IWR6843ISK: MSS_ESM got HVMODE_ERR(ESM Group1 - 34) in running.

Hello Takeo,

What do you mean this doesn't occur in normal operation? Under what conditions do you see this error?

From table 3-15 in the TRM, this error is related to the IO supply. Are you moving or touching or otherwise disturbing the board when you see this issue? What is the communication with the external device doing when this error occurs?

Regards,

Jackson

Hello Jackson,

Thank you for reply.

It may occur one hour after the start of sensing, or it may not occur for more than five hours.
Of course, the operation is the same, I have not touched the board.
Therefore, I don't understand the occurrence conditions.
We are currently working to assess the situation with our hardware developer.

Communication with external device
・UART => Configuration(frameCfg, chirpCfg, and so on), SensorStart by CLI
・SPI => Communication every 80ms. IWR6843 is spi master, external device is spi slave.

It is hard to believe that this type of communication would cause I/O error.

I also saw this ticket.:
e2e.ti.com/.../awr6843aop-meaning-of-mss-esm-bit34

Regards,
Takeo

Hello Takeo,

This does indeed indicate an issue with the IO supply. I am trying to get more details on the specific cause of occurrence. Does the device crash with this error or does anything else on the board go wrong?

Thanks,

Jackson

Hello Jackson,

Nothing appears to be going wrong.CLI(UART) is fine, SPI is fine too.

The ERROR OUT signal had dropped to Low, so I read the ESMSR register and it was HVMODE_ERR.
Restarting the system resolves this error.

RFLdoBypassCfg is follow. (3.3V IO Supply)

rlRfLdoBypassCfg_t gRFLdoBypassCfg =
{
    .ldoBypassEnable   = 0,
    .supplyMonIrDrop   = 0,
    .ioSupplyIndicator = 0, /* 3.3 V IO supply */
};

int32_t MmwDemo_mssDataPathConfig(void)
{
    /* Has the mmWave module been opened? */
    if (gMmwMssMCB.isMMWaveOpen == false)
    {
        retVal = rlRfSetLdoBypassConfig(RL_DEVICE_MAP_INTERNAL_BSS, (rlRfLdoBypassCfg_t*)&gRFLdoBypassCfg);
        if(retVal != 0)
        {
            System_printf("Error: rlRfSetLdoBypassConfig retVal=%d\n", retVal);
            return -1;
        }
        :
        :
    }
    :
    :
}

Regards,
Takeo

Hello Takeo,

I think this monitor is mostly intended for startup behavior, and the trigger is likely less than a critical condition. I am still checking on the specific trigger mechanism for this ESM bit, but if the device continues to function normally and after clearing the bit the error does not come back (other than the infrequent method above) it should be ok to log and continue. Please give me until Wednesday to respond with more info.

Regards,

Jackson

Hello Jackson,

Thank you for your suggestion. We look forward to hearing from you.

Regards,

Takeo

Hello Takeo,

I am sorry, it is taking longer for this information. Please give me until next Tuesday to respond. But I think you are not blocked right now in your development from your responses. Hopefully that is ok.

Regards,

Jackson

Hello Takeo,

Sorry for the delay. This error indicates that there is a mismatch between the IO voltages in some areas. So something is seeing a 3.3V as 1.8V, likely momentarily. The error is mostly made for startup measurements, so seems likely it is some momentary glitch. 

Can you please confirm again that even when the error appears, the device continues working while it is active? Or does the error clear itself after a short time?

If you are still seeing the error, when you see it can you also read bit 15 from memory address 0xFF201514 to see what voltage is latched when the error occurs?

Regards,

Jackson

Hi Jackson,

Thank you for your information.

We have reviewed wiring, then we are retesting.Since then,the error no longer occurs.
We'll see how it goes in a bit.

> If you are still seeing the error, when you see it can you also read bit 15 from memory address 0xFF201514 to see what voltage is latched when the error occurs?
OK. I'll try it if error occurs.

Regards,
Takeo