Self replicating CNC for 194 (or more) countries

During the last weeks I did some science with my homebuilt linear drive. Of course the spring scale used to simulate side load is also self made and my "kilogram" prototype is actually a very special 3.557kg version of the platinum-iridium alloy in Paris:

Getting geared motors seems to be a problem in some parts of this planet, thus the transmission is simple, but works fine:

Currently some long runs will show if the mechanics keeps it's accuracy over time...

During 3rd round of HackadayPrize I will do some experiments with linear drives. The main question is if "normal" threaded rods can move a CNC machine with high precision. Furthermore I will deal with discovering alternative methods of motion sensing, especially the camera sensors of optical computer mice. Let's start with a general video about linear drives that ends with a DIY printer:

Software an schematics are available on my pages:

http://homofaciens.de/technics-machines-cnc-v3-1_en_navion.htm

...finally I got an optical mouse with an A2051 chip so that I can read the raw camera data:

The video about those interesting motion sensors:

During the next section of the HackadayPrize 2016 I will try to optimize the linear drive for my CNC and see if the sensor chips of computer mice are suitable for use as linear sensors.

5 weeks went quickly (where are all the hours?) and it's time for a short summary. The main intention of this section of HackadayPrize2016 was to evaluate motion sensing. As shown in the past 2 logs, optical sensors from old printers or scanners and even en electromechanical solution work fine. I wasn't successful in using an optical mouse (yet) since the two devices I own don't support the PS2 protocol (they talk definitely USB only). The Arduino software works in principle as I could test with my ball mouse:

The cheap optical mice I have ordered did not arrive yet. I will make a video about those interesting option of motion sensing later.

The analog reading of optical sensors as suggested in a comment might become an option in the next section when I'll try to examine the accuracy of DIY linear drives.

Using hall sensors does not have significant advantages (if at all) over the optical sensors and since I am short in time I won't use them in my design (but keep them in mind).

I have build a simple electromechanical rotary encoder using three resistors and some wire:

More infos are on my project page:

http://homofaciens.de/technics-base-circuits-encoder-disc-2_en_navion.htm

Only one analog input is needed to read the sensor. I will do some experiments if this type of sensor can be used to build a linear drive, soon.

I have disassembled an old printer:

and created a video about how to reuse the optical sensors:

Next i will have a closer look at the four DC motors I salvaged from the HP 8500 all-in-one printer:

During the next weeks (the "anything goes" section of the HackadayPrize) I will start building linear drives mainly using parts from old printers, scanners, mice and whatever I find useful to disassemble:

First thing I need is motion detection. Printers use transmissive photo sensors as end switches and rotary encoders:

I have already demonstrated in a video how to use them in drives. I will address in more detail how to figure out the pin configuration of unknown sensors and what resistors to use on the sensor and the LED side. Old computer mice with balls were build around such sensors, too:

...but this one is still in use while I am typing this log, thus it's no good option...

...must get another one, but they are hard to find these days. Newer, optical computer mice have an interesting option - a tiny camera:

and a "conventional", transmissive photo sensor at the scroll wheel.

Please leave a comment if you know other sources for simple motion sensors.

Hall sensors as used in optical drives (disc motor) or 3 wire computer fans are another way to sense motion (thanks to Robert for his email):

[Logan] posted a link in the comment section using grey scale patterns:

https://botscene.net/2012/10/18/make-a-low-cost-absolute-encoder/

Have a look at anaQuad from [esot.eric] that uses raw analog input of sensors to increase resolution of motion sensing by software - interesting approach.

What else can be used from the old Printers/Scanners?

The motors:

and Gears:

They all have motor drivers (h bridges) that would be useful, too, but I have no experience in reusing those chips. Please leave a comment if you can give some hints about using those motor drivers.

Cordless screwdrivers or power toys are sources for motors and gears as well (there must be one somewhere in my cellar...)

There is a lot to be done the coming weeks...

The first CNC of my 3.x series is composed of cardboard and paperclips:

CNC v3.0 on YouTube

The build instruction is available on my project page:

http://www.homofaciens.de/technics-machines-cnc-v3-0_en_navion.htm

That simple machine shows some additional characteristics of my 3.x series::

5.) Stand alone machine (build in "computer")

6.) made with schools and universities in mind