TMS320F28023: What is the difference with the automotive grade device

Verna,

The only functional difference in the PTT and the PT-Q1 is the allowed temperature range that the device is specified to operate at.  For the PTT it is 105C, but for the PT-Q1 it is 125C.  Not sure if your start up conditions are in this range of temps or this is not a factor for the system.

The internal device is identical between the 2 devices, so there should not be a functional difference in the logic.

If, on a failing system, you see that a reset solves the issue then I think we would want to look at the goodness of the power rails when XRSn is released.

We would need to understand what behavior the C2000 has when it is failing.  Is XRSn toggling(indicating the watchdog has tripped), do we see a droop in the power rails, etc.  It would be best if we could replicate this with JTAG/CSS connected to determine where the device code is at when this happens.  Else you would need to insert some GPIO toggles in the code to better understand how far into the execution the failing device goes, etc.

All this to say unless we are outside the temperature range of the PTT device, I would not expect any difference in behavior to the Q1 grade device.  Q1 rating gives a higher level of quality needed for automotive grade applications(in addition to the temperature rating). 

Best,
Matthew

HI 

When the board is powered on, but the DLP is not on. Manual reset TMS320F28023PTT XRSTn can be restored to normal,

Manual reset TMS320F28023PTT, tested the power supply VDIO33/VDADC33/VDD18, voltage did not drop. Only MON_RESET will run again,

Manually resetting DLP120_RESET alone does not work. Manually reset the XRSn of the MCU. The program is provided by TI. We have not done development and debugging, so we cannot simulate debug and capture REG through JTAG. The power sequence is shown as follows:

Verna,

I am going to try and loop in the DLP team for comment. 

Just to confirm, on a "T" rated device that fails to boot properly, if you reset the C2000 MCU then the system works correctly?

So the condition only happens on initial power up?

Best,

Matthew

Hello Matt, Verna - 

I do not see anything obvious about the implementation to explain this phenomena. I also expect the two devices to operate the same way.

As a quick sanity check, Verna, have you tested this behavior on multiple systems? 

Regards,

Ross

hi，

Not only initial powering, the C2000 MCU system can be restored to normal after reset every time.

The other question: they have not used the MCU on other systems

Hi Verna -

Thanks for the reply. I would recommend to test this behavior on another PCB (with the same components), so that we have more than one data point.

Regards,

Ross

Hi，

In the same MCU, the fault will occur on other PCB boards, but the recurrence probability will be different. RC circuit is added to XRSn of 100% faulty parts, and the pull-up resistance of DLP RST is improved after it is changed from 10K to 4.7K, but the probability that the two boards can start up normally is different.