Here is what I posted as a comment on Wayne's (HelpMonkey's) **broken link removed**
Wayne, I see that you are touting the circuit I posted for you on www.electro-tech-online.com as your circuit #4, with no attribution, and with an implication that you designed it. I don't mind if you market it as a kit, but it would be nice to if you had left my name on the schematic diagram, and acknowledge where you got it.
btw- This circuit does not work on the same principle as your CDS circuit #3. In fact, my circuit solves the HUGE problem that circuit #3 has, which is that the motor drive current NEVER goes to zero, causing the batteries to discharge even when the panel is at rest.
As OZnog said, it is much better to use a microcontroller to do all the work as there are so many variations to consider. These can all the done with a micro ($1.50) and a few parts. You can even use a $5.00 servo to do all the work as it has inbuilt gearing and use the solar panel to give you feedback. You can make the whole system "intelligant," for no additional cost, as it will search back and forth to pick up the best position. It will also "look" in the morning for the sun. All at the cost of 10 microamp quiescent.
I did some reading about using LEDs as photo-detectors, and got motivated to try an experiment. I tried some moderately bright red LEDs that I have laying around, similar to these. They are in a water-clear T1 3/4 lens. I do not know their viewing angle; i'm guessing they are the 30 degree variety, because they were bought for indicators.
Shunting an unbiased LED with a 1megΩ resistor, under outdoor illumination on a sunny day, I get about +1.5V anode to cathode, using a Fluke 73 DMM (>10meg input impedance). The LED is directive, consistent with the "viewing angle" shown on the referenced data sheet.
I then connected two LEDs in parallel, which approximately doubled the output voltage, indicating that the current produced by the LED (into the 1megΩ resistor) also doubled. I then reversed one of the LEDs so that the pair is connected in inverse parallel with the 1megΩ resistor, and bent their leads so that their optical axes diverged by about 45 degrees.
The inverse-parallel 2 LED array produces +- 1400mV as the array is panned across the sun, with zero volts out when the optical axes of the two LEDs roughly bisects the angle formed by the two diverging LEDs. Seems like this could be used to drive a Sun Seeker, because the 2 LED array puts out a signal which goes from +1400mV to zero to -1400mV as the sun angle changes.
To construct a Sun Seeker using the inverse-connected two LED array, you may have to put a small baffle that splits the angle between the two LEDs, orthogonal to the sun's path. In the morning, when the array is parked looking at the wrong horizon, the baffle would cast a shadow onto the west LED, allowing the East LED to dominate to drive the seeker toward the East horizon. After the seeker locks on, the baffle would be edge-on to the sun, so the differential mode of the array would not be effected by the presence of the baffle.
Using this idea, I modified your circuit to include a deadband near the balance point, and made it so that the motor switching transistors use a Darlington connection to minimize power dissipation while the motor is running.
Attached is a schematic and a running LTSpice simulation. The Solar Flux is simulated using a current source, which produces a voltage across the LEDs similar to what I measured outside. I use two comparators to implement the deadband region near the balance point, so the motor is switched off cleanly once the seeker centers itself. The transistors dissipate less than 1/2W when switching 1A. Substituting a PFET and NFET for the transistors lowers the part count, if you have the FETs.
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?