V. Because the battery pack’s 4.8 V is less than the solar panel’s 4.9 V, power from the battery pack will not flow. Current only flows from a higher voltage to a lower voltage. Therefore, no alkaline battery power is used.
When the solar panel and the battery pack are at the same voltage, they will both contribute to running the device. This is a neat trick for extending the life of a device with an otherwise underpowered solar panel. Every little bit helps take some burden off the batteries.
Finally, when the solar panel isn’t receiving enough light, its voltage drops below the battery pack’s voltage. In that case, the battery supplies all of the voltage through D2.
In all cases, there is no device-perceived transition in switching between the power sources. The device receives power all of the time.
Assuming the solar panel has excess capacity at various times (either because the target device is not consuming its peak power or because the solar panel is receiving extra light), it would be nice to store the extra power. To do so, we simply add one more diode (D3) and a path from the solar source to the rechargeable source.
Schematic of a simple solar panel charger circuit.
As before, the solar panel provides power to the device through D1, or the battery provides power through D2, depending on which power source has a higher voltage. When the solar panel has the higher voltage, solar power flows through D1 to power the device and through D3 to recharge the battery pack.
Why add diode D3 instead of just removing D2 to allow the battery to get recharged? Well, that would remove the reversed-battery protection.
Why not add D3 but then remove the D1 path completely? Well, then the solar panel power would have to go through two diode drops to reach the device, which would require a solar panel with a slightly higher voltage than before.
The extra diode costs less than 25 cents in quantity 100. Given the versatility of the 1N5817 diode (we use them all the time in motor drivers), a group buy of quantity 1000 will bring the price down to a dime each.
The cost of a diode is a lot less expensive than replacing a device due to a reversed-battery mistake or purchasing a more powerful solar panel due to a double diode voltage drop.
Most consumer battery chargers are “smart chargers” that supply constant-voltage or constant-current in a manner that the particular battery chemistry prefers. These chargers also shut off recharging at a certain voltage, temperature, amount of time, or when it detects the cell changing consumption.
The simple solar recharger circuit doesn’t have any safeguards against overcharging, nor does it care about optimal battery longevity. To get away with such a simple circuit, you must pick a solar panel whose operating voltage is around the desired battery voltage and whose total power output doesn’t exceed the maximum trickle charging rate.
For NiMH cells, the maximum charging voltage is 1.6 V and the maximum current rate is 0.05 C for up to 20 hours. For a three pack, that would be 4.8 V (3 × 1.6 V) and 125 mA (2500 mAh × 0.05 C = 125 mA). I’m not worried about maximum number of charging hours, because sunlight isn’t going to exceed 20 hours in Chicago.
The solar panel used for my weather station device is a Panasonic Sunceram II BP-378234 which has a maximum (open-circuit) voltage of 5.5 V and a maximum (short-circuit) current of 43 mA. At first this would seem to exceed the maximum charging voltage. However, the real-world measured voltage during charging never exceeds 4.25 V because the voltage of a solar panel drops significantly when attached to a load. In fact, the official operating voltage is only 3.4 V, which is slightly less than the desired operating voltage of 3.6 V.
In summary, a simple direct solar charging circuit works safely only if the solar panel is significantly less powerful than the battery. That ensures the solar power is never too much to overcharge the battery and thus doesn’t need to be electronically monitored to disconnect when the battery is full.
Frankly, I wouldn’t use this circuit with lithium, any fickle chemistry, or any expensive battery. NiMH is probably the least expensive and most compatible, although lead-acid may also work. You can always add a current-limiting regulator if you’re not sure if your solar panel is too powerful or not.
Okay, let’s see some graphs of how the solar charger performs in a real life...