TMS320F2800137: The code can run normally when JTAG debugger is connected., but it cannot work on standalone W/o JTAG debugger.

Ke,

What is your flash entry point? Did you try setting a HW breakpoint and debug the problem?

Regards,

Manoj

Hello Manoj,

I can run & debug code from the JTAG emulator but I don't run from Flash without starting from the emulator. 

And the flash entry point is 0x80000 Bank0 Sector 0. Thanks.

The LED5 is route to GPIO24. We test the level about R27, it always high level. Could you give us some suggestion/Dir?

Shaoxing,

Please try to emulate standalone boot (check BOOTROM chapter) , set a breakpoint at address 0x80000 and hit run. When you hit run, your PC should reach 0x80000. This would confirm that your boot pins configurations and hardware setup is fine for standalone boot.

Regards,

Manoj

Hello Manoj,

The customer board can run with JTAG emulator. 

Any latest update on this issue? Did you hit the breakpoint at 0x80000 when you tried emulate standalone boot?

Hi Manoj,

We have wrote the 0xA5FF to the BOOTMODE_Pin when JTAG debugger is connected, and then restart the CPU, it can not work. Also hit the breakpoint at 0x80000, it can not run this address here. 

If customer is using default boot mode pins (GPIO24 / GPIO32), then both GPIO24 / 32 should be pulled HIGH during device powerup to correctly boot to flash. If the GPIO24 / 32 are only pulled to 1.24V, then the device will not boot to flash. This is the reason why your standalone boot is failing. Did you try to understand why these pins GPIO24 / 32 are not getting pulled high properly. Doing ABA swap help us identify whether the issue is board related (or) device related issue. I would encourage you to try this experiment before shipping the devices for failure analysis. If the unit by itself is fine, it is just you have faulty boards.

Regards,

Manoj

Hi Manoj,

We test the GPIO24 voltage about the faulty board and the good board as below:

1.1.2V always during the faulty board power up. We have remove the R7/1.5K Ohm

2.3.3V first, and then low level 0V during the good board power up. We do not have remove the R7/1.5K Ohm

Thanks. Any comment and suggestion here?

Shaoxing Ke said:

The LED5 is route to GPIO24. We test the level about R27, it always high level. Could you give us some suggestion/Dir?

What is the purpose for R7 resistor and LED5? I don't have enough information to provide any recommendation / suggestion

On a faulty board, if you remove R7, does GPIO24 go get pulled upto 3.3V and stays high?

On a good board, why does GPIO24 go low? Is it pulled by your application code.

Hi Manoj,

They use the GPIO24 to control the LED5 after the chip work. And the chip-GPIO24 Pin → LED5 → 10Kohm pull-up resistor. We did removed the R7 from the faulty board, LED5 light will not affect the measurement of the chip. But we can still request the customer to remove the LED5- to test again. Hope we can give constructive advice to customer to troubleshoot the issue. Thanks.