For this project, we took extremely heavy inspiration from previous Engineering 4 Go-Kart projects. We talked to some of the previous students, and identified the priorities in our design. Our philosophy was to create a robust and reliable go-kart frame that was relatively simple to fabricate. One of the major challenges that we foresaw from other years was that extremely complex geometries are difficult to assemble, and actually may detriment the reliability of the cart. Additionally, the availability of thick steel tubing allowed us to use simple geometry to create a strong frame capable of safely carrying our weight.

All three of our members, Charlie, Hunter, and Tyler, have extensive experience in FTC robotics. From this experience has arisen a prioritization of modularity for iteration. This is reflected in our final designs, which utilize numerous 3D-printed parts and separately designed subsystems that combine to create a robust cart that is also easily iterated upon. 

The priority of our CAD design was to create an easily fabricated, robust, and reliable go-kart frame that could be constructed out of 1" diameter tubing, wooden panels, and supplementary 3D-prints where necessary. 

One of the major goals for this project was to create a reliable, robust, and well-documented custom Electronic Speed Controller (ESC) for the go-kart. Previously, other years have had significant issues with the reliability and troubleshooting of the Commercial Off the Shelf (COTS) ESC, and that is a problem that we wanted to eliminate this year. 

This was the primary aspect of the go-kart that I worked on. This aspect of the project was incredibly challenging, as I only just started with electrical projects over this previous summer (summer 2025) and did not understand the fundamental principles behind a BLDC ESC. Additionally, the high current requirements for our motor (40A max) created strains on design. However, all these challenges culminated in an incredibly educational and rewarding experience. I've learned the basics of electrical design, proper iteration strategies, the fundamentals of BLDC operations, and have expanded my introductory knowledge of electrical engineering. 

Ultimately, I moved too quickly to design a PCB. From my first few PCB designs, it was clear that I wasn't ready to take on the task of creating such a complex PCB from scratch. The first lesson that I learned from this round of iteration was the consideration of back-EMFs in the system. High current and high voltage lines had the potential to create disruptive electromagnetic fields that could interfere with logic lines. Second, I learned the importance of compact PCB design that was well thought out prior to laying traces. I created a convenient pattern that wasn't incredibly thought out prior to tracing, which created a messy and impractical design. 

Both of these designs still have traces too small to accommodate the 40A required of this ESC. However, the ESC PCB V1.1 has traces capable of accommodating 10A, which is more than enough for preliminary testing with a small BLDC. 

Neither of these PCBs utilize the easy to use, standard DevKit style ESP32s. The lack of the DevKit style boards creates an issue with interfacing and code iteration, something that would not be easy to deal with. 

While I would not learn this lesson till the third iteration, the lack of half-bridge drivers would be a fatal issue in the ESC. Simply, the ESP32 does not have the PWM precision to allow for direct actuation of the MOSFETs, especially not with intermediary MOSFETs.

Additionally, the lack of capacitors would mean voltage spikes, and that the BLDC would not get the necessary current to achieve significant torque on startup, something incredibly important in creating a go-kart. 

Finally, I learned the importance of separating high-voltage high-current lines from logic lines. This eliminates many of the risks associated with back-EMFs. 

Overall, the schematic design for this iteration of the ESC is relatively similar to the previous one. It incorporates the same fatal flaw of direct actuation from the ESP32, and has the same intermediary MOSFETs. However, there were some marginal improvements to the design, such as the incorporation of additional control systems. Specifically, this design uses toggle switches in combination with servos to create the potential for some fun features in the future.

Additionally, this schematic layout is now properly organized and easy to navigate, something that was not the case for my previous designs. It also uses the much needed capacitors, allowing the BLDC to have a steady voltage and the necessary current for start ups. 

This PCB prototype incorporated a more robust power system, with better separation between high current and logic traces, eliminating some of the risk of back-EMFs. Additionally, the layout of the half-bridges is much better thought-out and organized than previously, specifically designed to allow the more robust power separations.

This design again uses the inconvenient and weak DuPont connectors. This was a particular issue in the context of the voltage regulators, which may have needed to draw more current continuously than the DuPont connector was able to provide.

This PCB makes marginally better use of the space available. However, it is still wonky and spacious, and could definitely be easily condensed.

The lack of half-bridge drivers would be a fatal issue in the ESC. Simply, the ESP32 does not have the PWM precision to allow for direct actuation of the MOSFETs, especially not with intermediary MOSFETs.

Another lesson learned was the unfortunate reality of working with DuPont connectors. While convient for prototyping, these are a poor choice for final designs, as they are somewhat unreliable (easily disconnected) and a general pain to work with. In the future, the preference for connection interfaces will be terminal blocks. 

Iteration 4 had radical changes in the overall PCB schematic. First, I switched back to the original gate-drivers, this time with their specifications specifically matched to the motor we were planning to use. Second, I pivoted away from the standard DevKit-C ESP32s, in favor of the NodeMCU ESP32s. This was for no particular reason other than I am more familiar with the NodeMCU. 

This schematic also incorporates all of the necessary capacitors, resistors, and most recently, diodes that will allow the ESC to function as designed. This schematic incorperates the same organizational priniples as the previous, except for the lack of section labels. 

This PCB prototype used many of the lessons learned throughout the iteration process. It utilizes distinct separation between high-current high-voltage traces and logic lines, preventing devastating back-EMFs. It also utilizes a highly compact design, best represented through the control modules for the different half-bridges. Finally, it switches out the unwieldy DuPont connectors for terminal blocks, creating better, more robust interfaces. This is the prototype that was finished at the end of semester, and is the one that is going to be fully prototyped.

One concern over prototyping is the incredibly small size of the components, some being sized less than a milimiter. This should be a relatively small issue, as outsourcing assembly with the PCB manufacture is a relatively small cost that would have great benefit to us. 

• PVC Pipe Cutters

• Tube Notcher

• PVC Notcher

• Hot Glue Gun

• PVC Welder

• Laser Cutter

• Prusa MK4 and MK4S 3D Printers

• Bambu Labs P1S 3D Printer

• Drill and Drill Bits

• PVC Pipe

• PLA (Various Colors)

• 3/8" Bolt

• 3/8" Locknuts

• 1/4" Wood

• 1/8" Wood

• Woodscrews (Various Sizes)

• Steering Rack

• Tie Rods

• Wheels

• CV Axle

• Motor

• Chain (Unknown Type)

• Aluminum Stock

• Seat

• Brake

Fabrication was a relatively straightforward process that generally followed the CAD designs. One major change made during the fabrication process was the removal of the side guards, as we found they had little use or impact, and simply mde the design more difficult to fabricate. Attached below are the different images of the stages of go-kart fabrication. 

The next step of this project was to replicate our PVC pipe go-kart design onto a chassis made out of steel. Many of the fabrication steps were the same, except the hot-glue and PVC welding was now replaced by TIG-welding. 

Unfortunately, we were not able to document much of our steel prototype, as our last few days were spent finishing the build and the electronics (as well as this portfolio). When we return, we will be sure to update the portfolio with all the images of our final prototype. 

• Bandsaw

• Tube Notcher

• Steel Notcher

• TIG Welder

• Laser Cutter

• Prusa MK4 and MK4S 3D Printers

• Bambu Labs P1S 3D Printer

• Drill and Drill Bits

• 1" Steel Pipe (1/8" Wall)

• PLA (Various Colors)

• 3/8" Bolt

• 3/8" Locknuts

• 1/4" Wood

• Steering Rack

• Custom Tie Rods

• Wheels

• CV Axle

• Motor

• Chain (Unknown Type)

• Seat

• Brake Assembly

As of the end of the first semester, we have a functional go-kart (as in, it can perform basic driving functions). We have not yet gotten permission to drive our go-kart out on the field, so our preliminary testing was all conducted within the premises of KRLL, as not to disturb others on campus.