In my design i use two CAN-FD-modules MCP2518FD. wireing seems to be correct according to documentation. I measure while communication on CANH 0V to GND → no communication - should be ~3V. Claude Code analysed the schematic an told: ATA6561 STBY = HIGH = Standby-Modus.
I should disconnect the STBY from 3.3V to GND.
…really?
GPIO 11 → SDI (ESP32-S3-DEV_Kit)
GPIO 12 → SCK (ESP32-S3-DEV_Kit)
GPIO 13 → SDO (ESP32-S3-DEV_Kit)
GPIO 10 → NCS (ESP32-S3-DEV_Kit)
GPIO 15 → SDI (ESP32-S3-LCD-1.28" waveshare display)
GPIO 14 → SCK (ESP32-S3-LCD-1.28" waveshare display)
GPIO 13 → SDO (ESP32-S3-LCD-1.28" waveshare display)
GPIO 16 → NCS (ESP32-S3-LCD-1.28" waveshare display)
CANH ↔ CANH
CANL ↔ CANL
GND ↔ GND
3.3V ↔ 3.3V
JP1 closed → 120 Ohm ist done on both sides.
Symptoms
SPI initialization succeeds on both nodes (begin() returns 0, no error)
Internal loopback test passes: sending a frame with CANFD_NO_BIT_RATE_SWITCH in InternalLoopBack mode is received correctly
No physical CAN communication possible: TX FIFO fills up immediately (no ACK from bus), no frames received
CANH measures 0V against GND — even at idle, where it should read ~2.5V (recessive state)
Setup
We are using two of these modules in a CAN bus system:
Node 1 — bridge_ESP32 (ESP32-S3): MCP2518FD connected via SPI (FSPI, 40 MHz crystal), polling mode (INT=255), 125 kbit/s, NormalFD. Library: ACAN2517FD v2.1.16.
Node 2 — horizon (ESP32 + TFT display instrument): Same module, same library, same settings, CAN_TX_ID=0x280, CAN_RX_ID=0x200.
Both nodes connected: CANH↔CANH, CANL↔CANL, GND coupled. JP1 closed on both (120Ω termination active).
Any Ideas? is the breakout board really to be modified?
Thanks 4 help
Matthias
Hello Matthias,
Thank you for the detailed description!
I believe that the problem you are facing is due to the transciever being powered via 3.3V instead of 5V. You must bring 5V to the VCC lines, as the ATA6561 chip only operates on that logic level. The SPI lines themselves have an onboard voltage regulator, so you can connect them to the ESP32 devices without a logic level converter, it is just the VCC line that needs to be powered with 5V.
Please try it out and get back to us with the results!
Hello Josip,
with 5V on VCC I measure 2.6V on CANH - seems working.
When I use 5V i get issues with GND. maybe the reason I still get no communication running. At weekend I will try to fix it.
You should change the description on your website according to the operation voltage to be constistent: “Operating voltage: both 2.7V and 5V” @ Welcome | Soldered Documentation and add the onboard voltage regulation for the communication pins.
ESP32-S3-LCD-1.28" waveshare display and ESP32-S3-DEV_Kit - both have no 5V output Pin to catch the 5V from the boards. …i will see @ weekend.
Hi together,
major topic in communication problems was:
(1) MCP2518FD needs VCC = 5V. With VCC = 3.3V no communication.
(2a) 1st MCP2518FD needs common GND with ESP32-S3-DEV_Kit (1st communication partner) AND
(2b) 2nd MCP2518FD needs common GND with ESP32-S3-LCD-1.28" waveshare display (2nd communication partner) AND
(2c) both MCP2518FD needs a common GND (with too much GND-drift no communication).
[with galvanically isolated USB power supply i got 2.6V GND-drift between 1st and 2nd MCP2518FD - seems too much. With same GND: no problems]
=> communication runs with ESP32, but with higher electrical efford, than expected.
hope it helps
Matthias