100 x 100 photo-resistor led matrix

So i have an idea about somehow maing a 100 x 100 matrix od photo-resistors and LED's. it will hopefully all be mounted in a hole board, like the ones in stores that hang things on.

the goal is to make it so that a shadow that is projected on it (by standing in front of it), will light up with the LED's.

maybe inmliment into it some cool features like only lighting up an outline around you, or making it sparkle, or both at once.

what im looking for here is ideas on possible presets (like the outline sparkle thing), and ideas of how the crap i would make a matrix this large for cheapish.

i know that if i ebay the resistors, and the led's i can get them for probably around 75-100$

Maybe it would be easier just to use a camera than a photoreisitor aray, but anyway, im looking for ideas of any sort.

if anyone wants to make this themselves i'm fine with that, i want to see it done, but if you got the idea from me, please mention me somehow in documentation of it :)

and for the record, its not a sure thing that this will ever get done even if possible.

OH GOD I LOVE THIS!

This is how they used to do scrolling text outside of building (like in NYC).  They would put the text on  a long strip of film and project it on a large array of sensors. Each sensor controlled a light bulb outside.  I would love to see something like this replicated.

 

There are much smarter folks than me here that can help you with this. It will probably be multiplexers or switch registers or something that you will need. Someone will be by soon to make a suggestion, I am sure.

…and for the record…

You should totally do this. You have the skils, you are a very talented person.

First idea that comes to

First idea that comes to mind is to use the LEDs as the sensors too… here are a few links to give you an idea:

http://thebox.myzen.co.uk/Workshop/LED_Sensing.html

https://www.robotshop.com/letsmakerobots/node/125

https://www.robotshop.com/letsmakerobots/node/17166

Just to hitch onto the back

Just to hitch onto the back of this post…

Multiplexers and shift registers are great for controlling many outputs, but with sensors in the mix as well you should also investigate things like I/O expanders. Depending on what micro or other controller you use, it may be cheaper just to use many little micros that each control a small segment of the whole matrix, and then have the micros synchronise (if needed) on a serial bus.

thanks

thanks for the compliment, the only real reason that i can think of not to do it is that it would cost quite a bit of money i think. like if i just had the PR’s and the LED’s wired togeather and put on a parallel powersource then it could be cheap (comparativly) but that wouldnt do anyhting but light up the non shadow parts of the board.

question/ problem

if i was to sense the light with the LED’s themselves, wouldn’t that make it impossible for me to emmit light out of them at the same time, since that would corrupt the data

Just a thought

I have read more than once that LEDs can/will appear full or nearly full on even when only driven at 80 to 90 % duty cycle (PWM). If you could manage to run your light test inside the times when the LEDs are off, you would be able to sense and respond all at once.

Right on birdmun - I have

Right on birdmun - I have tried this technique myself and you can definitely power the LEDs with a high duty cycle while still using them as sensors.

In my Allura project I used this to light up the column of LEDs that was detecting the most light, however it’s hard to see in the video because the LED ‘stylus’ overpowers the light from the LED matrix.
I can’t remember the exact numbers, but I think it was only a matter of a few tens of μS between starting to read the LEDs as sensors and detecting a ‘hit’ from the reverse leakage current. Even with a low-end 8bit micro controller you should be able to read the LED sensors many times a second and still have them turned on for >90% duty.

**I suppose that would make sense **

I guess that if i do it like that it would work, but im more worried about how im going to controll and get the signal from soo many sources.

may be stupid

Yes, may be I am too stupid to understand what do yoou want do ?!

But, can you explain precisely or may be do a little sketch of what you have in your mind ?

 

Thx.

here is what you are looking for

what i am thinking about is having a large scale matrix, (100 x 100 seperated about 1 1/2 inches by use of a board with holes in it like the ones that are used in stores to hang things on) of led’s that will each have an individual photo resistor that will controll it. with this, your shadow will controll the lights since the photo resistor will take the data from where the shadow is and use this to controll the lights, and i would like to make it not just be one setting (like shadow = led on) but i want it to be able to have presets like maybe the shadow makes the led’s sparkle or maybe only the outline of the shadow will light up. 

It would be cool to do a

It would be cool to do a small proof of concept. Maybe a line of resistors/leds instead of a big grid. I have a couple of MAX7219CNG chips which would probably do the job. It would be cool if they could be used for inputs as well as outputs but I’m not sure they would work that way.   

Sounds like a cool project!


true

thats true, i think i may do that, something like 8 led’s in a line. that still doesnt solve the controll 10 000 led’s problem, and the 10 000 input problem as well

start "petit"

Napoleon Bonaparte left his little island of Corse to begining the emperor of Europe and control 3/4 of the world !

Start by a matrix of 4x4 Leds

yeah i might do that

yeah i might do that

The idea is not to light up

The idea is not to light up the LEDs at the same time as you are using them as sensors, but to switch back and forth between light and sensor so quickly that you can’t see the difference. The LED can still spend most of its time being a light, so it will appear just as bright as a normal, full-time LED.

If you want to do this project on a big scale (at a reasonable price) I suggest you get comfortable using some cheap microcontrollers =)

Microchip and Atmel both have a lot of very nice and very cheap options that will work for you.