Hi, I am trying to send classical CAN messages using FDCAN with STM32h723vgT6 - based board. I am however having difficulties sending any packets (and receiving).
Hardware setup:
- DM-MC02 board, based on STM32h723vg (https://gitee.com/kit-miao/dm-mc02)
- SIT1042 CAN controller, directly built-in the board. Thus, the respective CAN_H and CAN_L are directly accessible through the board's pin
- (Device) ZDT_Emm_V5.0 stepmotor controller. Anyway it is just the device I am trying to send data to (https://blog.csdn.net/zhangdatou666/article/details/132644047)
- CANable 2.0 - inspired USB device.
What I am trying to do:
I want to be able to send and receive classical CAN messages using the DM-MC02 board to drive my controller. Because of my controller's specs, I need to send classical CAN messages with extended IDs at a bitrate of 500 000, by configuring FDCAN (what's supported by STM32h723vg) in normal mode. I would prefer using the CAN1 port of my board.
I am using STM32CubeIDE.
What works:
I am able to drive the stepmotor controller with my computer by using python-can through the CAN USB device, which indicates that for the most part my wiring seems okay, and that I do have a somewhat working CAN network. I did check the termination resistor values, all my CAN_H and CAN_L cables are twisted together.
What does not work:
My setup is rather simple, the CAN network is made of the DMMC board, CAN USB drive and the Stepmotor controller. For now, I wanted to send data from the board to the CAN USB (and see the messages via python-can or Cangaroo). However, I have no signal from my board.
Basically I am unsure what is missing in my board's configuration or code. I have followed many guides to configure the .ioc, I even found some code by the manufacturer (not the settings I would like to use though, it is for FDCAN 1 250 000, which I am unsure what to think of by the way as, at least for python-can, 1 250 000 seem to be not a supported rate, thus I could not test with it much). For the timing configuration, I have used Kwaser's calculator for FDCAN, I found this configuration (from my understanding, the Data should not matter much with classical CAN) :
Please find attached my .ioc file and source code, and here is the core of the code.
if(HAL_FDCAN_Start(&hfdcan1) != HAL_OK)
{
Error_Handler();
}
/* Configure Header */
uint8_t TxData[] = {0xF6, 0x00, 0x05, 0xDC, 0x0A, 0x00, 0x6B, 0x00};
TxHeader.Identifier = 0x0100;
TxHeader.IdType = FDCAN_EXTENDED_ID;
TxHeader.TxFrameType = FDCAN_DATA_FRAME;
TxHeader.DataLength = FDCAN_DLC_BYTES_8;
TxHeader.ErrorStateIndicator = FDCAN_ESI_ACTIVE;
TxHeader.BitRateSwitch = FDCAN_BRS_OFF;
TxHeader.FDFormat = FDCAN_CLASSIC_CAN;
TxHeader.TxEventFifoControl = FDCAN_NO_TX_EVENTS;
TxHeader.MessageMarker = 0x0; // Ignore because FDCAN_NO_TX_EVENTS
/*...*/
/* Add message to Tx buffer */
if(HAL_FDCAN_AddMessageToTxBuffer(&hfdcan1, &TxHeader, TxData, FDCAN_TX_BUFFER0) != HAL_OK)
{
Error_Handler();
}
/* Send Tx buffer message */
if(HAL_FDCAN_EnableTxBufferRequest(&hfdcan1, FDCAN_TX_BUFFER0) != HAL_OK)
{
Error_Handler();
}
/* Polling for transmission complete on buffer index 0 */
while(HAL_FDCAN_IsTxBufferMessagePending(&hfdcan1, FDCAN_TX_BUFFER0) == 1) {
__NOP();
}Basically, I found that the code gets stuck in the while(HAL_FDCAN_IsTxBufferMessagePending(&hfdcan1, FDCAN_TX_BUFFER0) == 1) loop, indicating that the buffer has some difficulties getting processed and cleared.
When using "HAL_FDCAN_GetProtocolStatus(&hfdcan1, &status);", I can see that the status struct's DataLastErrorCode and LastErrorCode are set to 7 (No change since last read).
When trying to receive messages with a similar code, the program would also get stuck as if there were no new messages.
As I cannot make that simple setup work I am a bit at lost on what to do, perhaps the STM32 community could have new ideas.
Please feel free to ask for any details, and I would appreciate any help.
Thanks !
Another question,
In the datasheet the transceiver is an 8 pin package:
while in the schematics the transceiver has the pin 9 called "EPAD":
.. not sure if this the same transceiver part number..
Hello, thank you for your suggestions it helped me for debugging and narrowing down the issue !
While I have now a setup that allows to transmit and receive messages from the STM32 (Hooray !), something is still off so I would like not to close the thread yet.
Summary of the issue resolution
Two things:
1. Thank to your suggestions, I found out that the CAN controller was indeed not being powered correctly. Resolving this lead to this new status struct values:
- Last Error Code: 3 Acknowledge error
- Activity: 24
- Error Passive: 1
- Warning: 1
2. This pointed out that there might have been some issue with the CAN configuration itself as well. And indeed there is, but in a surprising way.
New issue
While STM32CubeIDE reports that, according to the clock and CAN config, the Nominal Baud Rate should have been 500 000 bit/s, meaning 500kHz signal frequency, an oscilloscope measurement revealed that it was instead outputting at 480kHz ! With a quick and dirty tweak in the timing values to raise the nominal baudrate to compensate
allowed to reach ~504kHz, allowing CAN communication.
As I find this finding great (something works!), I wanted to ask whether someone would know if something is going out, perhaps it is just the external oscillator that is not accurate...
