As submitted for the final report

submission-final

Added ability to export power comparisons to CSV

v1.1.1 - Latest PCB with improved voltage regulator

The plan is to use these boards for the microcontroller side and use the v1.0.4 boards for the other side. This board should be capable of working on both sides.

Changes

• Upgraded / hopefully fit for purpose voltage regulator.
• Reverse polarity protection.
• RGB LED.
• Swapped accelerometer interrupt pin to be RTC capable.
• Thermal isolation for the temperature sensor.
• Fixed USB polarity so it should hopefully work when plugged in on both sides.
• Set I2C pullup resistors to use the same value as USB CC pins.
• GPIO pins changed for ease of routing.

v1.1.0 - Fixing silkscreen for swapped lines.

No boards of this version were ordered.

Fixed the I2C SDA and SCL lines being swapped. The addressing for each temperature sensor was flipped for each side. The silkscreen on the v1.0.4 boards now has SDA and SCL incorrectly labelled - they are now flipped.

The SPI bus labels for SDI and SDO have been flipped to match up with the correct naming schemes.
Both these issues have been fixed / adapted for in software. The only impact to the v1.0.4 hardware is the I2C pinout in the silkscreen being flipped over.

v1.0.4 - First PCB order

As ordered for the first PCB (set of 10). These boards have some issues on the microcontroller side. It is recommended to use a later iteration for the microcontroller side and use the v1.0.4 boards for the other side.

Changes

• New form factor, uses an ESP32s3 MINI instead of WROOM.
• 4 Layer PCB

Known issues

• The voltage regulator can't supply enough current to run the microcontroller side, especially once WiFi starts. The solution during testing is to add a shottky diode across the regular as suggested in the manual and run the unit from a LiFePo4 AA sized cell.
• I2C and SPI lines are flipped over / mislabelled on the board.
• The ESP32s3 generates a fair bit of heat, which affects the temperature sensor.
• The USB connector has the D+ and D- pins flipped on one side, so the connector only works when plugged in one way.
• The interrupt pin of the IMU isn't connected to a pin capable of being read by the low power processor.
• There is a solder bridge between the designated battery voltage pin and ground on the fully populated test board. The battery voltage was rerouted to another pin.

v0.0.0 - Original outline from the project proposal

PCBOrigDesign

PCB original design