Low Power and Solar Power

Since it is evident a 9V battery won’t last near long enough I decided to start playing with solar power to charge a NiMH 9V battery.

Doing some research I found someone that was using solar cells from garden lights to charge a 9V battery. I went to Harbor Freight and got a NiMH 9V battery and 8 of these lights for $20:


I parted the lights and got some switches, LEDs, and 1.2v AA batteries as well:


I connected 4 of the solar cells together and built a charging circuit:


This was kind of clumsy to deal with so I connected all of the solar cells onto a piece of cardboard. 4 cells didn’t seem to generate quite enough voltage so I tied in a 5th. The other 3 are available, but unused:


I only had a few hours of sun, but I did get the battery charged enough for an Uno to run for about an hour.

Next I will program the Uno to sleep all except for a few events a day and see if the solar cells can keep the battery going.

The next experiment can be found here.

This entry was posted in c-arduino, Low Power Arduino. Bookmark the permalink.

6 Responses to Low Power and Solar Power

  1. Viper Joe says:

    Hey Big Dan, those of us Cyber Stalking you from unknown origins would really like to see a pic of the back side of the carboard with all the solar panels. We want to see how you have wired up all the panels to make one big one. Thanks…

  2. Dan TheMan says:

    phht, there’s no cyber in your stalking – I had to move 300 miles to get away from you🙂

    They are simply connected together in series – red to black, red to black. Right now only 5 are connected which is providing about 10V in hazy conditions. I believe each cell maxes out at 2.5V, so on a sunny day I should see 12V.

  3. Sean Straw says:

    I have several comments and questions on this one.

    Disclaimer: I’m merely a software geek with a hardware bent, though I work for a hardware manufacturer in the grid-tied solar power industry.

    Is the intent really to drive an _UNO_ or might it be to eventually drive something like an ATTiny or similar?

    Sleep modes on the AVR processors (ATMega, such as the ATMega328 used in the Uno, or ATTiny, such as ATTiny84, 85, 24V, etc that I use in various projects) are capable of dropping the uC power consumption to |- BATT+

    In a normal array series, you’d have a suitably rated schottky diode between EACH module, but because you’re using these tiny modules, the voltage drop across the diodes would quickly drop your overall system voltage, so just use the one to limit current flow towards your battery.

    Note also that when reading the voltage from a solar module without a load connected, this is a “no load” reading. The voltage when there is a load connected will be much lower.

    Lastly, I believe boost circuits are worthy of consideration for driving small projects. In your application, it would allow you to use a smaller supply (battery or solar) to drive something which requires a small current at a higher voltage.

    I developed a voltage booster derived from a Joule Thief (but using a Zener diode in reverse bias instead of an LED) that provides me with 5.5V for projects. This circuit optionally charges a 5.5V ultracap (which functions similar to a battery but without the memory effect), and supplies my uC circuit with power – from a 2V solar cell similar in size to a single one of the cells from your Harbor Freight solar lamps (FTR, WalMart is another source of cheap solar stakes – I glommed onto a bunch for $0.97 apiece a while back for projects, though those had the stubby 2/3AA cells). You can use a boost circuit to charge a higher voltage battery as well. I scavenge Li-Ion cellphone and laptop batteries from battery recycling at the office – often the batteries just don’t have the maximum capacity they once had, but still function well for little electronic projects (which aren’t as power hungry as a cellphone). The boost charge from a small solar cell won’t exceed their charging current, so I’m not concerned about overcharging them.

    Another advantage of a boost circuit is that they’re considerably more lax when it comes to the input voltage than a direct battery circuit – I can drive my boost module from one or two healthy 1.5V batteries, or I can run it off of a single 1.5V battery which has been run down to 0.75V (or less, but the oscillator tends not to start below 0.7V input) — I still get 5.5V on the output. It’s all a function of how much current you need to pull out of the system and how constant that current needs to be (the ultracap helps here). If you were driving a 5V circuit off of 6V, and the batteries dropped off to x, shut it off, and bring it back on when the cap is < y), which would help to conserve energy if the supply is a battery, since we don't need to continue to run the boost if we have suitable output already. The uC of course would make use low power sleep mode and itself only be powered on when the boost module is on (I have power switches on the modules)

    Now, I aught to get off to work. Missed my standup (status meeting) this morning.

    • Dan TheMan says:

      Thanks for the comments Sean.

      I dropped this project last winter mainly to wait for better wx (not much sun here in the winter to work with). So far I’ve been busy with other things and not returned to it.

      I did buy a decent solar panel from harbor freight (well, compared to what I was using). But it is probably overkill for my eventual idea.

      My plan is to build a gate alarm for my front gate. While it has a spring to keep it closed, I sometimes prop it open and forget about it. Having built a nice little monitor for my back gate this would be slam dunk except for the fact that there is no electricity; hence my interest in battery/solar.

      When I finally get around to doing this, it will be based on something like the RBBS where there are no LEDs or other non-essential circuitry.

      When I get back around to this I will make use of your suggestions. I especially need to do some experimenting with boost circuits.

      p.s. – I’m glad I don’t have to do stand-ups any longer!

  4. Sean Straw says:

    I designed the boost circuit so that I could power ATTiny based circuits from solar: charge a battery or cap from a solar cell, and have the ATTiny do it’s thing – take some environmental measurements, store them, go to sleep, and after so many readings, power up a radio transmitter module (found cheap on aliexpress – it’s all made in China, may as well order it direct), to transmit the data to a base station (a matching radio receiver module hanging off of a Beagle Bone Black), where it can be stored in SQL and queried, generate email notices that can be sent out on the internet (= basic smartphone notification without needing a special app), etc.

    The same approach could be used for monitoring the state of a gate and signalling some device (perhaps another ATTiny, but with a buzzer or lightshow hanging off of it instead of a sensor). I ran some basic range testing with a tiny circuit sending a pulse on the transmitter, and a similar circuit flashing an LED when it received the pulse. Set down the transmitter, and start walking away from it while holding the receiver unit and note how far you get before it misbehaves. In my case, > 350 feet or so. This was without any sort of ECC (Error Correcting Code), whereby a corrupted signal could potentially be corrected. Other radio modules have even greater range.

    Come to think of it, your gate project could broadcast the current state of the gate say every 5 minutes, or using an interrupt on a make-break (physical, or reed) switch when the gate is latched and unlatched. The interrupt method has the benefit of being able to effectively provide a well-timed log of the open and close times of the gate. Logging of times could be managed at the receiving end, so the gate sensor logic needn’t worry about having an RTC. An “i’m still alive” transmission could be sent every ‘x’ hours, so you know a transmitter hasn’t gone dead.

    You can store data into the EEPROM of an AVR processor – so you could have multiple gate transmitter devices, and each could have the same sketch loaded to it, but a different serial number stored in EEPROM., which can be part of it’s report (and used for defining the retransmit period, etc). This means you could have _ONE_ receiver unit. listening in and storing information for a variety of sensors on your property.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s