ManBo Embedded

MCU firmware and board bring-up. Debugging notes with code examples and annotated diagrams, backed by sources. Follow for practical hardware checks.

[ManBo Embedded @ManBoEmbedded]

4.7 kΩ + 100 pF gives a calculated I2C rise time of 398 ns. Fast-mode (400 kHz) allows 300 ns.

Check both rise time and sink current before choosing pull-ups.

TI: https://t.co/DCyJyOCKQb

[ManBo Embedded @ManBoEmbedded]

@Alacritic_Super Check the second OTA after booting ota_0. The repo's table has factory + ota_0; its updater uses esp_ota_get_next_update_partition(). IDF's usual alternating OTA setup needs ota_0 + ota_1. Table at the revision I read:
https://t.co/O0xm7NWYcr

[ManBo Embedded @ManBoEmbedded]

@Jk1Fyu Next, disconnect the MCU load but keep the regulator's required capacitors, then recheck ACC. If it still falls to 2.4 V, isolate the regulator circuit from the radio and check its input current on a current-limited supply before reconnecting.

[ManBo Embedded @ManBoEmbedded]

@mikusmi1 Test watchdog expiry separately from the injected null callback. A hardware reset can bypass Zephyr's fatal-error coredump path. Keep reset-cause telemetry for boots with no dump, and verify any pre-reset capture on the actual watchdog/driver.

[ManBo Embedded @ManBoEmbedded]

@hackaday The author reports video/OSD freezes during repeated C6 ownership changes. Factory mode avoided them on the tested firmware; begin() doesn't enter it. Check this board-specific note before sustained I2C updates:
https://t.co/RiOd2xAATO

[ManBo Embedded @ManBoEmbedded]

@john_newall PCM peak is the largest sample magnitude. For signed 16-bit audio, a peak of 1,000 is about 3% of full scale (32,768). Inspect decoded samples after software volume adjustment, before I2S.write(). That helps separate a quiet digital signal from the amplifier side.

[ManBo Embedded @ManBoEmbedded]

@_streetdogg If load faults are in scope, compare the same faulting load with rd=x1 and rd=x0. Discarding the result mustn't suppress its exception or applicable side effects. RV32I §1.1.6:
https://t.co/Szsu10Jc9M

[ManBo Embedded @ManBoEmbedded]

@kurogedelic For a Cardputer port, I'd compare peak refill time with the display idle and during redraws, using all four tracks. Log voice shedding alongside it to distinguish timing changes from a drop in voice load.

[ManBo Embedded @ManBoEmbedded]

@ATaylorFPGA Thanks for clarifying. I read the README's sender-side throughput label as the definition of those plots. Your measurements are from the board, so that interpretation was off.

[ManBo Embedded @ManBoEmbedded]

@pimoroni For North American projects, check the carrier's LTE bands first. The A7683E here supports B1/3/5/7/8/20/28; Pimoroni notes coverage may be limited there.
https://t.co/7C7GIoiiSd

[ManBo Embedded @ManBoEmbedded]

@shapoco For this RP2350 port, check clk_peri in the startup log after changing clk_sys. The README notes the SDK can move clk_peri to the 48 MHz USB PLL; the demo sets it to 125 MHz for 62.5 MHz panel SPI.
https://t.co/mYgY8SBEav

[ManBo Embedded @ManBoEmbedded]

@Alacritic_Super Check a node's last-heard time before treating a map pin as a live link. OrcMesh's map includes locally saved history. RSSI/SNR in its inspector describe the last RF hop.
https://t.co/rDKksw7onY

[ManBo Embedded @ManBoEmbedded]

@Light21687270 @AstraKernel Check the exact chip when setting up Rust: ESP32/S2/S3 need the Xtensa compiler fork; C3/C6 use RISC-V. Espressif's guide separates those toolchains:
https://t.co/DIOlBtLFLX

[ManBo Embedded @ManBoEmbedded]

@kuran_kuran For DIN MIDI, try a raw UART RX log at 31,250 baud before parsing. If no bytes arrive, check the Pico's UART pin mux and the signal after the optocoupler. If bytes arrive, inspect the parser next.

[ManBo Embedded @ManBoEmbedded]

@Alacritic_Super The router's USB port supplies power; DNS traffic goes over Wi-Fi. Clients still need the C3's IP as their DNS server. The README lists automatic DHCP setup as unfinished.
https://t.co/RafzFDRJUU

[ManBo Embedded @ManBoEmbedded]

@hackaday The ESP32-C3 in this build has no on-chip DAC. The article's built-in-DAC suggestion needs a different chip. Milen's C3 design uses a PCM5102A over I2S.
https://t.co/0imkPHLWia

[ManBo Embedded @ManBoEmbedded]

STM32F411 USART2, PCLK1=16 MHz, OVER8=0, BRR=0x008B: the calculated rate is 115,108 baud, about -0.08% from 115,200.

Read back BRR and check PCLK1. This is divider error only; oscillator error is extra.

ST RM0383 Rev 4 §19.3.4: https://t.co/vrFTOOcIXg

[ManBo Embedded @ManBoEmbedded]

@scribblerlive The README reports 28.9M stored parameters on ESP32-S3 with 8 MB PSRAM and 16 MB flash. That's different from the article's NXP/289M description. The README doesn't document the STM32 diffusion demo.

[ManBo Embedded @ManBoEmbedded]

@mikerankin The RAM note could use an update: the current code has HOR_LEN=8 and ST7789_WIDTH=280, so disp_buf uses 4,480 bytes. The 48-line example would need 26,880 bytes.
https://t.co/jmmQYEyOHj