Solar power to trickle charge a robot's batteries.
Martin_H
Posts: 4,051
I have an idea for a demonstration project idea that I could use a bit of help with because parts of it are a bit above my EE skills.
The project idea is to create a robot that can explore its environment for a longer amount of time than battery power normally allows. To achieve this goal I would use a solar panel as an adjunct power source which is large enough to power a BS2, but not the motors. The idea is that when the robot is moving, or in shadow, the batteries will provide the power. But when sitting still in bright light, the solar panel will power the BS2 and trickle charging the batteries. The robot's task would involve some sitting still in the sun to obtain power, with some movement based upon outside events and programming. The robot does need to survive harsh condition either, I'd take it inside if it started raining.
I plan on using either my Boe-bot or CBA robot with a five or six cell NiMh pack. So I believe I need a solar panel with a higher voltage than the batteries and a check diode to prevent the panel from discharging the batteries when the robot is in shadow. There are plenty of inexpensive 9 volt solar panels (e.g. http://www.parallax.com/product/750-00031) that I think should have enough oomph to power the BS2, but I need assistance with the interface to the batteries.
For example if I'm using a 9 volt solar panel that provides much less current than the batteries, do I need a charge controller between the solar panel and the batteries? If I simply connect the solar panel and check diode to Vcc and Vss, what would happen? What voltage will the CR servos receive? The nine volts from the solar panel or the six volts from the battery pack?
The project idea is to create a robot that can explore its environment for a longer amount of time than battery power normally allows. To achieve this goal I would use a solar panel as an adjunct power source which is large enough to power a BS2, but not the motors. The idea is that when the robot is moving, or in shadow, the batteries will provide the power. But when sitting still in bright light, the solar panel will power the BS2 and trickle charging the batteries. The robot's task would involve some sitting still in the sun to obtain power, with some movement based upon outside events and programming. The robot does need to survive harsh condition either, I'd take it inside if it started raining.
I plan on using either my Boe-bot or CBA robot with a five or six cell NiMh pack. So I believe I need a solar panel with a higher voltage than the batteries and a check diode to prevent the panel from discharging the batteries when the robot is in shadow. There are plenty of inexpensive 9 volt solar panels (e.g. http://www.parallax.com/product/750-00031) that I think should have enough oomph to power the BS2, but I need assistance with the interface to the batteries.
For example if I'm using a 9 volt solar panel that provides much less current than the batteries, do I need a charge controller between the solar panel and the batteries? If I simply connect the solar panel and check diode to Vcc and Vss, what would happen? What voltage will the CR servos receive? The nine volts from the solar panel or the six volts from the battery pack?
Comments
First of all, the charge rate of the cells is at something like 1/10th the rated amp hours. So you are only going to get so much activity out of one sun day.... a full charge may take 10 hours.
And, there is the issue of the size and weight of the solar panel that will actually provide the maximum rate of charge. You may suddenly realize that you have a rather large and bulky robot due to the solar panel size.
And if you get past those two hurdles, there is the actual care and feeding of NiMH cells as the don't like to be overcharged, they don't like excessive discharge, and the should stay out of the sun and heat for long life.
Personally, I gave up on the idea of a mobile solar robot and gave up on the idea of using the NiMH and decided that the Lead Acid Gel cells with trickle charging from solar cells was a lot easier to engineer. But that doesn't mean that your goals are impossible.
These days, we have buck and boost switching regulators that can take a good solar charging scheme and tailor the voltage outputs to whatever you motors and microcontroller require. You may find that you can get by with cells NiMH cells and have them fully charged in one day's sunlight. It is only when you desire too much from the batteries that you begin to find you needing too many solar panels.
You might also consider using cheaper and less charging sensitive NiCad for solar panels. NiMH may not stand up well to trickle charges.
The sunlight hitting the panel is a huge variable, cloudy vs clear, angle of view (whether the cell is perpendicular to the sun). So you rarely have consistent output from the cell. Use a Schottky blocking diode to reduce the drop to 0.3V (typical rectifier diodes are ~0.7V) and shoot for high no-load voltage out of the solar cell, I'm swagging maybe 12V no-load output in full sunlight to charge a 6V battery pack. The higher the voltage difference, the more hours of charging you will get, from lower light levels throughout the day. You will rarely have optimal conditions with maximum solar panel output voltage, clear sky bright sun at 12 o'clock, so try to get as much as you can at other times.
The batteries set the voltage in the system, not the solar cell. They are like a giant filter cap. If you hook a 12V no-load voltage solar cell in full sunlight to a 6V battery pack, the total voltage will be something like 6.1 V, so your servos have nothing to fear.
As you said, you will be getting at best a low-current trickle charge from a reasonable size solar cell, so IMO there is little risk of overcharging the batteries even without a charge controller. Someone more knowledgeable than I will challenge that statement, but I look at the number of solar garden lights out there which are nothing more than a solar cell, a diode, a NiMH cell for power, plus an LED and a comparator circuit to switch the LED on when the solar cell output falls to zero or thereabouts.
Even when NiMH cells get overcharged via trickle charging (like many cheap cordless phones), they heat up a bit and dissipate the extra energy that way. No big deal, battery packs still last several years in constant use.
Stick your battery pack, solar panel, diode and two multimeters out in the sun all day and check the voltage & current every hour to see what happens, you're probably good, especially for a light-use robot experiment. If you were going to leave this thing unattended outdoors 27/7/365, you might need a charge controller. That could cost more than your batteries and solar cell combined, so name your poison.
Of course if you wanted to get slick, your bot could make & break the connection to the solar panel with a latching relay (you did stock up when I posted about those from Electronic Goldmine, right?). That way, you eliminate the diode (full voltage, no 0.3V drop) for faster charging, and the relay only needs an occasional microscopic blip of current to toggle it on or off. Leave the cell connected and charging mostly, but occasionally break the connection and use an ADC to measure battery voltage vs. solar panel voltage and act accordingly.
My interests were to have a solar panel on a mobile outdoor bot, solar panel mounted so a servo could tilt it on one horizontal axis. By rotating the robot (via drive wheels) and tilting the panel, it could always face the sun for maximum power output.
We seem to be talking about a robot here. A machine that can run around and do stuff. It has to carry all it's batteries and solar panels with it. That's all heavy stuff and moving it will eat energy. SLA's might be out of the picture due to their weight.
We know this is possible in some way. The Mars rovers have been roving for ages. They have the advantage that they are never out of the sun half the time. They can just stop and recharge, perhaps doing some low energy expenditure experiments, when they are ready the can move on a few meters more with the solar energy acquired.
But here on Eath, then what?
Energy input is really unreliable. You may be in the shade any way. Then you have to remember where the best sun bathing spots were and run back to them when you are tired.
There is a reason why vegetables don't move around much here on Earth:)
Anyway, you might find David L. Jones interesting on battery management: http://www.youtube.com/watch?v=A6mKd5_-abk
I didn't buy the latching relay but now I wish I did. This is a really neat idea because you can have a robot act as its own charge controller.
Is it this heavy?>> http://www.dhgate.com/store/product/novelty-mini-solar-power-robot-insect-bug/135923679.html
Or is it this heavy?>> NASA GROVER
The Grover is like what erco describes, except bigger.
Smaller than a bread box, but larger than a BEAM robot like the solar bug.
I revisited Wikipedia to verify that I was providing good opinions about NiMh.
Actually, it seems that I got some things wrong.
1. NiMh will handle trickle charges from C/10 down to C/300 quite well.
2. The power density of NiMh is significantly superior to NiCd or Lead Acid -- so it is well worth making this your prefered choice for a solar powered robot.
My other concern is about a low voltage shut off. This is necessary to protect the NiMh from damage from excessive discharge. I have previously mentioned several times that there are voltage regulators that have an additional Enable feature that can be used to shut down before damage occurs to your batteries. This would be a wise option.
Regarding high voltage protection, I would rather use blocking diodes to get a 9V solar cell to properly mate with a 7.2VDC battery pack than waste 30% of the solar power with a linear regulator. Careful matching of solar cell, blocking diodes for voltage drop, and the right number of cells should provide a lot more power availble to charge the cells and drive the robot.
Agreed. In addition if it is not practical to match the solar cell to the battery a switching regulator can be used to increase or decrease the voltage from the solar cell to match the battery voltage. Not quite as good as an MPPT charger, but a big improvement.