I want to make a scanner, or maybe a "modeler".

Background info keywords:


https://www.physicsforums.com/proxy.php?image=http%3A%2F%2Fi.dailymail.co.uk%2Fi%2Fpix%2F2014%2F01%2F15%2Farticle-2539962-1AAFD4DD00000578-400_306x317.jpg&hash=18c044d58272f31902c0c6194ba64eba   <- PICTURE



I want to make something that detects where each position is on a Z-axis (their x and y coordinates are fixed). So just a third number per-pin would allow me to map out the entire layout of the pins. I am not too sure if this has been done before or not. If anyone knows of a product that does this I would really appreciate more info.


- Something that takes feedback from each pin and maps it out in the computer.


- How to read the feedback? (what sensor can I use, note that there will be hundreds of these sensors, so they have to be preferably small. Are there some very small linear position detectors? is an optical detector better? (I WOULD PREFER IT TO BE STURDY, i wanna play with it).

Im thinking it should be something similar to how TVs work, with a pixel-matrix but instead this would be a matrix of position feedback, and I think it would be simpler than a TV since each channel only gives out 1 feedback (Z-position) instead of an RGB code (3 codes per pixel). 

Anyone have some ideas on how to develop this project.

Thank you!

david606 (author) 2 years ago

Let me be more clear, I will simplify my problem further.

Can anyone help me find something that will sense what I require in the attached picture.

david606 (author) 2 years ago

Thank you for the comment, its really insightful as to how the video I linked could be made. However, I wanted to make a simplified version of this, and again, I am looking more for a feedback than a command-generator. I want a pin-screen with something that gives me feedback on how deep the pins are. So, no actuators would be necessary, just some sort of sensor.

That project seems way too complicated for a simple hobbyist like me at this time. but you're ideas seem worth looking into. Thank you for the feedback tho.

I have seen these videos when the project started and also read a bit about their progress.
To get to the level we see in your linked video it takes a lot of money, hardware and coding.

With no time and not enough to start the project I might as well offer my collected thoughts on it, maybe it helps you to find some ideas.

I wanted to keep parts available and me capable of getting it together, so instead of linear actuator I thought of a combined spring air pressure system instead.

As normal air cylinders are too bulky something home made hs to be done, I settled on plastic tubes as cylinders with teflon pistons - easy to do on a lathe.

Each piston is connected to a dyneema string going through a pin hole at the end of the pipe.

From there through various "pulleys" to little stepper motors equipped with indexing discs - the kind you find in old ink jet printers for example.

By default the system is pressurised and all motors set to stop by simply putting the holding current and voltage on them.

If a piston is activated by the software, the motor is released.
Through the indexing disc the software can get the accurate position of this piston and once the set distance was travelled the motor is set to full stop again.

To get the piston down the motor is activated to wind up the string until it reaches the end point.

Good things about this design:

Very fast movements possible.

Relativly easy to build.

Bad things that might need to be addressed:

A compressor is needed.

A lot of motors are needed.

A way to protect the electronics from the electricity generated by the motor when releasing is neceassary.

Changes I thought about:

Do all movements with the stepper motors, this however requires to use either piano wire or thin, flexible steel slders.

Using a steel piston and electromagnets on the pipe to position the piston, but this limits the psotions to a few fixed ones.

Going a similar way like the pro's and use actuators, but even a 8x8 matrix would explode in terms of building costs.

For the positioning data, have you though about 3D images?

These old fun ones that you can generate on your PC are made from from a 2D image in BW, the shade of grey indexes the depth of the pixel.

So basically the darker a point is the further away it is.

Like this you can use a 2D image for the shape and a grayscale pic to deliver the info on the depth.

But I highly doubt this can be done with an arduino.
You need too many inputs and outputs, unless of course you use byte addressing and the corresponding hardware after the arduino to get the signals working.

You are addressing a lot of points in three dimensions at reasonable high speeds, if you want more than a display but instead something actively reacting to the input you need a faster processing base.