Solar Energy Project: Working with Beam Circuits

For the solar energy project, I wanted to try and build one of the walker bots that I had seen as I was browsing around the beam resources. All of these bots called for ICs that I didn't have, so while I waited for an order to arrive, I played around with building some of the simpler beam circuits. The first circuit I built was the Miller Solar Engine from solarbotics.net This process was helped me learn about the IC of the 1381 voltage detector ( datasheet ) and the concept of hysteresis. In the circuit above, the 1381 tracks the voltage level of the 4700uF capacitor as the solar panel charges it. When the 1381 voltage trigger reaches its upper switching voltage, it brings the base of the NPN transistor high which allows current to flow through the transistor to drive the motor. Once the capacitor has drained below the 1381's lower switching voltage, the 1381 brings the NPN transistor's base low, and current is blocked until the capacitor is charged above the 1381's upper switching level again. The difference between the upper switching and lower switching voltage of the 1381 is what is called hysteresis, and what's great about hysteresis, it seems, is that it determines how long your load will be driven. By changing the capacitor across the 1381 ( C1 in the diagram ), the hysteresis can be changed. Here's two videos of an oscilloscope tracking the pulses of a motor when the Miller Engine has a 0.22uF capacitor at C1 versus a 10uF capacitor. Notice how the lower switching voltage is smaller with the 10uF capacitor, so the hysteresis is longer. This results in longer, less frequent pulses from the motor.

BEP Pummer circuit >>>
Outside of my picture at left there's two rechargeable AAs and a 5.5v solar panel responsible for charging them. This circuit uses the solar panel as a sensor as well as an electrical source. The circuit detects the voltage coming into the solar panel and doesn't allow current to flow from the batteries to the LED unless the solar panel voltage is low ( i.e. it's dark outside... )
Resources:
74AC240 datasheet
Make Magazine's "Pummer Dude!"
Costa Rica Beam

Next I turned to using the Maxim 8212 to figure out how to achieve greater flexibility with trigger voltages and hysteresis. Some quick characteristics of the 8212 are that it gives a high output when triggered, and its upper voltage trigger and hysteresis can be set by different resistor settings. This circuit allows a load to be driven for a set voltage drop. So, for instance, I can set my motor not to turn on until 6 volts is reached and to stay on until the voltage drops to 3 volts.

I used the following resistor settings in the 8212 setup to achieve a trip voltage of around 5.5volts.
( If you bother to check these settings against the official formula, don't be surprised if they don't match up. I gave up on the formula eventually and just started trying out different resistor values at R2 and R3 and measuring voltage with a multimeter. )

Maxim 8212 circuit >>>
R1 => 100K
R2 => 300K
R3 => 122K
R4 => 1K

Resources:
Solarbotics pdf on the PM3 Power Module
Maxim 8212 data sheet

Solar powered bicore light seeking head
On the left above is Mark Tilden's schematic for a phototropic bicore that I found a description of at this website. Using photodiodes connected in series to a 74AC240 Octal Buffer, this circuit will turn a motor in the direction of the photodiode receiving the most light. I successfully got this circuit working, and then I combined it with the Maxim 8212 circuit as shown above. Now the phototropic bicore is also photovoltaic, and when the charge from the 4700uF capacitor goes higher than the 8212's 5.5V trip voltage, the motor turns in the direction of the photodiode currently receiving the most light.




























And since solarbotics.net unfortunately uses frames so I can't link directly to their discussion of the 74AC240 Octal Buffer ( and since I'm not sure I feel experienced enough to rephrase the description... ) I'll repost their definition here:

"Octal Buffer / Line Driver with Tri-state Outputs The '240 is often called "the bicore chip," because we can take advantage of the 240's inverters to turn a single 74*240 into a bicore (in this case, only 2 of the inverters are used, and the rest are used for upping the current so you can drive a motor directly). The '240 also has tri-state outputs, so an enable line can be used to turn its outputs on and off simply (good for adding reversing capability to a 'bot). Since the '240 gives us "vanilla" (non-Schmitt) inverters, it is usable for either grounded or suspended bicore designs (but better for suspended). See also the 74*04, above (which is very similar to the 74*240, but without tri-state outputs, and with about half the output current capabilities), and the 74*244, below (which is basically a '240 with non-inverting buffers). Note that pins 1 and 19 are the enable (tri-state control) pins. Signals on these pins pass through an inverter before getting to the buffer's tri-state control terminals, which means that a low voltage at 1 & 19 would be a high voltage at the buffer, so the buffer would be enabled. A high voltage on those pins would be a low voltage at buffer's tri-state controls and would turn the buffers off. So tie pins 1 and 19 to ground to enable all of the inverters inside the '240 chip."