I am currently in a scientific research class at my school, and I've decided to build a self-cleaning solar panel that detects debris using a change in weight that triggers a windshield-wiper mechanism to clean the surface. I am extremely new to the world of circuits and engineering, and I would appreciate any ideas you may have on executing this project. The system would run on energy collected from a solar panel and be activated when the weight of the panel reaches a certain threshold. I am restricted to a student research budget of about $100, and I already have small solar cells that I will be using to test this circuit.

I have done research on load cell circuits and the Wheatstone Bridge concept, but I am confused on how to use the weight collected to activate a separate mechanism and which materials would be best to use. Any ideas and suggestions are welcome, and feel free to ask any clarifying questions! Thank you!

 

I think weight will be a very difficult measure to use as a proxy for "dirt on the panel". You'd have to measure milligrams per square foot to be able to detect dust, and at that level of precision there will be problems with moisture absorption and other interfering noises. It might make more sense to just run the wiper at preset intervals.

Or, you might be able to devise a dual beam sensor. One beam is somehow kept (or made to be) clean and the other sees the same debris as the panel. If the exposed beam loses power compared to the clean one, you know the panel is probably dirty also.

Just thinking out loud.

 

I have to strongly agree Wayneh in that using weight will be difficult something as simple as a slight breeze would have more effect likely than the weight of dust or dirt on the panel. I would think about embedding a tiny photo cell under the glass and sense the light taking into consideration cloudy days and night darkness. This is where I like the dual sensor idea mentioned. The delta is what you are looking at, the trick is making sure one sensor is always clean. I just do not see using weight working out very well.

Ron

 

A wash and brush up every night would be sufficient ?

 

A wash and brush up every night would be sufficient ?

Oh I agree but schools love to mete out research projects and it does make the students think. Every night at sundown swish and swab.

Ron

 

I would go with some version of wayneh's idea of a light beam (perhaps IR) reflected off the glass at an angle and sensed at the other side.
I think just one beam would be needed.
When the reflected intensity drops a certain amount from the baseline clean reflection, you would know it needs cleaning.
You could automatically calibrate the baseline clean intensity reflection right after its cleaned.
I would think the reflected intensity would be fairly sensitive to dust and debris on the surface.

 

Well, it would work for snow.

 

The weight sensing idea sucks- but don't take my word for it - Time for some Scientific Research!

Take a dirty piece of glass about the size of the average solar panel, scrape the dirt off with a razor blade and weigh it.
Compare the weight of the dirt to the weight of the average solar panel.

This gives you an idea of how vanishingly small the "signal" is you need to detect.
Here is a data sheet- it's a 350 watt panel, weighs 20 Kg.
Lets say you get 100Mg of dirt, that means you are looking for (20,000 / 0.1 ) = 1 part in 200,000.

Now compare this with the resolution of a fancy expensive digital scale- not even close.

 

I was thinking of using the design more to detect the weight of heavier debris, such as wet leaves, acorns, or snow that accumulates on panels that are difficult to reach (such as on roofs). However, the measurement of intensity of a light beam may work in both cases (both layers of dirt and larger pieces of debris).
Do you have any suggestions for components that can measure intensity of a reflected light?

 

I would think measuring the output from each of the cells could give you an indication of non-uniformly distributed debris (like leaves). Uniform dust would be hard to measure that way so maybe needs a different approach. Maybe a dust sensor. Have an exposed piece of glass that you put a beam through. Degradation of the beam indicates dust or dirt.

 

You may want to start with a Google of terms like "reflective light sensor" or similar. Here is an example of an IR Reflective Sensor just as an example. You would embed a mirror in the panel and when it becomes dirty the reflected light will diminish. Years ago I worked with a dew point analyzer. They used a mirror with a reflective light sensor. The mirror was chilled and when the mirror began to fog over the air was at dew point and the temperature noted. Th link is an IR emitter and receiver but you get the general idea. Just allow for the convergence point distance.

Ron

 

Years ago I worked with a dew point analyzer. They used a mirror with a reflective light sensor. The mirror was chilled and when the mirror began to fog over the air was at dew point and the temperature noted.

Excellent approach to reverse engineer for this project.

 

And if the sensor itself gets dirty ?

 

And if the sensor itself gets dirty ?

You clean it.
But that shouldn't happen often since the emitter and sensor would be facing downward.
And if it's auto-calibrated after every automatic cleaning of the panel, the emitter/sensor would need to get quite dirty before it wouldn't work.
The auto-calibrate cycle could also tell you if the signal is getting low enough that an emitter/sensor cleaning is needed.