Color sensor with BASYS3 VHDL

Hi, I need to make a circuit which does the following thing:

When it sees a red object it will send output 0 until it sees a green objec=
t (like a well colored cubic toy), after it sees green it will send output =
1 until it sees red again. ( If it is hard to implement I'm ok with just ou=
tput 0 when it sees red and outputs 1 when it sees green, I mean it is ok i=
f the outputs not continuous but it would be great if they are continuous)=
=20

I have basys3 and have to use VHDL. I think I should use TCS34725 sensor ht=
tps://www.adafruit.com/product/1334 but I'm not sure I just assumed, I woul=
d be really glad if someone helps

Some extra questions: If it is possible I will implement it on a toy RCcar =
is it possible to connect it with Bluetooth module to get inputs or how sho=
uld I do it ?- I can use another board JUST FOR CONNECTIONS my main Project=
 should be on BASYS-3

Thanks
=20

On Monday, March 18, 2019 at 2:28:50 PM UTC-4, utkud...@gmail.com wrote:
> Hi, I need to make a circuit which does the following thing:
>=20
> When it sees a red object it will send output 0 until it sees a green obj=
ect (like a well colored cubic toy), after it sees green it will send outpu=
t 1 until it sees red again. ( If it is hard to implement I'm ok with just =
output 0 when it sees red and outputs 1 when it sees green, I mean it is ok=
 if the outputs not continuous but it would be great if they are continuous=
)=20
>=20
> I have basys3 and have to use VHDL. I think I should use TCS34725 sensor =
https://www.adafruit.com/product/1334 but I'm not sure I just assumed, I wo=
uld be really glad if someone helps
>=20
> Some extra questions: If it is possible I will implement it on a toy RCca=
r is it possible to connect it with Bluetooth module to get inputs or how s=
hould I do it ?- I can use another board JUST FOR CONNECTIONS my main Proje=
ct should be on BASYS-3
>=20
> Thanks

Are you familiar with finite state machines (FSM)?  I think FSM will be the=
 best circuit for this task since you want it to remember which color it la=
st saw. =20

The inputs to the FSM will need to come from a much more complex circuit th=
at reads the state of the color sensor via I2C and makes a decision as to t=
he color it is seeing.  This will need to pick three states of the input co=
lor, red, green and neither. =20

To get the color data to make these decisions will require a controller cir=
cuit to read the registers in the chip.  Looking at the data sheet it looks=
 like this is not completely straight forward but requires a sequence of op=
erations to make the chip sample the light through ADCs.  So you should mak=
e sure you understand this process.  You may want to read the library code =
to see the sequence of operations implemented there while correlating that =
with the data sheet. =20

So read the data sheet, look at the code and make sure you understand what =
is required to get the data out of the color sensor.  Once you design that =
circuit you can try testing it by just driving I/O lines with the data valu=
es and looking at them on the oscilloscope.  If you pump the samples out se=
rially, you can see it on the oscilloscope with the low order bits toggling=
 furiously but the higher order bits fairly stable but changing as you chan=
ge colors. =20

When you think you are reading the color chip correctly, add the circuit to=
 weight the values and produce the two signals "red" and "green".  See if y=
ou can get these to work. =20

I would say you should construct a test bench and simulate it, but explaini=
ng to you how to construct a model of the color chip might be more than you=
 can do.  If you think you can do that, the test bench would be much better=
 than working on the actually FPGA because you can "see" any signal in the =
design.=20

Rick C.