About eTPU Resolver Extrapolated Channel Trigger on S32K396

[小小灬li]

Hi.I am using EB to configure the eTPU Resolver on the S32K396.
Currently, I encounter an issue when calling Etpu_Resolver_Ip_GetExtrapolatedOutputs(&Etpu_Resolver_Ip_Instance_I0, &pResolverOutputsExtrapolated);—the extrapolated output values are always zero.

From the code, I noticed that the Extrapolated channel is configured as rising-edge triggered, but in the demo project I cannot find where or how this trigger actually occurs, nor which configuration enables it.

Additionally, the manual states that the Extrapolated thread can be triggered in three ways.
I tried the first method (Host Service Request) by calling Etpu_Resolver_Ip_Sample(&Etpu_Resolver_Ip_Instance_I0);

inside the ATO Notification (on ATO Update 2nd) callback. This indeed triggers the Extrapolated channel, and I can then read valid data using the function above. However, I am not sure if this is the correct or intended approach.

For reference, I configured the project based on S32K3_ETPU_SW_4.7_2.0.0_CD04_D2502_MotorControlExample in EB. My environment details are:

• RTD version: SW32K3_S32M27x_RTD_R21-11_5.0.0

• eTPU version: S32K3_ETPU_SW_4.7_2.0.0_CD04

• Compiler: GHS

Could you please help clarify how the Extrapolated channel is supposed to be triggered in this configuration, and whether my current method aligns with the intended design,Thanks!

[TomasKulig]

Hello Li Zhe,

I have just found the last piece of puzzle. We are missed to configure the input glitch filter for the eTPUA CH 7. i think there is not the RTD/MCAL driver for it so you can use the code below and it will work also on your side.

/* Configure the IGF fort the extrapolation trigger signal routed to the eTPU */

void IGF_Init(void)

{

IP_IGF->MCR[8].MCR = IP_IGF->MCR[8].MCR|IGF_MCR_FGEN(1); // Only enable the filter in bypass mode to enable signal pass from the pad to eTPU channel input

}

Regards,

Tomas.

[TomasKulig]

can you share with me the elf file, please?

When using the Design studio for the compilation can you share the whole project please?

Thnak you,

Regards,

Tomas.

[小小灬li]

Yes, please give me your email address.

[TomasKulig]

i believe that you are properly routing the LCU output to the eTPU channel A7 in the TRGMUX_APP. I think you are missed the IMCR configuration for the eTPU channel A7

/* Assign eTPU input from TRGMUX */

Siul2_Port_Ip_SetInputBuffer(NULL_PTR, 0U, 1, 458, 4U);

This configuration is mentioned in the TRGMUX spreadsheet by note for the orange labeled cells:

[小小灬li]

In addition, based on the same project setup, I added a test in the ATO Notification (on ATO Update 2nd) function to manually write the HSR of the Extrapolated channel. This works correctly. The following image shows the function I added in the interrupt callback.

The following image shows the Extrapolated data that I can observe in the debugger.

If I comment out the call to Etpu_Resolver_Ip_Sample(&Etpu_Resolver_Ip_Instance_I0);, I observe the following state in the debugger, where all Extrapolated values remain zero.

[小小灬li]

I previously configured this register, and after replacing my earlier configuration with the function you provided, the eTPU is still not being triggered.

As shown in the figure below, the trigger signal has a pulse width of 8.4 µs.

Below is my EB configuration where LCU0_LC0_OUT1 is routed to eTPU A7 channel.

This is the TRGMUX configuration I observed in the debugger.

[小小灬li]

Thank you for your previous response. I would like to follow up with an additional question regarding the triggering of the eTPU Resolver Extrapolated.

In my current setup, the PWM trigger signal is first routed to the LCU through TRGMUX. The LCU applies the required delay, and then its output is routed—again through TRGMUX—to both an external IO pin and the eTPU Extrapolated channel.

On the oscilloscope, I can clearly observe the expected signal on the IO pin. However, the eTPU Extrapolated channel still does not get triggered.

Could you please advise what might cause this issue. Is there any additional configuration required for the Extrapolated channel to accept trigger signals from the LCU via TRGMUX.

The two images below show the Resolver channel configuration in EB and the corresponding peripheral initialization code in my project.

[TomasKulig]

[TomasKulig]

Hello Li Zhe,

It depends on your application. I believe that you will control the motor. So you need to trigger the extrapolation in the same time as you are measuring the motor current to have all data necessary for motor control algorithm in the same time (phase currents, Udcbus, position and speed of the motor).

We are using the same trigger as we are using for the BCTU/SARADC phase current measurement for the extrapolation in our motor control. It is generated by the FlexPWM. And routed through LCU and TRGMUX to the eTPU. The LCU is there to do the trigger pulse longer to be visible by the eTPU logic.

You can check the configuration in the S32K396 Single motor control kit SW.S32K396 BLDC/PMSM Motor Control Development Kit | NXP Semiconductors. The code is using the RTD low level driver but the eTPU configuration is closly the same.

Regards,

Tomas.