In looking into a very simple photovore, it appears that I will need two stage amplification from most phototransistors. It might be easier to use some sort of photocell. Should I look into driving something like a high gain darlington to run small motors from a photocell, or are those little round cells more like voltage devices, from which I could run a MOSFET?

Phototransistors (two stage) may still be a good idea, since I know they are tuned to standard IR LED's (frequently 880nm), narrowband filters. These may be easier to set up so that robots will "march", following an IR LED on each other's rear ends.

These claim to deliver 20ma. This might be enough for a small vibrator motor... is it? $0.20 ea, in 860 and 940nm central band.
http://www.goldmine-elec-products.com/prodinfo.asp?number=G18693
http://www.goldmine-elec-products.com/prodinfo.asp?number=G13767

This is a cheap way to get an IR LED, but it does not list the wavelength:
http://www.goldmine-elec-products.com/prodinfo.asp?number=G16816

These are even cheaper ($0.10), but again, we have no idea of the wavelength, voltage or current either for that matter.
http://www.goldmine-elec-products.com/prodinfo.asp?number=G18101

We need to get ideas together and order SOON. If we go with photovores, we need to consider the mechanical construction too. Hot glue is your friend. How do I identify fast-switching (shotkey?) diodes from my grab-bag?

Do those little photocells produce a voltage from light? Is it enough to drive an n-channel mosfet? Either way, I think we can build a photovore in 2 parts. A phototransistor might need a second stage to drive a motor, still only 2 parts to drive the motor.

Any ideas besides photovores?

Tried this out. It works with a simple, general purpose NPN transistor. Resistor around 200-1K ohm for desired sensitivity.

If we can get 1K pots, it will be nice. One pot per photovore may be sufficient. One side can have fixed gain, change the other side to match (mostly). I believe this circuit can be soldered BEAM/ants-marching style. We just need motors with fairly long shafts, and a mechanical design. You want large mechanical advantage on the motors, so I have done this before by using the shaft itself as a very small wheel.

I can get a large grab-bag of photoresistors from electronics surplus or similar.

As for linkdumping, Goldmine emailed out some interesitng cds cells. dual elements in one can, could probably be used to control left/right.

http://www.goldmine-elec-products.com/prodinfo.asp?number=G18730

at $0.99 they are not cheap, but not too bad for two in one package. curious about how mask/blinder them.

they also have a $2 assortment
http://www.goldmine-elec-products.com/prodinfo.asp?number=g14025

other useful bits and discounts can be seen on their main page.
http://www.goldmine-elec.com/

Not sure where the supplies from Halloween went, but I may have a pack of 5 or 9 battery snaps at home. (so we are not making these solar powered?)

9v non alkaline batteries can be picked up for 2 for $1 at Dollar tree. may not run for as long of a time, but like most batteries included kits, would last for a the first night.

we should still have some rubber spline (pretty much a rubber tube with some gripper tread on the outside. we found it worked pretty good on the 3v hobby fan motors last year. (not big, just increases the shaft to about 1/2cm or so.

Quote from adric on March 2, 2012, 13:45
we should still have some rubber spline (pretty much a rubber tube with some gripper tread on the outside. ... (not big, just increases the shaft to about 1/2cm or so.

Sounds about perfect! Enough for 20-30 motor shafts? We can attach it to a 2mm (diameter?) shaft?

Had some issues with the motors running too fast.
Added a current limiter.

Unfortunately, component values will differ with motor. Will need to do some tweaking. The current limiter value greatly effects the range needed on the pot for desired sensitivity.

Another issue with current limiter value choice is that motor can stall. It can take the current given without turning. This might push more towards the lower end of the current limiter resistance.

On the mechanical side, I have found that you will not want the motors too far apart. We want one motor firing to move the robot forward, while curving to one side. When the wheelbase is too wide, one motor firing tends to pivot the robot (nearly) in place. This could be interesting though. Placement of the battery will change the robot maneuverability. It could be fun. Very simple electronics, but people can tweak the mechanical design for desired characteristics. (Like move the battery forwards and back to change turning radius)

I still have plans to try a maximally minimalistic driver... just and LED as photodiode driving a small MOSFET. Sensitivity of this driver will need to be all mechanical (shades/aperture/pointing direction)

This is still fun for me. I kind of prefer building and tweaking to playing with things. I'm like that with RC planes. Once I get them flying well, I kind of loose interest. I'm not so much into precision or stunt flying.

Here's a video of the lower sensitivity red-led based sensor

http://youtu.be/ihImbBBV4oY

Awesome! Nice writeup too! Im really hoping you will have the parts, so that sunday i can get down and try and make one!

The response of the two legs was very different, and hard to tune.
I also fear that the sensor lag is making it harder for this to follow lights. It is a neat sensor, but I may go back to CdS based solutions for attempts to make a minimalist photovore.

All Electronics has the nice pot's that will clamp to a popsicle stick, $0.25 ea. However, they only have power MOSFET's. They are nice, nut they are $0.50 ea.

Electronic Goldmine has $0.13 transistors which will work well, but only slider pots for $0.25. This may not be such a bad choice, since these pots are usually well over $1 ea. However, they may not mount easily to popsicle sticks. Another positive is that I think they are encased, so they can be glued to the sticks.

After having built some 16 light followers, here are some notes on streamlining the procedure.

• select matching CdS cells
• Place motors on table in a position similar to where they will be in the final robot, with the same side of each motor facing you. Lay the diodes on the motors such that the stripe is towards the inside. Solder the diodes in place.
• Power up each motor by placing the + lead on a battery on the motor tab toward the stripe on the diode, and the - lead on the other side of the diode. Make sure that one motor spins clockwise, and the other motor spins counter-clockwise. If this is not the case, reverse one of the diodes.
• Locate the drain pin on two transistors, and bend it away from the other two pins.
• Strip about 3mm from the ends of 4 wires. The wires should be at least 5cm long each. Bend the stripped end of each of these wires into a small hook.
• Arrange the potentiometers as they will be mounted on the robot. One on the left, one on the right, facing forward.
• Arrange the transistors so that the gate and source pin are aligned with an edge and sweeper pin on each of the potentiometers. It will be desirable for the bent drain pin to point up (or forward) in this arrangement on the constructed robot.
• Hold the gate pin of a transistor to the desired pin of the potentiometer, and help hold it in place with the hook of one of the wires. Solder this 3-way connection together.
• In a similar manner, solder the source pin to the corresponding leg of the potentiometer with a wire end. It is best for the wires to head in opposite direction from these solder joints so that they are not likely to short when the robot bumps into things.
• Cut and attach wheels to the motor shafts.
• Place both motors in their desired relative locations on the finished robot, with the tab attached toward the striped end of the diode towards the center.
• Put a generous dab of hot glue on each motor, and attach one side of the popsicle stick frame. You will have a few seconds to adjust the alignment before the glue cools. When the parts are aligned, wait for the glue to harden.
• Flip the motor assembly over, and glue another craft stick to the other sides of the motors for added strength.
• Place the wheels of the motor assembly on the table and attach the + end of a battery to the motor tab attached to the striped end of one of the diodes. Attach a lead from the - side of the battery to the other side of the same diode. The motor assembly should try to drive away. Note which direction it drives. Place an X on the stick that the motor tries to drive away from.
• Repeat for the second motor, making sure that both motors try to drive the frame in the same direction, with X marking the stick which will be the backward facing side if the frame.
• Place the third craft stick where you will want it on the final robot. This stick usually forms a tail where the battery will be mounted, so most of the stick will extend out from the X side of the motor frame. Mark the desired locations on this stick where the motor frame sticks will cross.
• Put a dab of hot glue on each of these lines, then balance the motor frame assembly on the third stick, in the pools of hot glue. You have a few seconds to check the balance and alignment of this stick to the frame before the glue begins to harden.
• When aligned, wait for the glue to harden.
• Hold the robot tail down on the table.
• Place some hot glue on the forward facing side of the motor frame, and mount one of the potentiometer assemblies. Be careful that the placement of the hot glue does not interfere with the motion of the potentiometer adjustments.
• Repeat for the other potentiometer assembly
• Solder the bent drain pin on the transistors to the motor diode lead away from the stripe on it's corresponding motor. If the two leads cannot be connected without risk of shorting other leads, use a small length of insulated wire to make the connection.
• Repeat for the other NON-stripe motor-diode lead to the corresponding bent transistor drain pin.
• Solder the diode leads towards the stripe on the motors together, and solder them to one side of the CdS sensors. If the leads are not long enough to place the sensors where you want, use an uninsulated wire to connect the stripe end of the diode leads together and extend forward to the location where you wish to the one lead from each of the sensors.
• Solder the wire to the gate pin of one transistor (center pin on TO-92 package) to the other lead of the sensor on the opposite side. Repeat for both sides. This is a criss-cross connection where each motor/transistor/potentiometer assembly is attached to the sensor on the opposite side.
• Solder the positive battery terminal wire (red) to a diode lead toward the striped side of the diode.
• Take the remaining source wires that were soldered to the transistors, and solder them together with the minus (black) lead on the battery connector.
• Plug in battery and test

Soon after http://www.instructables.com/id/Simple-Light-Following-Robot/ was posted, it got featured under technology.