LSM6DSV16BX: Does MEMS Studio route MLC to INT1 by default in the generated .h file?

[RotemShechterN]

Hi everyone,

I am developing a custom board using an STM32U5 and the LSM6DSV16BX. I am using MEMS Studio to train and configure the Machine Learning Core (MLC), but I noticed that the generated .h configuration file seems to set INT1_MLC1 bit at MLC_INT1 register by default.

Even though I haven't explicitly set up physical pin routing in my project, the exported C array includes bank-switching logic that forces this configuration. I see the following sequence in my generated header:

    { .type = MEMS_CONF_OP_TYPE_WRITE, .address = 0x01, .data = 0x80 }, // Switch to Embedded Bank
    // ... [other configs] ...
    { .type = MEMS_CONF_OP_TYPE_WRITE, .address = 0x0D, .data = 0x01 }, // MLC_INT1: Route MLC1 to INT1
    // ... [other configs] ...
    { .type = MEMS_CONF_OP_TYPE_WRITE, .address = 0x01, .data = 0x00 }, // Return to Main Bank

Because of this, INT1 pin is firing on MLC state changes. Since this is a custom design, I need strict control over my interrupt routing.

My questions are:

1. Can someone confirm if routing the MLC to INT1 is indeed the default behavior of MEMS Studio?

2. If it is, is there an option or setting within the MEMS Studio GUI to change or disable this routing before exporting the .h file?

Thanks in advance for the help!

Best regards,

Rotem

[Federica Bossi · Accepted solution]

Hi @RotemShechterN ,

Here our answers:

1. Can someone confirm if routing the MLC to INT1 is indeed the default behavior of MEMS Studio?
   Yes this is the default behavior of MEMS Studio
2. If it is, is there an option or setting within the MEMS Studio GUI to change or disable this routing before exporting the .h file?
   There is no way to generate header file without routing the MLC to INT1 in current version of MLC tool, we will provide this functionality in next versions of the tool.