2011-01-15T21:54:00.000-08:00

Spent a few hours cutting birch today, got a rough prototype of the CNC mill assembled (dry fit). Next will be putting bushings on the rails, installing steppers and leadscrews, and assembling the spindle.

2011-01-13T12:43:00.000-08:00

Going to do tutoring this semester - C and C++. If you're looking at this for that reason, hello! E-mail is pretty self explanatory, stan AT stansanders DOT com.

Projects update:

CNC MILL

Stepper driver boards are all done in CAD, and parts are all in. After some issues with the PCB mill, ended up not getting those done. May end up just ordering them from a board house at this point.

ROBOT ARM:

Just started ET 202 - Intermediate Robotics. Instead of building the Lynx Motion arm kit, I'm engineering my own. If my math is right, the arm should be able to lift/push with 2lbs of force (compared to a measly 1/4lb for the Lynx arm).

STM32 DESIGN CHALLENGE:

Judges reviewed my entry and have advanced me to Stage II - Prototyping. Should be receiving an STM32 Discovery dev kit shortly. Uses an Arm Cortex M3 - very cool! My entry is a self-contained CNC controller box which will function similar to certain photo printers. User will insert a memory card with their gcode on it, and run it. No muss no fuss no linux. Unit will use the M3 for the brain to interpret gcode, handle a basic UI (lcd, few buttons) for changing machine calibration and such. Will use my DRV8811-based stepper driver boards inside, so this unit should handle small machines just fine (up to 2 amps per stepper phase).

That's all for now.

Projects

2010-12-29T18:14:00.000-08:00

All the parts for my desktop cnc mill arrived! I was going to use an existing stepper driver design using the Allegro A3982, but nobody on this continent has the chip. I ordered some really cool stepper driver chips from TI - DRV8811 and DRV8821. They're 28 pin SSOP chips, so I made a simple breakout board in eagle to allow me to breadboard the chip. With help from the nice folks at NSL, I exported gerber files from eagle, and milled the boards on an AccurateCNC 560H pcb mill. If you've never seen one of these, they're fascinating to watch.

Here's the computer screen during the milling operations. Those tiny lines are the paths the mill takes to remove copper, creating the circuit. The yellow circles are holes that get drilled out at the end.

Here are the boards during the milling process. The outlines of the pads and traces have been milled away, but there is still lots of extra copper that still must be removed. The large rectangle is a heatsink - the TI driver chips have a pad on the bottom which will be soldered to this in order to dissipate heat generated by the large amount of current used to drive the stepper motors.

Here is a finished pcb. At the end of the milling operations, the mill drills out all the holes, and mills the board edges out to separate it from the original large board.

I also am learning how to do nice surface mount soldering. After destroying one board and spending one hour being frustrated, I watched some youtube videos and everything worked great.

I picked up a dead playstation (the original one) for $5 from goodwill. I removed the CD components and the main board. I'm in the process of mounting a parallel port where the original player 1 controller would plug in. I'm planning on putting my power supply, parallel port interface, and stepper motor drivers inside the case. It should look pretty cool when it's all done.

The spindle for the mill is going to be powered by a 12 volt 2 amp motor made by Jameco, part number 232047

I also got an older dell machine, which will be the computer that runs the mill. I am planning on installing EMC - Enhanced Machine Controller. It's a Linux distribution based on an older release of Ubuntu, and is free. I haven't fiddled with it yet but it has a good reputation from what I can find. You can read about it here: Linux CNC

The Guy-Manuel helmet is currently shelved while I work on some other things. The master is 75% done, just need to add the ear pucks and the raised frame hoops. Once the cnc mill is finished, I will be able to mill out the circuit trace details for the ear pucks, and prototype the rest of the led boards and control circuitry.

I finished writing the code to run the large rainbow bars, and the 16 crawling LEDs right below them. I'm using a TLC5940 (another TI chip) to give me PWM on all these channels. Serial data is sent from an arduino running my control code and so with only a few data pins, I get more pulse width modulated pins than the whole unit has to begin with. The best part is the chips can be daisy chained. I have a prototype board ready to be populated, and if it works I will make another version that is 2 layers so I can avoid having so many pin headers and jumpers.

This is a demonstration of some of the code for the rainbow bars. The code allows me to store a 2-dimensional array of binary keyframes which represent the animation. The keyframes are cycled every N-cycles of the pwm clock (depending on what speed is desired). Meanwhile, the code constantly calulates fading values. This allows me to plug in only a few keyframes, and get a smoothly animated display.

School starts in a couple weeks. I'm hoping I can get the cnc mill finished before then.

_

Projects