Tested the force sensitive resistors

I tested the FSR, using an Arduino Nano and Arduino IDE
A 10 kΩ resistor was used to transmit the signal through analog pins on the arduino nano, and allowed for the FSR to accurately sense pressure

Finishing touches + changing parts

I modified the battery mount underneath the robot, so that the screw holes won’t conflict with the screw holes for the 40 amp buck converter

Additionally, I decided to use the Pimoroni Servo 2040 instead of the Pololu mini maestro, to decrease costs
Chamfers and fillets were used to modify the robot as finishing touches, and the leg design was copied and positioned along the frame

After this, I'll wait for the force sensitive resistors to arrive to begin testing those

Changed buck converter + Power distribution modules

Despite the listed stall current for the MG996R on the TowerPro website being 1.4 amps, many vendors on amazon and other sources list 2.5 amps
Realistically, the servos would never all stall at the same time, but to be safe, I decided to use a
40 amp buck converter
Due to the height of the 40 amp buck converter, the walls of the frame had to be raised.
Additionally, a 40 amp rocker switch was added and will be mounted to a wall on the frame. The idea is to have the battery adapters go directly to the 3 buck converters. The 5 amp buck converters will power the electronics independently of the 40 amp power switch that only toggles power to servos. This allows for easier debugging.

One 5A buck converter powers the pi 5, and the other powers the Pololu servo board

2 distribution modules, each 30 amps, are used to distribute power to servos

New electronics configuration on second plate

Redesigned frame + changed parts

Since the leg design was mostly completed, I decided to circle back to the frame. Due to the current design of the frame, and the positioning of electronics, it’s very difficult to add necessary parts like servo controllers, as well as separate buck converters for the servo controllers and raspberry pi. Additionally, the buck converters would most likely be unable to sustain the servos for extended durations without overheating, because their heatsinks are nearly touching the frame :(

The first change was moving the battery underneath the frame to make more room for electronics, also helping to lower the center of mass for better stability, at the cost of lower ground clearance. A gap was made on the back wall of the robot for the battery cable to get inside the frame

Additionally, the two 15 amp buck converters were swapped out for a more reliable 30 amp buck converter that has a set output at 6 volts.

The second plate was changed as well, and the Pololu Servo2040 board was added for controlling the servos. 2 power distribution boards, 15 amp 2x12 pins (12 ground and 12 voltage), are added and positioned on the left and right sides for powering the servos. This will help keep wiring more manageable, and the parts and their location allows for easy maintenance and future modifications.

Modified leg design

I made another change to leg design to further reduce weight

The supports for the middle segment of the leg are removed, as there is less load at those points than at the hip joint. I made the femur thicker to compensate for that change.
Chamfers are added around screw holes, and the ball bearings is added servo mount for the hip servo

Because it would be very difficult to attach the center servo bracket to the hip servo mount with the ball bearing if it were printed in one piece, I separated it into two parts that attach to each other after they’re positioned properly

Redesigned tibia + Force-sensitive resistor

Because the metal legs are heavy and relatively expensive, I began working on a 3d printable design to replace them. However, I still wanted the hexapod to be able to detect when the leg was touching the ground.
A lot of research was done on FSR’s, and the measuring range, accuracy, cost, and dimensions were taken into account to finalize one.
After multiple iterations, I came up with this design:

The bottom part of the leg would have a force sensitive resistor in between a 3d printed part and a hemispherical silicon bumper
The force sensitive resistor being used has a 1 cm diameter

The 3d printed part has a flat edge for the cable of the force sensitive resistor to travel upwards, where it will connect with wires that lead inside the frame
These parts would be encapsulated by a durable 3d printed TPU shell that is attached with a screw

Femur modification + ball bearings

Changes were made in leg design to better support the weight of the leg, mainly due to the metal tibia

Ball bearings will be installed at the three rotating joints of the leg to reduce friction, distribute mechanical load, and, ideally, ensure smoother and more consistent leg movement.

Additionally, the femur is also changed so that the two parts are no longer connected, as this connection makes it a lot more difficult to print and assemble

These changes should make it lighter, easier to 3d print, while still helping to support leg weight.

Created the leg

I decided to take a break from figuring out wiring and hardware, and did a lot of research on other hexapod designs, and how the leg design would work

Based on this research, I created a preliminary leg design with three sections and three degrees of freedom. To enable the robot to detect ground contact, I plan to use metal hexapod legs from an existing assembly kit. These legs include force-sensitive resistors embedded in the rubber “feet,” which detect when the leg touches the ground.

Ground-contact sensing will be important for confirming the leg is in contact with the floor, and will improve navigation over different surfaces. The rest of the leg structure (besides the servos) will be 3D printed to keep the design lightweight and customizable.

Redesigned frame+added buck converters

With more space within the frame for electronics because the servos are mounted more outwards, I wanted to test if the battery could fit within the frame with a better design. Unfortunately, I had to get rid of the triangular extrudes as they made it more difficult to make changes to the frame.
A battery mount was created that the raspberry pi would also be mounted to

2 buck converters would be used, with one on each side, as the servos and electronics like the raspberry pi and servo driver board would require a lot of power
A 3d model for a 15 amp buck converter was imported, and were positioned on the battery mount

This design is really sketchy though, the heatsinks for the buck converters are directly facing the bottom of the frame, and the lack of a distribution board would making wiring very difficult and messy

Redesigned frame+created servo mounts

When checking over the design of the frame, I noticed that it is too large to fit on my printers build plate

Simply moving the servos inward so that the frame could work as it's really close to fitting, but it would cause them to intersect with the battery and also decrease space within the frame for electronics.
The servos already take up a lot of space inside the frame, making it harder to add necessary electronics

To fix this, I created basic servo mounts so the hip servos could be attached to the outer part of the frame, freeing up space inside the frame, and allowing all the parts to fit within the build plate

Designed battery mount

I added some triangular extrudes to decrease weight and have more space for wires to go through
An Arducam ToF camera is added to the front of the frame for cheap obstacle detection using point clouds

An 8 amp 11 volt battery was chosen for better runtime
I tested out positioning the battery within the frame, but realized it took too much space and would make wiring more difficult

Then, I tried out having the battery mounted underneath the frame, and added a plate on top of the frame for additional electronics

Parts such as buck converters and batteries will be attached to the lower platform, and parts like the raspberry pi and power distribution boards could be attached on top of the second plate for easier access

Designed the frame

I spent a lot of time researching open source hexapods online
Each leg would have 3 servo motors for better movement
Having the legs evenly spaced out around the frame will make coding easier to do in the future
MG996R servos are low cost and have high torque, so I decided to use that as the motors for this robot
To start out, I decided to design a basic frame that I would build outwards from

A raspberry pi 5 will be used for better computing power and the potential for adding cameras and other sensors in the future