Solar Panels and Dark Activated Motors

[John Mossbacher]

Hello. I'm working on a project and could use some help. The end goal is to design a platform that responds to the solar panels surrounding it. The panel will be on motorized wheels and should move in the opposite direction of where the least light is being received (or in the direction the most light is being received). I'd also like for the device to power itself and be able to store energy when it is not using it. I have a basic idea on how to do those things so my question is: How can I determine how much power I need to generate by the panels to move certain object weights? I'd like to be able to generate enough power to move between 10-50 lbs across a flat, but mildly rugged terrain.

Any suggestions would be greatly appreciated.

 

[BradtheRad]

Your wheels must have a large diameter, if they are to get over mildly rugged terrain.

To store electrical power, a lead-acid battery is probably your best option. However this increases your weight. Requiring more powerful drive motors.

Your drive system is not too different from a motorized wheelchair. You may pick up some tips if you look into the methods for constructing them.

Although the total weight may be less than a human being, you need stable construction nevertheless. A chance breeze against the solar panels could topple the unit.

 

[John Mossbacher]

Thank you for the response. I'm now thinking the correct order of operation would be to build the unit first to determine how much power it needs, then match it with the correct proportion solar panels. Ultimately, I'm looking for something that is sits 4" above the ground, between 1.5'x1.5' or 1.5'x4.0'. It will be bottom heavy so tipping is not a primary obstacle as of now.

The other big challenge for me is to figure out which sensors to use to tell it when to stop moving (so it doesn't run off a cliff or continuously drive the wheels into a wall) I think I'd like to use laser sensors over physical switch sensors.

 

[chuckey]

I do not understand why you are doing what you are proposing. The position of the sun is well known, both in height and its bearing, so a look up table for where you are and a tracking unit would be enough. Mount your panels on a fixed pole that you turn around ( 180 degrees in less then 24 hours), so the pole takes the static weight. Mount the panels on a hinge at their centre (so their static weight is balanced) and move them according to the elevation of the sun. Using this system it reduces the power consumption of the tracking system to a few percent of what you are proposing.
Frank

 

[John Mossbacher]

The project I am working on is to design a robot that will stay out of the shade (or stay in the shade, depending on the preference). For this reason, I need the design to be mobile.

 

[Audioguru]

The lousy old 741 opamp will not work from a supply less than maybe 10V because it was designed to use ONLY a 30V supply.
Its inputs do not work properly if their voltage is within 3V from the positive or negative supply. The output does not go low enough to turn off your transistor and does not go high enough to properly turn on your transistor.
Use a modern Cmos rail-to-rail opamp instead.

 

[betwixt]

You also need at least one more sensor to be able to tell which direction to move in. With a single LDR/comparator you can only assess the light level, not where it comes from. In a practical system you will probably need several more and some decision logic to guide the motor toward or away from the light source.

Brian.

 

[John Mossbacher]

I agree. Right now, I'm brainstorming whether I should hook a sensor to each panel to monitor the amount of voltage each puts out to tell the correct motor(s) to move in the direction where the most photons are being absorbed by the panels -or- try and come up with a more customized photoresistor system where "no light" means go. The diagram I posted above was a concept that certainly needs expanding on but if I had one of those systems on each motor, it might work. I also need to figure out how to have these motors communicate with each other, like you suggested. This is my first real project and I'm just figuring it out as I learn so every piece of new information I get is another piece to the puzzle. Any more suggestions on parts I might need? I know I need transistors, photoresistors, motors, solar panels, barrier sensors, maybe a dynamo or a battery, inverters, wires and a breadbasket of course...

But what else?
Thank you

 

[betwixt]

a breadbasket!! - it runs on electricity not dough :lol:

You need to consider your exact needs before looking to buy parts. The first thing to consider is how is it powered, obviously, if it has the ability to move into the shade, you can't use solar power directly or it would never be able to move again afterwards so you need an alternative power source or enough PV power to run the unit and charge a battery at the same time.

I would suggest the next step would be to carefully work out an algorithm to achieve your needs, try flow charting the whole operation and do it in the finest detail. Not just "if dark move forward" steps but a complete chart to show which motors should be powered, which direction and at what speed. When you have the flow chart, the next step is to look at the possible usage scenarios, for example, what happens if it is totally dark or there are two light sources or if this is to be used outdoors, what if it's foggy but bright and all the sensors see the same light level? Factor these into a new flow chart.

When you are satisfied that all eventualities are covered, start looking at the mechanical demands on the unit, the size of motors and their power requirements, how much energy you can get from the PV and how big the battery has to be. You might be surprised at the power needs of motors and how little energy you can produce from light.
You will then need something to make it follow the paths in your flow chart, you can do it without a microcontroller or save yourself hundreds of hours by using one. A micro makes the control functions VERY much easier to implement. Also check things like the optics, make sure you can sense the light direction accurately enough, bear in mind that a PV panel has almost 180 degrees acceptance so they aren't much good for direction finding.

Only at this stage should you start looking for parts, there is absolutely no point in buying parts you might need then finding them unsuitable. Do the design first then buy the parts later.

Brian.

 

[John Mossbacher]

I meant to say breadboard, haha. Autocorrect.. I suppose it would be infinitely wiser to put everything down on paper first. I already filled up my dry erase board and came to the conclusion that I'm in the information gathering stage still. This forum has done wonders for me already. I'll be getting back to you all with some more advanced questions once I get to that stage.

Thank you for the help.

 

[chuckey]

Now that you have this automated robot strolling around, the first thing it has to do is to find some light, or if there is none, go to sleep -literally. So a start would be to have four mini solar cells facing at 90 degrees to each other and a detector system to decide if there is enough light about to try and tune into it, or go back to sleep for an hour or until there is a worthwhile amount of light about.
Once your four mini solar cells have given you the " some light about" signal, you can compare their outputs to get a rough direction and rotate your robot to this direction. Then you need to get into the focussing mode which means four more photo cell/LDRs, but this time orientated in the same direction but each looking up through a separate tube. the tubes being splayed slightly. This is so when the tube array is actually pointing at the sun, the output from each photo cell/LDR is the same. If the tubes are not facing the sun properly, a pair of the detectors will give a higher output then the other pair.
This way you can use the Left and Right pair to rotate the robot, and the up/down pair to get the azimuth correct.
Frank

 

[John Mossbacher]

I like that idea. This way, the robot can utilize its' energy more efficiently. I was thinking about having it react whenever a panel stopped receiving as much light as the others. Another obstacle I'll face is having a small shadow (like the silhouette of a branch or its' leaves) interfering with the system, effectively "pushing" it away or into a corner. Maybe I can have the sensors take a reading every 20 minutes or so to avoid that problem. So much to think about..

Thank you for your suggestion about the LDRs in a tube. That is a good innovation.

 

[John Mossbacher]

I like that idea. This way, the robot can utilize its' energy more efficiently. I was thinking about having it react whenever a panel stopped receiving as much light as the others. Another obstacle I'll face is having a small shadow (like the silhouette of a branch or its' leaves) interfering with the system, effectively "pushing" it away or into a corner. Maybe I can have the sensors take a reading every 20 minutes or so to avoid that problem. So much to think about..

Thank you for your suggestion about the LDRs in a tube. That is a good innovation.

Still working on the project. All good things take time for sure.