S32G2 LLCE Flexray

[PXH1]

Hello,

I want to implement LLCE Flexray function on S32G2. My versions are S32G_LLCE_1_0_7 and SW32G_RTD_4.4_4.0.2.

The development board S32G-VNP-RDB2 and the self-made board BP0 and BM0 are connected in pairs, and the two boards are shared. Since both boards use the same pins, download the node1 routine to the homeboard and run the node2 routine in debug mode.

It's always running WHILE

The variables for POCStatus are as follows

It's the same phenomenon on both boards. But Flexray shouldn't have two coldstarter nodes to link to.

The return value of the function enabled in the cold start is not E_OK;

We want to know the cause of this phenomenon and the solution?

Thanks

[Celeste_Liu]

Dear PXH1,

Thank you for your question. I will support you on this issue.

FlexRay indeed requires a minimum of 2 nodes to function properly. However, when using a custom board, there are several factors to consider. The return value of the function not being E_OK in the cold start indicates that there might be some configuration or setup issues.

One possible reason could be related to the clock settings. Make sure that the clocks for the FlexRay controller on both the reference board and the custom board are properly configured and synchronized.

Based on the description doc, it seems that the development of this project is based on the EVB board rather than the RDB board, so you need to change some configurations. Pay close attention to any differences in pin assignments, register settings, and initialization sequences. The following are the links for reference:
https://community.nxp.com/t5/S32G/FlexRay-test-of-S32G-VNP-RDB2/td-p/1758084
https://community.nxp.com/t5/S32G/S32G-LLCE-FlexRay/m-p/1828808

In addition, some network configurations are also required. You can refer to the following link:
https://community.nxp.com/t5/S32-Design-Studio/S32G274A-FlexRay-Nodes-Not-Coldstarting/m-p/1858736

I hope these additional suggestions can assist you in resolving the issue.

Best regards,
Celeste

[PXH1]

hello,

I checked my configuration based on your answer,

First of all, I need to correct my question, the return value of the function on cold startup is E_OK, I wrote it wrong at that time, sorry

Second, I went to check both engineering and development boards as well as homemade boards and pins. They do correspond one to one. The clock is also the same as the configuration in the link you provided. The flexray configuration items also match the replies in the link.

In a recent test, we had a new question,

As explained in the chip manual in the figure above, TXEN should be LOW when Flexray was waiting for the link, but we measured and found that it was HIGH, which was due to our misunderstanding or the need for configuration.

[Celeste_Liu]

Dear @PXH1 ,

Thanks for your reply.

According to the relevant introduction in the Reference Manual, as shown in the figure below. I believe your understanding is correct, and TXEN needs to be low.

Best Regards,

Celeste

[PXH1]

hello
However, this pin is reused in the routine as LLCE_FLEXRAY_LLCE_FR_TXE_A_b, not GPIO, so it cannot be pulled up or down directly.

Is this phenomenon something wrong with the routine? Or call some functions too little?

Thanks

[Celeste_Liu]

Dear @PXH1 ,

As can be seen from Figure 500. "FlexRay block diagram" in Chapter 56 of the S32G2 Reference Manual, the TX_EN signal will be inverted. Therefore, it should be correct to measure it as high. I don't think the problem lies in this.

As for the problem you encountered, I am unable to reproduce the issue based on your environment configuration. I am very sorry about that. If there is any progress in the future, please feel free to ask.

Best Regards,

Celeste