Unique I2C address for top/bottom variant #3

gemenerik · 2025-11-06T15:48:50Z

The deck's I2C address is selected by the Crazyflie via the I2C_ADDR pin (PC15). When PC15 is pulled high, the deck uses address 0x31; when low, it defaults to 0x30. It required a sizeable 100ms boot delay to ensure PC15 is stable before reading it.

PA12 (ADDR_SELECT) is used for LED position detection. The deck tests this pin at startup using pull-up/pull-down states to distinguish between floating (no LED), driven-low (bottom), and driven-high (top) states.

The I2C protocol was extended from v1 to v2 with two new commands:

CMD_GET_LED_POSITION: Reports whether the deck is mounted on top or bottom based on PA12 detection
CMD_GET_LED_CURRENT: Reports per-channel current readings (R, G, B, W) in milliamps. ADC channels 4-7 are sampled in round-robin during the main loop and converted using sense resistor values (3.4Ω for red, 1.0Ω for others)

Requires:

Add simultaneous top and bottom Color LED deck support crazyflie-firmware#1561

Detection is based on LED pulling ADDR_SELECT (PA12) high / low

Instead of auto-detecting deck variant via PA12, the Crazyflie now pulls PC15 (I2C_ADDR) high to select address 0x11. Added 500ms boot delay to ensure the Crazyflie has set PC15 before the deck reads it. PA12 (ADDR_SELECT) remains available for future I2C commands.

Copilot

Pull Request Overview

This PR implements dynamic I2C address configuration for the color-led firmware, allowing the device's I2C slave address to be selected at runtime based on GPIO pin states. The base I2C address is 0x10 (OwnAddress1 = 0x20 in 8-bit format), with the option to increment to 0x11 (OwnAddress1 = 0x22) based on pin PC15.

Key changes:

Added runtime I2C address selection based on PC15 (I2C_ADDR) pin state
Introduced PA12 as ADDR_SELECT input pin with pulldown (currently unused in code but configured)
Modified GPIO initialization to support two new input pins for address configuration

Reviewed Changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 2 comments.

File	Description
color-led.ioc	Updated STM32CubeMX configuration to add PC15 and PA12 as GPIO input pins and changed PA12 from output to input mode
Core/Src/main.c	Implemented dynamic I2C address selection logic reading PC15 after 500ms delay, and updated GPIO initialization for new pin configurations
Core/Inc/main.h	Added pin definitions for I2C_ADDR (PC15) and ADDR_SELECT (PA12)

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Core/Src/main.c

Extends I2C protocol (v1→v2) with two new commands: - CMD_GET_LED_POSITION: Reports top/bottom mounting via PA12 detection - CMD_GET_LED_CURRENT: Reports per-channel current readings (R,G,B,W) LED position is detected once at startup using pull-up/pull-down testing to distinguish between floating (no LED), driven-low (bottom), and driven-high (top) states. Current monitoring reads ADC channels 4-7 in round-robin fashion during main loop and converts to milliamps using appropriate sense resistor values (3.4Ω for red, 1.0Ω for others).

…ucture

Copilot

Pull Request Overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 3 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Core/Src/main.c

gemenerik · 2025-11-13T08:39:44Z

I2C address changing not working consistently

…shing instructions in the development guide

Makes the flashing process more transparent

Refactor driver from single-deck to dual-deck architecture with context-based state management. Previously the driver only supported one deck at a time. Now both variants can operate simultaneously with separate state, parameters, and logging. Driver changes: - Context structure holds per-instance state (init, colors, thermal, current, position) - Two deck drivers: bcColorLEDBot (PID 0x13) and bcColorLEDTop (PID 0x14) - I2C addressing: bottom at 0x30, top at 0x31 (GPIO 11 controls address selection for top) - Separate bootloader reset functions for each deck (both use DFU address 0x64) - LED controller callback now forwards to all initialized instances Parameter and log group changes: - Renamed from colorled to colorLedBot and colorLedTop for separate control - brightnessCorr shortened to brightCorr (firmware name length limits) - Detection parameters: deck.bcColorLedBot and deck.bcColorLedTop - New log variables: ledPos, ledCurR, ledCurG, ledCurB, ledCurW Protocol enhancements (v1 → v2): - Add CMD_GET_LED_POSITION and CMD_GET_LED_CURRENT commands - Increase RXBUFFERSIZE to 9 bytes for current readings (4 channels × 2 bytes) - Task loop refactored with dedicated polling functions (thermal: 100ms, current: 1000ms) - LED position polled once during initialization (hardware-fixed value) Note: LED current readings are PWM-based and only provide stable values at full intensity (100% duty cycle). This feature and position detection are primarily for production testing. Luminance correction improvements: - Red LED value updated to 139 lumens (from 90) accounting for actual circuit current - Red's lower forward voltage (2.1V vs 2.9V) results in 1.54× higher current despite higher sense resistor, requiring luminance adjustment GPIO initialization now includes proper error handling with debug messages. Example app updated to auto-detect and select appropriate deck parameter group. Requires: - bitcraze/color-led-deck-firmware#3 - bitcraze/deck-ctrl-firmware#4

…loper note for STM32 option bytes

Copilot

Pull Request Overview

Copilot reviewed 6 out of 6 changed files in this pull request and generated 8 comments.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Core/Src/main.c

Core/Inc/main.h

cmake/gcc-arm-none-eabi.cmake

Core/Src/main.c

Refactor driver from single-deck to dual-deck architecture with context-based state management. Previously the driver only supported one deck at a time. Now both variants can operate simultaneously with separate state, parameters, and logging. Driver changes: - Context structure holds per-instance state (init, colors, thermal, current, position) - Two deck drivers: bcColorLEDBot (PID 0x13) and bcColorLEDTop (PID 0x14) - I2C addressing: bottom at 0x30, top at 0x31 (GPIO 11 controls address selection for top) - Separate bootloader reset functions for each deck (both use DFU address 0x64) - LED controller callback now forwards to all initialized instances Parameter and log group changes: - Renamed from colorled to colorLedBot and colorLedTop for separate control - brightnessCorr shortened to brightCorr (firmware name length limits) - Detection parameters: deck.bcColorLedBot and deck.bcColorLedTop - New log variables: ledPos, ledCurR, ledCurG, ledCurB, ledCurW Protocol enhancements (v1 → v2): - Add CMD_GET_LED_POSITION and CMD_GET_LED_CURRENT commands - Increase RXBUFFERSIZE to 9 bytes for current readings (4 channels × 2 bytes) - Task loop refactored with dedicated polling functions (thermal: 100ms, current: 1000ms) - LED position polled once during initialization (hardware-fixed value) Note: LED current readings are PWM-based and only provide stable values at full intensity (100% duty cycle). This feature and position detection are primarily for production testing. Luminance correction improvements: - Red LED value updated to 139 lumens (from 90) accounting for actual circuit current - Red's lower forward voltage (2.1V vs 2.9V) results in 1.54× higher current despite higher sense resistor, requiring luminance adjustment GPIO initialization now includes proper error handling with debug messages. Example app updated to auto-detect and select appropriate deck parameter group. Requires: - bitcraze/color-led-deck-firmware#3 - bitcraze/deck-ctrl-firmware#4

Dynamically detect top/bottom variant and adjust I2C address accordingly

0dc57c8

Detection is based on LED pulling ADDR_SELECT (PA12) high / low

gemenerik self-assigned this Nov 6, 2025

gemenerik mentioned this pull request Nov 6, 2025

Support for Top and Bottom Color LED deck variants bitcraze/crazyflie-firmware#1553

Closed

gemenerik marked this pull request as ready for review November 10, 2025 14:05

gemenerik requested a review from Copilot November 10, 2025 14:05

Copilot AI reviewed Nov 10, 2025

View reviewed changes

Core/Src/main.c Outdated Show resolved Hide resolved

Core/Src/main.c Show resolved Hide resolved

gemenerik changed the title ~~Detect top/bottom variant and adjust I2C address accordingly~~ Unique I2C address for top/bottom variant Nov 10, 2025

gemenerik added 3 commits November 10, 2025 15:10

Move i2c_address def inside main

4a52b0a

Update compiler flags for debugging and enhance development guide str…

63cf66d

…ucture

gemenerik requested review from ArisMorgens and Copilot November 11, 2025 14:22

Copilot started reviewing on behalf of gemenerik November 11, 2025 14:23 View session

Merge branch 'master' into rik/dynamic_i2c

4602c78

Copilot finished reviewing on behalf of gemenerik November 11, 2025 14:24

Copilot AI reviewed Nov 11, 2025

View reviewed changes

Core/Src/main.c Outdated Show resolved Hide resolved

Core/Src/main.c Outdated Show resolved Hide resolved

Core/Src/main.c Outdated Show resolved Hide resolved

gemenerik marked this pull request as draft November 13, 2025 08:39

gemenerik force-pushed the rik/dynamic_i2c branch 3 times, most recently from b6d33aa to 5339d56 Compare November 13, 2025 15:13

Update I2C base address and adjust address selection logic

1eb319c

gemenerik force-pushed the rik/dynamic_i2c branch from 5339d56 to 1eb319c Compare November 13, 2025 15:18

gemenerik added 2 commits November 19, 2025 13:42

Add custom target for flashing firmware using cfloader and update fla…

388e4d8

…shing instructions in the development guide

Document direct cfloader command instead of CMake wrapper target

d5aa7d4

Makes the flashing process more transparent

gemenerik mentioned this pull request Nov 20, 2025

Add simultaneous top and bottom Color LED deck support bitcraze/crazyflie-firmware#1561

Merged

Update flashing instructions to include radio parameters and add deve…

c5b4b38

…loper note for STM32 option bytes

gemenerik marked this pull request as ready for review November 20, 2025 13:59

gemenerik requested review from Copilot and evoggy November 20, 2025 13:59

Copilot started reviewing on behalf of gemenerik November 20, 2025 14:00 View session

Copilot finished reviewing on behalf of gemenerik November 20, 2025 14:03

Copilot AI reviewed Nov 20, 2025

View reviewed changes

gemenerik and others added 4 commits November 24, 2025 10:45

Improve led detection function docstring

162cdd9

Correct I2C address switch comment

0d4c03d

Correct pulled LOW->HIGH comment

fe17052

Remove redundant default case handling in ADC channel index update

0127d61

gemenerik requested review from tobbeanton and removed request for ArisMorgens and evoggy November 24, 2025 13:21

Update flashing instructions to use correct target names

fe6e58d

evoggy previously approved these changes Nov 24, 2025

View reviewed changes

Core/Inc/main.h Show resolved Hide resolved

cmake/gcc-arm-none-eabi.cmake Show resolved Hide resolved

evoggy reviewed Nov 24, 2025

View reviewed changes

Core/Src/main.c Outdated Show resolved Hide resolved

Lower ADC reading frequency to reduce overhead in LED current monitoring

1800157

gemenerik dismissed evoggy’s stale review via 1800157 November 24, 2025 14:50

Improve loop timing consistency by implementing a 1ms delay mechanism

c6fc356

gemenerik merged commit f30f355 into master Nov 24, 2025

gemenerik deleted the rik/dynamic_i2c branch November 24, 2025 15:56

Unique I2C address for top/bottom variant #3

Unique I2C address for top/bottom variant #3

Uh oh!

Conversation

gemenerik commented Nov 6, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

gemenerik commented Nov 13, 2025

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

gemenerik commented Nov 6, 2025 •

edited

Loading