Electric Tri-Rear Quadcycle v1

Problem: The breadboarded circuit has grown beyond the available space for on-bike drive testing. Component count and package sizes now exceed the volume of the current enclosures.

Expectation: Repackaging onto a compact protoboard should significantly reduce footprint and wiring bulk, improving fit within the existing housings.

Fallback: If it still doesn’t fit, consolidate into a single large enclosure and split the model into A1-printable sections. Use fast, strong joints (e.g., tongue-and-groove or dovetails with heat-set inserts + M3 bolts, or snap tabs with a cross-bolt) to connect the pieces.

I'm using my school's fabrication studio to make the motor mount, but since it is summer, the studio is technically closed, so my access to fabrication tools is greatly limited. As a result, the time it is taking to finish the motor mount has been significantly extended. It should have taken 2 days or less to finish the mount, now it seems more like 2 weeks. 

3D scanning the bike required numerous attempts and great patience. 3D scanning is a practice makes perfect sort of thing. The quality of a scan is closely associated with experience.

Problem: Prototyping a custom 3-phase rectifier took longer than planned; with the semester starting, I grabbed off-the-shelf units as a stopgap—but they’re too bulky for the current layout and packaging.

Temporary solution: Pause this feature and revisit it in eTRQC v2, where I’ll design a roomier mount and refine a more compact in-house rectifier.

Problem: My past designs used a single supply, so I had to rethink power switching for two rails—36 V (traction) and 3.7 V (peripherals)—and how to fully shut off both at the same time.

Solution: Use high-side P-channel switches: two IRF4905s, each driven by an NPN BJT (2N5551 for the 36 V rail, 2N2222 for the 3.7 V rail). The BJTs’ bases tie to the power button’s “work” pin, so one control signal drives both MOSFET gates, giving a simultaneous power cut to both batteries.

Problem: The controller’s STOP input is active-low and expects open-collector behavior; driving it HIGH backfeeds the line and causes runaway spin.
Solution: Use a 2N2222 in open-collector mode—LOW (sink) asserts STOP; released (floating) disables it for normal run.

Problem: I bought a 12–48 V e-scooter throttle before I fully understood the motor controller I/O. The controller expects a 5 V speed input, but I assumed the throttle would map cleanly to 0–5 V. I then tried to “fix” it with the controller’s speed-limit trimmer (wrong tool), which only clamps top speed—not scale/offset. Result: at the trimmer’s min, the throttle spans 0.00–0.75 V (full twist); at max, it jumps to ~4.65–4.67 V—basically unusable with the throttle.

Solution: Add an MCP6001 rail-to-rail op-amp with a 100 kΩ trim pot to set the gain/offset so the throttle delivers a clean, controllable ~0.1–4.9 V to the speed pin.