Other Parts Discussed in Thread: AM3352
Tool/software:
Due to issues encountered with the previous thread, we have created a new one. We apologize for the delay in setting up this new thread.
----Previous question---------
When powering up, the "SWIO" voltage may not be output about once in 1000 times.
I am using a combination of AM3352BZCZ80 and TPS65910A31A1RSLR.
Is there anything that could be causing the startup failure?
Please let me know if there is anything that needs to be checked.
-Other voltages are output normally.
-When the output is NG, other power supplies also start up faster than usual. At this time, the rise slope is distorted.
-Forcing pin 1 of the TPS65910 to low will restore "SWIO".
-Pins 1, 40, 37, 45 of the TPS65910 are connected to the AM3352.
Pin 33 "PWRON" is open.
*I am not fluent in English so I am using a translation. Please excuse any strange sentences.
Thank you in advance for your consideration.
---Answer--Field Tolson ------
-
- Are any interrupts status thrown when this occurs?
As mentioned I'll try and dive into some of the documentation and other issues to see if I can find something similar in the meantime.
----------
>・Do you notice this on other/multiple devices?
>・Do you have multiple boards that experience this behaviour?
>・I think an A B A swap could be helpful, if not already done.
I have two units, each with a different lot of TPS65910 installed, but the same problem occurs.
>・Could you provide any captures of this behaviour?
> When the output is NG, other power supplies also start up faster than usual
Capture is not ready. The occurrence frequency is low and the waveform cannot be captured well.
I'll try a little harder.
>・Is above statement pertaining to input or output?
It is the output power supply that has a quick rise time and distorts the waveform.
The input power supply rises in the same way every time.
However, it seems that the problem is more likely to occur when the power is turned ON after leaving it for a few minutes, rather than when it is turned OFF/ON repeatedly.
>Are any interrupts status thrown when this occurs?
It is unclear whether the interrupt was activated at the time the problem occurred, but when I checked after the problem occurred, it was Hi.
> As mentioned I'll try and dive into some of the documentation and other issues to see if I can find something similar in the meantime.
We apologize for the inconvenience, but please continue to check.
-------Current state-------
I haven't found the answer yet.
My guess is that in rare cases, the PWRHOLD signal gets asserted too early after VCC7 is applied, causing a system failure.
I suspect the TPS65910 is failing to initialize.
