3D printed ball bear hinge have insufficient clearance and space in the shaft area for the other part attached to the other aluminum profile to rotate smoothly. We made it tighter and thinner for clearance and was able to attached it to the legs.

Once everything was assembled and attached to the body, we realized that the overall angles of the aluminum connected to the motor and servo housing are not leveled so changes had to be made. As such, changes to the angle where the rocker joint and motor + servo housing is attached to the aluminum profile was adjusted so that all the pieces are leveled and stable We also designed a part that can fit into the groove of the wheel for the coupler to be attached to.

The changes made to the motor and servo housing allowed the legs to be leveled and stable. Upon attaching both legs and testing out the motion, we found out that the ball end joint is slightly angled and not parallel to the legs. This caused the right leg to be bent inwards if the left leg is tilted up, due to the pivoting of the differential bar. As such, the differential bar was made to be longer to fix this issue. Also the PCB was ordered 🙂

 

17/7: We continued exploring the fuse, voltage divider, PCB connector and Ina219

18/7: We had our 3rd presentation to showcase our progress and changes of our project to our professors

20/7: Abigail have returned to Singapore! so we had her look at her current and previous designs so that she can understand what went wrong and what was successful. As mentioned in other posts, one of our current challenges are applying the appropriate bearings to our design for the rocker bogie’s leg to rotate smoothly. As such, we explored the use of hinges and the use of shaft in ball bearings. The use of hinge requires a tapping screw that we do not have, while the ball bearing application requires us to design another 3D part to be attached to the AP.

21/7: Failed to connect to Dlink camera as they have their own server

 

10/7:  All the changes implemented into on our 3D models have been made and the second prototype parts were printed. First prototype was disassembled to reuse the aluminum profiles and we were able to assemble 1/4 of the rover. However, we failed the first trial the PCB and are still deciding what will be used for communication on the user level, a source for battery (drone?) and good voltage regulator (protection?). We will also be using a 24V battery, step down 3.3V for logic and 5V for power (giga, rpi, camera, router &servo).

12/7: We decided to throw away pi, use wifi direct to giga, decided on a one 5V step down regulator and decided to add imu

13/7 : Besides more printing and assembling, we finally got to cut the aluminum profiles today. We decided on linear bush bearings and metal rod for pivot and tested giga(blink) over wifi

14/7: The body of the rover and the remaining legs was fully assembled, however the legs cannot be attached to the body of the rover yet as we lack the necessary screws required. The rocker joint and rocker to chassis parts will be combined as putting them into 2 separate parts are not necessary and troublesome. We are still improving the pcb design and hope to purchase linear bearings and couplers soon.

We finally got all the material to put the legs of prototype 1 together and got to assembling but came across some unexpected issues a long to way. The use of deep-groove bearings was not applied correctly in parts that requires rotation as there is still not enough clearance between the pieces to rotate smoothly. As such we have to explore other options and ways to make the bearings work.

The size of the nuts that came with the M8 screws purchased previously was also too big for the nut grooves placed in the parts, so it requires amendments. We also decided to combine the servo housing with the motor housing leg pieces together, and turned the ap from the servo housing to be fully 3d printed instead of using aluminum profiles.

Additionally, a 3D printed mold was done to assist in the making of the seed pods and provide a consistent round shape of ~3.5cm

03/07:

We discovered that the aluminium profiles require a specific 20mm x 20mm aluminium bracket so the surface change of our 3d models have to be tweaked. The grooves on the back of the 20mm x 20mm aluminium bracket was measured and taken into consideration so it can fit on our 3d printed models

– Changed dev mode to 2

– Got the giga to control the motor (Pam too fast for motor)

– Got PCA9685 to control motor

– Tony suggested using PCA9685 to control all pins for DRV8874

04/07: We printed the 3D parts that connects the aluminium profiles on the side and the part that connects the wheels to the body (joint). We also printed the servo housing and legs We also that the 3D model of the differentiator is too big (40cm+) for the lab’s 3D printer as it can only print a dimension up to 20x20x20cm. As such we broke the differentiator model into 3 parts, using the tongue and groove method and screws to secure the pieces tightly together. We changes the 3D printed piece below the differentiator into an AP for more strength, it has its own 3D model design and print to connect those 2 pieces together.

Test motor strength & current, decided on servo min & max of 145 & 545, use pca9865, assembling of rocker bogie middle legs with aluminum profile

I2c board not working, tested motor with 1298 n to be working. We also set up a GitHub account & use motion study to test out the movements of our 3D design to prevent future errors and collision, especially the leg pieces

The ESP32 arrived today, so we tested the servo with it. We also learnt that translational + rotational steering is called differential serve. We flashed raspbian and got pi to boot, and pi 3 is working, but it can only run melodic and not Nordic, hence we are installing pi4 to install Ubuntu 20 for noetic.We hope to control 26gp 555 with the driver and esp32 tomorrow

The 3D print of our revised seed disperser (prototype 2) was collected, however further revision will be required. The design comprises of a long , thin, rod that makes it difficult to print (requires a support) and very fragile due to its plastic material.  We also explored different types of brush and brushless motors, and how we can utilize Arduino in our design. Additionally, we acquired motors of the same model, raspberry pi and started printing the first prototype of the rocker bogie’s legs

The 3D modelling of the seed dispenser was revised from having a total of 5 holes of 3.5cm to a total of 3 holes of 3.5cm to reduce the overall size and weight of the seed disperser.

Some items were also procured (list) and the total weight of the system was lowered from 20kg to 10kg to decrease the strain placed on the support system. Additionally, the use of rocker bogie system in our vehicle was confirmed to disperse the seeds and travel through the expected rocky terrain. We found a suitable motor model for it (36GP 555) and acquired servo motor for the seed dispenser system