Solar powered standalone Arduino?

I've been searching for a way to build a sun-powered arduino, using a battery and solar cell. Seems like it'd be pretty straightforward to buy off-the-shelf solar cells, re-chargeable batteries, and power conditioning gear. Has anyone done the research already?

Main thing I'm looking for is the right power conditioning gear to charge the battery from the cell, and not over-charge the battery. Need to know which batteries would be best, size of cell, specific partnumbers for gear, and so on.... Also, looking to understand if the arduino needs any voltage regulation off of a battery whose output may decline through the night.

Anyone know of any such project I can use as a point-of-reference?

hi

there are hundreds of different kinds of batteries, and thousand of different charger circuits. You might want to start by fixing some variables, like

a) How much current does your electronics design require, in milliamperes per day? You might want to consider using the Atmega8L as it is a low power version. There’s also a nice high-efficiency DC-DC converter used in the Arduino BT design that you could copy.
b) How large of a solar cell can you tolerate in your design, and how much current, in milliamperes per day can you get from that size?

Once you have that, it’s just a matter of comparing efficiencies for different charger & battery systems.

D

Totally agree that there's a huge amount of choice out there. Kind of overwhelming really.

I was trying to see if there was - for the "average" off-the-shelf arduino board - some recommendation on what power supplies and solar cells would be appropriate. Isn't it the case though that for a basic arduino setup that includes - say - a sensor and a ZigBee hookup that there's some overall estimate on the voltage and amperage required?

Was hoping that someone had done these estimates already, and found some recommendations on specific model numbers and manufacturers for the "average" arduino use-case.

Thinking about it, I guess the "average" could vary a lot depending on whether or not you are doing something that is wattage intensive (e.g. controlling a heavy-duty servo.) For my use-case, want to basically just power a Squidbee with sunshine...indefinitely.

Do the squidbee people read this board? Seems very likely they've already looked at this issue and may just have an answer ready to go.

hey

it's still a design thing... you have to do the math :)

D

hey

it’s still a design thing… you have to do the math :slight_smile:

D

Yes, exactly. How do I calculate the draw in amperage and the voltage requirements of an arduino-based device in a given configuration? Any pointers on where to find such advice?

How do i calculate the amperage and voltage requirements for an arduino-based device that has (for example) a temperature sensor and zigbee trasmitter that wakes up every minute and reports back to home base?

Unfortunately, that usually requires spending altogether too much time getting cozy with the data books for the devices in question. Basically, you need to calculate the draw of each device and its duty cycle - which will give you the number of millamp hours for a day's use. After you add a safety factor, and allow for the inefficiencies of the power supply, the solar panel the battery, and the charge controller, you should have a good idea how much power the system must take in each day to continue operating. The solar panel should be sized appropriately.

Use the maximum possible power consumption for the device and battery charging for reasonable solar energy reception. That would cover you for a project though it’s expensive if you wanted to make several thousand of them. In that case you’d either use your own generated data or try and figure it out from the datasheets and how the circuit works. Hrm… an Arduino to monitor the current of an an Arduino or other device and remotely report back via one of the radio devices. Busy Busy :slight_smile:

This link
http://www.techlib.com/reference/batteries.html
Seems to imply that mAh/ma=hours so mAh=hours*ma so 12 hours at 200ma neads 2400mAh just to run the Arduino at maximum for that time. There are also issues with the voltage dropping that you may need to address depending on your circuit.

For batteries, none of them are real easy though lead acid is not too difficult.

hey

it’s still a design thing… you have to do the math :slight_smile:

D

Yes, exactly. How do I calculate the draw in amperage and the voltage requirements of an arduino-based device in a given configuration? Any pointers on where to find such advice?

How do i calculate the amperage and voltage requirements for an arduino-based device that has (for example) a temperature sensor and zigbee trasmitter that wakes up every minute and reports back to home base?

It would varry on what your doing with the arduno why not look at the charge pump and solar panels from spark fun the charge pump is on the simon board they have in a tutorial. Their is eagle cad schematics for it and you can use that to re roll the older arduno eagle cad schematics. I don't know why they don't give the eagle schematics for the new arduno NG maby it is not realy a open design. 8-) ;D Also just make some aplication circuit and mesure with a multimetter the current and voltage comming from a battery through a resistor going to the battery terminal.

Hej,

Libelium people are most likely not reading this thread, why don't you post something specific to them or try to send them an email? Their addresses are on their website

/d

Even when you calculate the power consumption from the data sheets you still need to measure the actual circuits on the bench (and maybe even under the field).

A while back I built a small data collector that was to be deployed in a remote location. I did the calculations, built the circuit and measured it on the bench and ended up measuring a current demand about 50% higher than calculated. Further when the application was deployed, I discovered the current demand was higher yet further the small solar panel produced less than expected.

Moral:Nothing can compete with test, measure, and repeat!

Oooh, wow! This is really helpful advice. I didn't check "notify of replies" for this topic and thought discussion had stopped. Thanks everyone for the advice!

In the meantime, I've reached out to a bunch of people in the solar power community and read a text-book on home solar power. Kind of interesting. I'm going to check out the Spark fun project and see what that yields. The solar power people I've spoken with have offered some useful advice. Namely:

1) High capacity batteries (in terms of Ah) usually come in just 6V, 12V and 24V. 2) Might be easier and safer to stick with the "stock" equipment that's built around 12V systems 3) You can buy a "buck regulator" that basically steps the voltage down from 12V to something arduino can handle

The solution using stock equipment for home solar power could just be: A> Calculate the amperage consumption via bench measurement (thanks for the suggestion) B> Choose a 12V battery with enough Ah capacity

Constraint B1: Leave margin here since over-deep charging can shorten battery life Constraint B2: Leave even more margin for short and/or cloudy days C> Use text book tables to find how much sun on average comes your way. D> Find the right solar cell size Constraint D1: Has to output enough Ah to keep the battery topped off Constraint D2: Solar-cell charge rate has to be within about 10%+/- of battery capacity E> Hook the solar cell to a charge controller. Hook the charge controller to the battery. Hook the battery to the buck regulator. Hook the buck regulator to the arduino. F> Wrap everything up in whether proof casing G> Mount it with the solar cell pointed (basically) due-south with an inclination that matches your lattitude.

Bottom line here is that its quite possible to do this with stock equipment. Problem is, that it sure ain't cheap. :-/

Sorry if this is long-winded. Just trying to give back the benefit of the research I've been doing.

Thanks everyone for your help so far!

Hej,

the main problem in this whole stream of information is that you haven't considered Arduino's basic power consumption. Even if you put the processor to sleep, the power regulator in the board is constantly demanding some mAmps. Besides that you don't need another regulator, the board can be powered with 12V

/d

You have a point. It'd be a good idea to figure out a way to power down and power up the arduino as needed throughout the day instead of running it full time.

I could be wrong, but it looked to me like some buck voltage regulators can reduce voltage in a way that would reduce the overall Amp-hours you'd draw from the supplying battery. There are inevitable losses of power, but I think its possible you may (with the right regulator) extend the life of the supply battery by supplying arduino with less than 12V.

Having said that, it'd still be a good idea to find clever ways to power the arduino up and down as needed instead of running it full time. Maybe a low-power clock circuit, or maybe just run it during the day.

It would also be useful to know the insolation value for the region you are in.Insolation is an estimation of how much solar radiation you can expect to get over the course of a year in your part of the world.

There are some places where solar is just not even worth bothering with because the amount of usable sunlight they get is so low.

This is a neat idea.

I’d probably do it the brute force approach: Buy the biggest cell avaliable depending on cost to power ratio and your usage.
Then try it out. :slight_smile:

Remember that its ok to get a cell which is more powerful than what you need.
Getting one which is too small however isnt desirable.