AM6442: Output clock after reset

Hi Thomas,

There are several clock outputs in the device that can output system clocks for observation/debug, however none of them is available before pin mux setting done by SW. See section Clock Outputs in the device TRM and the Pin Attributes section in the datasheet.

• MCU_SYSCLKOUT0
• MCU_OBSCLK0
• SYSCLKOUT0
• OBSCLK0

If you are currently designing a PCB, then you may consider adding an external clock oscillator IC with multiple outputs, if you need to drive external devices with the same clock.

Regards,

Stan

Hi Thomas,

Please refer below FAQ

Look for Ethernet Interface

https://e2e.ti.com/support/processors-group/processors/f/processors-forum/1318441/faq-am625-am623-am620-q1-am62ax-am62px-am62d-q1-am62l-design-recommendations-commonly-observed-errors-during-custom-board-hardware-design-sk-schematics-updates-for-design-update-note

regards,

Sreenivasa

Thank you Kallikuppa Sreenivasa and Stanislav Stilyanov for your reply

I raised this question because I am implementing Design for Testability (DFT) on my board. Specifically, I wanted to verify whether the crystal oscillator is operational and whether its output frequency is was correct.

That's why I was asking whether a clock is available after reset, so that it can be read via Boundary Scan  without affecting the crystal oscillator.

Is it possible to read the crystale oscillator clock using boundary scan on the AM64 ?

Best Regards,

THOMAS

Hello THOMAS

Thank you.

Thomas Mongellaz said:

Is it possible to read the crystale oscillator clock using boundary scan on the AM64 ?

No.the oscillator pins are not included in the boundary scan chain 

Regards,

Sreenivasa

Hello THOMAS,

Quick note:

It is possible to have the clock output as mentioned above by stan and this will need programming the required output.

You could plan on configuring the clock out and performing some tests as part of the functional test sequence if that is something feasible.

Regards,

Sreenivasa

Hello Kallikuppa Sreenivasa

If we just want to know that MCU_OSC0_XI is oscillating at 25MHz, without programming any FW ,is it a good idea to check the signal PORz_OUT after the release of MCU_PORz , which should be release in the time RST26 at least  1840 ns? 

Does this time "1840ns" linked  with MCU_OSC0_XI oscillation ? What happen about this time if no oscillations on MCU_OSC0_XI?

Thanks for your help.

Hello Thomas,

Thierry HAUSER said:

If we just want to know that MCU_OSC0_XI is oscillating at 25MHz, without programming any FW ,is it a good idea to check the signal PORz_OUT after the release of MCU_PORz , which should be release in the time RST26 at least  1840 ns? 

The PORz_OUT is a buffered output of the MCU_PORz. Not sure if this indicates the 25M clock behavior.

The RESETSTATz is asynchronous reset, we have the min time not max time, and not sure if that would indicate the 25M functioning.

I am reviewing the inputs.

Are you not interested in knowing if 25M works or not?

Regards,

Sreenivasa

Hello Kallikuppa Sreenivasa,

Yes exactly , I would like to know if 25MHz is working or not, with the less as possible complex things to do, that's why I expected that monitoring PORz_OUT was a good idea.

Could you please check on your side?

Thanks

Hello Thierry Hauser,

Thank you.

I understand in case the 25M clock is not working, the processor is not expected to start. (will not start)

The reset status output can be used as indication to check 25M working.

RESETSTATz indicates the propagation of reset to all internal circuits. 

Regarding the PORz_OUT i need to check.

Do you have any concerns using the reset output?

Regards,

Sreenivasa

Hello Thierry Hauser,

I checked with the expert and the expert recommended to monitor the RESETSTATz pin to confirm if the 25M clock is working.

I understand from the expert the PORz_OUT does not have dependency on the 25M clock.

Regards,

Sreenivasa