Main Robot - Day 3 1/2

We worked on getting the steering system to stop at center using a touch sensor. The steering now stops at the center. We also added supports to the RCX brick so that it stays on better. Now that the final design is mainly done, we can start programming for the track.

(Touch sensor design)

(Brick supports)

Main Robot - Day 2

We adjusted the rear drive train a lot today. We first tried the TORSEN differential, but there was too much torque on the differential when the locking system (the part that makes both wheels spin equally when one has been spinning faster than the other) engaged, and the differential started flying apart. So we (Paul) built a limited slip differential that doesn't have quite as much torque on it when the locking mechanism engages. It works perfectly, but the drivetrain is geared down a lot so it doesn't move very quickly, but it seems pretty strong. It doesn't make its way up the stairs yet, but a bit of tweaking will change that.

Aside from that, we built a front bumper system that has two independent sensors that trigger when something bumps into either side of the car. It's programmed to reverse and change direction when it hits something then move forward agan in an attempt to avoid the obstace. It's pretty crude, but seems to work.
(TORSEN differential w/ "motor box" powering it)

(Limited slip differential)

(Limited slip w/ "motor box" attached)

(Car now)

(Another shot of the car now)

The bumper system is up front.

Main Robot - Day 1

Today we started building the robot that would navigate the final course. It consists of a front steering system (Figure 1) with rear wheel drive. The drivetrain consists of three engines, driving one single drive shaft (Figure 2) that gears into an open differential (Figure 3) which allows each wheel to spin independently for better turning.

Figure 1

Figure 2

Figure 3

Since the open differential is bad if the wheels start to slip (like on the sand) we might switch to a TORSEN differential system which has the good properties of an open differential (better turning) along with the fact that it is a completely mechanical system and transfers torque to the non-slipping wheel if one wheel starts slipping (property of a locking differential). The amount of torque transferred is a function of the torque bias ratio (TBR). A TBR of 3:1, for example, would mean that one side of the differential could handle 75% of the torque, while the other side only needed to handle 25% of the torque.

From wikipedia: "When the traction difference exceeds the TBR, the slower output side of the differential receives the tractive torque of the faster wheel multiplied by the TBR; any extra torque remaining from applied torque contributes to the angular acceleration of the faster output side of the differential."

Pathfinder Robot

We completed our pathfinder robot today. It is a powerful robot with one engine per wheel in the back, and two engines powering the single front wheel. Designed this way, the robot can easily climb over the stairs. The robot is as shown below (Snapple not included).



We (Paul) also completed a VERY COMPLEX AND AMAZING steering system that will be used on our final navigator robot. The single motor powers the steering mechanism, which will turn the wheels that will be added later.

Starting robot building

Today we began designing and programming our robot. There are no pictures because we played more with the programming language (NQC) than with the robot. More to come tomorrow...

NQC + Mindstorms

We got introduced to the Mindstorms legos today. We started getting set up to transfer programs to the robot by connecting the USB tower to the robot over IR. It took a bit of finessing, but we were able to get the brick to connect to the USB tower and transfer the firmware to the brick so we could start transferring programs. We compiled and transferred a simple program that made a few simple things happen on the robot.