PRE2019 3 Group4 Design Flaws and Future Improvements
On this separate wiki page the design flaws that arose during the assembly of the electronics in the casing will be described. Also, if applicable, a solution is given to a design flaw such that this may be prevented in the future.
Micro USB cable hole
On the right side of the design casing a cutout is placed for the entry of the micro USB cable for the RaspberryPi (RPi). The hole has been created according to the standard dimensions of a micro USB plug, however the casing of the micro USB itself had not been taken into account. This posed problems, since the micro USB is now too large to enter the design casing.
As a solution to this problem the RPi has been moved outside of the design casing for the final design and tests. Luckily this could be done easily due to the availability of the RPi extension kit with a large ribbon cable. Therefore for the future it is important to also take into account the casing of the micro USB itself when creating a cutout for the micro USB.
Inner Walls Casing
The 3D printing process did not go completely according to plan. This is partly due to bad communication with the 3D printing contact and partly due to 3D printer failures. From the 3D printing contact it became clear that in the final design inner walls had to be added for increased stability. By adding these inner walls, such that the power supply would be on the left side of the casing, followed by the two PCBs that control the solenoids in the middle, followed by the RPi on the right of the casing, these parts would be kept separated from each other.
It followed that indeed the stability for the casing increased. However in the 3D printing process something went wrong due to which the inner walls had been printed twice, thus encompassing twice the space in the casing. This posed problems with the placement of the two PCBs in the middle, since these did not fit anymore in the desired location. Eventually the decision had been made to place one PCB on the left side (on top of the power supply) and one PCB on the right side of the casing.
When manufacturing the final design, problems quickly arose with respect to the space of the casing. During the design process of the casing, the space had already been increased once before the final casing was 3D printed. However with the RPi extension kit PCB not assigned to any particular location in the casing, it was hard to place this PCB on a proper location.
Ultimately, with the decision to remove the RPi from the casing for the final design and tests, more space was available in the casing. By making proper use of this available space, the RPi extension kit PCB could be implemented in the casing. Regarding the future, a larger casing would not be needed if the PCBs would be designed and manufactured with surface mount (SMD) components. This would significantly decrease the amount of space needed inside the casing, and could probably even decrease the casing dimensions.
Future PCB Creation
To take into account the specified weight and dimensions that were desired in the crude approximation of the final design, a solution has to be proposed to meet these values. By drawing a PCB in advance with proper software (Eagle, CADstar, Altium), the dimensions of the PCB could be optimized with respect to the casing. Also mounting holes could be placed on favorable locations. For the PCB it is important to make use of SMD components to reduce unnecessarily high costs and to minimize the space of the PCBs. Through-Hole components would have extra exposed connections on the bottom of the board that could interfere with other electronics and the casing, and are therefore not preferred. When one decreases the space of the PCBs and the dimensions of the components, this would generally also lead to a decreased weight of the final product.
Communication with Software Team
Since the hardware and software have to be coherent with each other, it is important to have proper communication between the hardware and software teams. Since this communication was not entirely optimal in the beginning of the project, the hardware could not be fitted properly inside of the casing. Problems arose with the pin definitions in the software, because for example all read and write pins were assigned to one side of the RPi connector pins. Now, since the read cell is located on the left and the write cell is located on the right, this would result in cables being crossed over each other in the entire design. This is of course not the optimal method to tackle the implementation. Eventually with proper structural feedback from the hardware team and bright inputs from the software team, the pin definitions were optimized according to the placement of the PCBs in the casing.