Hi all!

My partner and I are creating a project for our analog circuit class. Our project requires us to build a purely analog project and we chose to create a car that follows a UVA flashlight in hopes of minimizing the interference with ambient light. We don't have a schematic of our project because we are having trouble brainstorming ways our circuit would be connected. Our car will be powered by a 9v battery, two motors, two comparators used for turning (open for suggestions), and an L293D motor driver IC.

We are now having trouble finding affordable sensors for our project and questioning if it's a good idea to follow through with UV light. We also found out that these sensors require true UV light (UV light that produces wavelengths > 365nm) to perform at an optimal level.

Some of the questions we have now are:

1. Is there an alternative method that would allow our car to move while being interfered with ambient light (such as using IR sensors) and how so?
2. Will the comparators work well with the L293D IC If not what other ways could I possibly connect my circuit in order for the car to turn when we need it to?

P.S. This is our first time building a project from scratch and we apologize if some of the stuff we said above does not make sense. If you guys have any questions we will try our best to answer them. Here's a rough draft of what our project will look like if you are curious. Thank you in advance!

 

hi bryan,
Regarding the UV option, fluorescent room lighting does also radiate background UV as well as visible light.
I would consider the IR option, modify your transmitting torch to send a pulsed IR signal.
The receiver can then be designed to respond to only an AC signal and not the the ambient light level.

Your diagram shows a separate motor for each of the back wheels and a single wheel at the front, how do you plan to steer the vehicle.?

Attached is a correctly orientated image of your project.

E

 

hi bryan,
Regarding the UV option, fluorescent room lighting does also radiate background UV as well as visible light.
I would consider the IR option, modify your transmitting torch to send a pulsed IR signal.
The receiver can then be designed to respond to only an AC signal and not the the ambient light level.

Your diagram shows a separate motor for each of the back wheels and a single wheel at the front, how do you plan to steer the vehicle.?

Attached is a correctly orientated image of your project.

E

Thank you Eric for the reply, We were planning on having the back wheels turn based on which comparator is being signaled. For example, If the comparator from the right side of the car (in the reference frame of the car) is being signaled from the sensor then the motor from the right side will stop and the left motor will continue to produce movement. This will then cause the car to turn to the right and vise versa. If the sensor is being signaled from the front then the car will move forward. We were also planning on using the L293D motor driver IC with the comparators.

 

hi bryan,
Post your draft design of the circuit when ready and we will help you with debugging.
E

 

I would put the "following sensor" on the other end. Make that idle wheel, a trailing wheel.

Use motors to steer.

 

Early RC multi-channel control used tone and tuned reeds for channel decoding. In other words analog. Have you considered modulating your light sources with high/low tones. For example, left could be high and right would be low. That would also help address interference from ambient light. You could make your own tone filters or use the LM567 chip that was used for tone decoding in telephones.

Then, integrate the left and right channels to get your comparator inputs. Of course, you will need to calibrate/zero the device as it will be hard to get two sources that give exactly the same intensity.

As for IR vs. UVA, I think UVA would have as many problems from ambient light as IR will have. However, IR for communication is more advanced, so you will have more choices and probably lower cost going with IR.

 

Early RC multi-channel control used tone and tuned reeds for channel decoding. In other words analog. Have you considered modulating your light sources with high/low tones. For example, left could be high and right would be low. That would also help address interference from ambient light. You could make your own tone filters or use the LM567 chip that was used for tone decoding in telephones.

Then, integrate the left and right channels to get your comparator inputs. Of course, you will need to calibrate/zero the device as it will be hard to get two sources that give exactly the same intensity.

As for IR vs. UVA, I think UVA would have as many problems from ambient light as IR will have. However, IR for communication is more advanced, so you will have more choices and probably lower cost going with IR.

I like that idea of tone filters i'll see if there is a way I can incorporate that into the project. Thank you for your input!

 

Here is the schematic design we came up with. We are going with an IR photodiode and we tried to put it together using a breadboard but it's not working. We are currently using a OPA347 as our comparators and IRF510 MOSFETs as switches to our motors. We were thinking if we could possibly convert our op-amp into a comparator/ transimpedance op-amp in order for our transistors to detect a voltage. If you have any questions, concerns, or suggestions about the design please feel free to ask, It will greatly help us out thank you!

 

By not working, what do you mean? Do the motors run at all or do they just not steer? What "control" tests have you done? ALWAYS think in terms of controls (i.e., a test with a known stimulus and known/expected result). For example, one control would be to block everything except the right photodiode and see what happens.

As a more general comment, using "DC light" signals will not work well, if at all. It's been tried and is why every IR control uses modulated light. A simple LM555/NE555 will drive an IR emitter for short range testing. Second, getting consistent analog light intensities is difficult. That is why almost all spectrophotometers today are ratiometric. That is, absorption is determined by the ratio of intensity of the beam that passes through the sample compared to a beam that does not.

 

By not working, what do you mean? Do the motors run at all or do they just not steer? What "control" tests have you done? ALWAYS think in terms of controls (i.e., a test with a known stimulus and known/expected result). For example, one control would be to block everything except the right photodiode and see what happens.

As a more general comment, using "DC light" signals will not work well, if at all. It's been tried and is why every IR control uses modulated light. A simple LM555/NE555 will drive an IR emitter for short range testing. Second, getting consistent analog light intensities is difficult. That is why almost all spectrophotometers today are ratiometric. That is, absorption is determined by the ratio of intensity of the beam that passes through the sample compared to a beam that does not.

The motors will not run nor steer with the setup above. The controls are the trickiest part of the project and the way I was thinking the car will run is having a motor run when the photodiode receives a signal from the emitter. This current will then be converted into a voltage through a transimpedence op amp for the transistor to close its gate and finally have the motor to work. That's the way I visualize this setup but I'm sure it's not that simple and I'm sure it probably has to do with the intensity you mentioned above. We've also attempted to work with one motor to see if we could get at least one of them to work. Anyways, Thank you for your input I'll definitely keep it in mind!