The Roboard mounted on a KHR1-HV.

The original KHR back cover fits with the Roboard well. The antenna is for 802.11g wireless networking. The wires going from the head to the back are for the USB capture cards.

The NiMH battery has been replaced with LIPO available here and here. After some cutting, the original cover fits and nicely holds the battery in.

Standard Kondo gyro sensors. Fit niceley inside the torso.

Compass/tilt sensor over i2c bus. It had to be mounted on the top like that becuase the inside of the torso shielded the magnetic sensors.

The sensor headers on the Roboard don't have power output. I could pull 5v from the usb ports, but they are all occupied. The i2c port is also occupied. My analog sensors all run at 5v, but the compass/tilt sensor runs at 3.3v. So here's 5v and 3.3v power.

This is an IR motion sensor, the same kind used for security. It should help identifying humans or animals. I am thinking of mounting it on the head.

The IR distance sensors are on the feet as before, and their plugs have been spliced up for the new arangement.

The 2 USB capture cards, previous backplate for head with a new hole in it, new casing for the cap cards, and the back of the previous head opened up with camera boards visible.

The cap cards in their casing.

The back of the head with the new hole in it, which the capture card casing will cover. The 2 wires are the analog audio/video from the cameras to the USB capture cards. The image of the back of the robot at the top of this page shows the complete setup.

Previously, the cameras transmitted a wireless NTSC signal to a receiver attached to a USB capture card attached to a computer. Now the USB capture cards have been directly attached to the cameras and Roboard. Unfortunately the NTSC step is ruining the quality of the images, and introducing interlacing issues. This analog step is rediculous but a temporary solution as I build yet another stereo rig, this time pure digital.