|
Overview of Ucoba's circuit |
|
Systems Breakdown: Pins 2, 3 &17, 18 - Phototropic Bicore (Pcore) Pins 4, 5 & 15, 16 - Motor Bicore (Mcore) Pins 6, 7 & 13, 14 - Drive Nv Pins 8, 9 & 11, 12 - Drive Nu (reverser) |
(note that these are the values that I used, by no means are they the standard)
74HC240
PD - Photodiodes
C1 - 0.22uF
C2 - 2.2uF (reverse timer)
R1 - 8M (the motor timing resistor)
R2 - 47k (Pcore and Mcore coupling resistor)
R3 - 6.2M (sets max. pulse from the Mcore)
R4 - 6M (dark adjustment)
R5 - 4M (reverse timer)
R6 - 330k (switch sensitivity)
The Pcore is the part of the circuit that makes it phototropic. It is an oscillator that varies with light intensity, in bright light it is very fast, the opposite in its absense. If one of the photodiodes is receiving more light than the other it will shorten the pulse of that one, while keeping the other's pulse the same. This is where the phototropic behavior comes to be. R4 isn't nescessarily needed. If you plan on having your bot roll under light when you can barely see it, then it would be wise to put one in. R2 is a coupling resistor that determines the sensitivity that the rest of the circuit (the Mcore directly) is to the Pcore, the higher that R2 is the less influential it will be.
The Mcore is the central oscillator . It provides a reference between the Pcore and the drive Nv, smoothing out the Pcore information if you will. This is dependent on R1, with the smoothing effect increasing in direct proportion to the rise in resistor value.
The drive Nv produces a limit on the pulse duratation from the Mcore. If the input from the Mcore is shorter than the time constant of the Nv then it will fire according to the Mcore's output.
The switches are feelers used for obstacal avoidance. When activated it will change the output of the Nu from high to low for a time equal to C2 times R5 (T=RC). This reverses the motor on the opposite side of the feeler that was touched. When both of the feelers are touched then both motors reverse. This is a briliant way to reverse a motor.
Notes on behavior:
In no light: the bot will oscillate at the frequency set by R4 (due to lack of light), which should be set at a high value as to get more out of the rest of the circuit. So in it's movement there is a great degee of waggle in its forward motion, ideal for seaching out a stray light source.
In mild Light: The waggle dwindles with respect to light intensity, but is still noticeable in daytime room lighting. Phototropicism isn't as prevailent as it is in a bright light source but will still end up going to the brightest source of light (eventually). My test for this used a desk lamp on the floor (the only light source) pointing down.
In bright light (sunlight): There is no waggle in sunlight, instead there is a continuous forward motion, wich is realativly slow. The phototropic effect is extremly sensitive in sunlight, usually the shadows cast are a great deal less intense of a light source. While the bot rolls around in the sunlight (usually heading towards the sun) and one of the photodiodes is in a shadow it will turn back towards the light. If it happens to roll out of the light all the way, it will (for the most part) turn around and head back towards it.
|
Random Notes: Changing the Nv resistors from going to ground and suspending it results in the following: one wheel (A) will start to turn, both then turn simutaneously, A stops while B keeps turning. Then a brief pause followed by the same cycle. It creates an angled direction. This only takes place under poor lighting conditions, otherwise it behaves the same. |
|
Construction tips: Place a physical barrier in between the photodiodes, (electrical tape on the inside of each works great for this) the reason being is that it greatly enhances phototropicism. It may be needed to put resistors in series with the photodiodes (if they are out in the open) for direct sunlight, sometimes the circuit will stall. This can also be overcome by covering the top of them with something (i.e. more electical tape) Use the D1 solar enginge, the PM1 solar engine or the 1381 DLSE to make it solar powered. |
| Further Reading: |
|
Tildens original BEAMAnt sketch Tilden's Telluride Handout Wilf Rigter's circuit analysis |
