# Troubleshooting These are the main problems we faced while setting up and testing the car, along with what we tried and what fixed them. ## Problem 1 — Faulty USB Drive ### Problem The JetPack installation did not finish when we used the first USB drive. maybe_a_problem ### What We Tried - Tried commands during setup. - Waited about 30 minutes. - Checked the USB drive again. ### Solution We used another USB drive and repeated the JetPack installation. ### Result worked The installation worked with the replacement USB drive. ## Problem 2 — NoMachine Connection ### Problem The laptop had NoMachine 7.8 and the Jetson had version 10. We could not connect, and NoMachine showed “The connection with the server was lost.” photo_no_machine_not_done_#crying ### What We Tried - Tried connecting with the existing versions. - Tried downgrading the Jetson to version 9.8, but the connection still failed. ### Solution We downloaded and installed the same NoMachine version on both devices. ### Result photo_no_machine_done The connection worked, and we could access the Jetson desktop from the laptop. ## Problem 3 — Yaw and Straight Driving ### Problem During track testing, the car kept steering left and right when it should have been driving straight. ### What We Tried - Tested straight driving at low speed. - Adjusted the yaw filter and steering limits. - Checked the yaw readings and steering commands. We found that corrections were calculated faster than new yaw readings arrived. The steering values were also rounded to whole degrees, which made the corrections uneven. ### Solution We added `[STRAIGHT]` to hold one target heading on each straight section. It uses fresh yaw readings, makes small, smooth corrections, and resets its correction history after turns. ### Result We made two versions with the same speeds and turns to compare them: **Without the yaw system:** [slow_without_yaw.py](software/jetson/slow_without_yaw.py) keeps steering centered between turns. https://github.com/user-attachments/assets/1d59722b-05b5-4f70-8cae-8201ca815710 **With the yaw system:** [slow_with_yaw.py](software/jetson/slow_with_yaw.py) uses `[STRAIGHT]` to correct the heading between turns. https://github.com/user-attachments/assets/0a3b3312-4361-43b0-a533-556fb77a1832 Both versions still use the IMU for corner turns. ## Problem 4 — Inconsistent Distance ### Problem The programmed distance was inconsistent. The robot travelled a different distance after each full lap around the field, which includes four corner turns. https://github.com/user-attachments/assets/edc7dd84-3243-4707-87c4-4f8222450dc0 https://github.com/user-attachments/assets/bc55a3ee-09cf-4bc0-b27b-0e08118dfad4 ### Solution We created a separate distance-checker program and repeated the same movement several times. This allowed us to verify the distance values and adjust the calculation until the robot travelled a consistent and accurate distance. ### Result The distance became more reliable and repeatable during full-field runs. https://github.com/user-attachments/assets/b283ef8e-e083-4e8e-adf9-65da4aa7fb71 https://github.com/user-attachments/assets/81d28cd4-cf9e-4bb1-9ad8-c071826f39cc ## Problem 5 — Worn Encoder ### Problem The encoder had been used heavily during testing. Over time, its readings became lower and unreliable, even when the robot travelled the same distance. ao ao1 ### Solution We replaced the worn encoder with a new one and checked its readings using the distance-checker program. ### Result The new encoder produced stable readings and restored accurate distance measurement. af1 af [Recorded tests](testing/) · [Test procedure](testing/PROCEDURE.md) · [Setup and calibration](software/SETUP.md)