Open Robot

The original DARwIn-OP Robot uses an on-board 1.6 GHz Intel Atom Z530 computer and a CM-730 ARM controller board. This controller board includes a 3-axis gyro, a 3-axis accelerometer and a Dynamixel bus to interface with the robot’s servos.

I intend to use an alternative electronics based on a more powerful ARM computer plus several peripherals. For the main computer, my current choice is the ODROID-XU which is a quad Cortex-A15 1.6GHz CPU with 2GB RAM.

Instead of using a controller board, I’ll be using a USB2AX to interface the Dynamixel servos and I’ll use I²C components for the gyroscope and accelerometer.

On the picture are shown the components that I already got and that I’ll be testing:

• I²C 12 bit 4 channels ADC (ADS1015 chip)
• I²C voltage-level translator (PCA9306 chip)
• I²C 10DOF with 3-axis gyro, 3-axis accelerometer, 3-axis compass, temperature and pressure sensors (L3G4200D, ADXL345, HMC5883L, BMP085 chips)
• Force sensing resistor (.1N to 100N)
• ODROID USB IO Board with GPIO/PWM/SPI/UART/I²C/ADC interfaces
• USB2AX Dynamixel interface

I’ll be using the native I²C interface I2C_0 that is available in the LCD connector (CON15) in the back of the ODROID-XU (the schematics are available through a customer email request to Hardkernel).

The LCD connector is an IPEX 40 pin with 0.5mm pitch. I got some a couple of cables from Ebay, but now I saw that Hardkernel (the maker of the ODROID) is also selling the cable. The pin configuration that I’ll be using for the I²C interface is:

• 4 – 1.8V
• 6 – SCL
• 7 – SDA
• 10 – GND

The ODROID-XU’s I²C interface is rated at 1.8V, so I’ll be using the PCA9306 to translate it to the rest of the I²C components which are rated at a higher voltage.

The DARwIn-OP requires a gyroscope and an accelerometer, but I am interested also in using the compass and the force sensors (through the AD converter) for a future version of the DARwIn-OP feet.

After doing some tests, I finished with all Idlers anodized. I am still waiting for some black dye to arrive, but the Idlers and Caps don’t use dye.

The process used was:

1. Polish the parts.
2. Wash the parts with soap and rinse in distilled water.
3. 1 minute bath in sodium hydroxide (10g/L).
4. 60 minutes of Type II anodizing: 12A/Ft² with 15% sulfuric acid (battery acid diluted 50/50 with distilled water) at 20°C.
5. 30 minutes boiled in distilled water.

Since I did my first test cuts with some aluminum 2011, I used this test Idlers to try the anodizing process first. The surface finish will depend on the aluminum alloy used. With the 6061 alloy I got a nicer finish.

I did get some parts ruined with some anodizing marks, specially in a few Idler Caps. At least I have a few spares and they are easy to make. I saw in a tutorial video in YouTube that the parts are never clean enough to anodize them. Anyway, I need to do more research and tests before anodizing and dyeing the frames.

I will also test using a longer sodium hydroxide bath. The aluminum oxide buildup by anodizing is noticeable since the ball bearing fits tighter in the Idler, there is space for a longer bath (since it eats up some material).

Now that I finished the Flange Bearings and Bearing Idlers, I am doing a test of the full assembly of the first leg of my DARwIn-OP clone.

After adjusting the bends of the FR07_H120 frames the idlers align perfectly.

Now I’ll continue making the rest of the HR07_I101 Bearing Idlers (still 9 to go) before starting to bend the frames for the second leg and start cutting the frames from the waist up. Anyway I am still waiting for some black dye to arrive before anodizing the frames.

I just finished uploading the remaining gcode files for cutting all the leg frames for a DARwIn-OP robot clone.

They are available in the Code Section and zipped together in File Section at the sourceforge repository for cloning the DARwIn-OP Robot.