TM4C123AH6PM: Failure with 15 Hz TDO frequency output

Hi Alan,

  I'm sorry to hear your problem. What debug probe do you have? If you are using LM flash programmer then it only supports ICDI debug probe. Do you have another debug probe such as XDS100, XDS110, or XDS200 that you can try? If you have these probes, then you can create a corresponding target configuration file or you might already have one. See below example. You can click the "Test Connection" to verify the JTAG scan-chain.

      Can you also answer me a few questions?

  1. How many uC do you see the problem? Can you give a percentage too?

  2. When you do unlock, did it succeed?

  3. Can you measure the device current draw on the suspected uC vs the good one?

  4. Can you measure the VDDC (1.2V) pins? Is it 1.2V?

  5. Are you saying without the debugger connected the TDO is producing a 14.7Hz output? Or this is with the debugger driving the JTAG interface? Is the TDO periodic? I have not seen/heard TDO outputting 14.7Hz signal myself. 

  6. Can you take a picture of the uC. I'd like to see the marking of the uC for its "Date Code" value.  

Alan Caravaggi said:

All voltages are correct, but I have noticed that a frequency of 14.7Hz comes out of the TDO signal.

Section 5.2.2.1 Reset Sources of the Tiva C Series TM4C123AH6PM Microcontroller Data Sheet datasheet (Rev. E) contains:

Note: If the device fails the initialization phase, it toggles the TDO output pin as an indication the device is not executing. This feature is provided for debug purposes.

The datasheet doesn't explain if there is any significance to the frequency on the TDO signal after a failure in the initialisation phase.

Thanks for reply.

I don't understand if it's an hardware fail or a memory problem.

1.  I had 5 failur on 8 board.

2. For the unlock procedure I use the EK-TM4C1294XL (evaluation board) and LM Flash programmer, but the software gives me the error "Target not found", I have try with working micro and the procedure goes right.

3. The fail micro draind 5mA, a good one drains 13mA at startup (boot mode) and when start the program 21mA.

4. The VDDC is 1,2V.

5. The frequency of 14,7Hz is continuous and without debugger (the other JTAG pins are High), if I put the micro in reset the TDO goes High like the other JTAG pin.

6. The chip reports:

980 YF

TM4C123A

H6PMI

03C3XEW

G4

Thanks

You're right, I had already seen the datasheet.
What I don't understand is if it's a hardware or firmware proplema, and the datasheet doesn't explain it.

Thanks

Hi Alan,

Alan Caravaggi said:

6. The chip reports:

980 YF

TM4C123A

H6PMI

03C3XEW

G4

   Please allow me some time and I will get back with you based on your provided information. 

Hi Alan,

  Can you please give me the chip marking for all five failure chips? Thanks. I want to know if they are manufactured from the same batch. 

Hi Charles,

these are the other chip:

- 980 YF TM4C123A H6PMI 81C9CQW G4

- 980 YF TM4C123A H6PMI 81C9CQW G4

- 980 YF TM4C123A H6PMI 03C3XEW G4

- 980 YF TM4C123A H6PMI 03C3XEW G4

We are now investigating on electrical problems of spikes due to relay activations, but for now we have not been able to find what creates the fault.

Hi Alan,

  Thanks for providing these information. Can you confirm that all these five chips show the same behavior where the TDO toggles at 17Hz or the behavior is different among them?

Hi Charles,

I confirm that all the broken micro have the TDO toggles at about15Hz

Hi Alan,

  Thanks for the confirmation.

  I also send you a friend request.

Hi Chester,

maybe we have found the problem. Until now, the programmers had assured me that nothing was written in the EEPROM at shutdown, after a double check of the program they finally informed me that they had added a write cycle in EEPROM at shutdown.
Also our chips are revision 6, which would all lead back to MEM # 04.
We are running further tests, but I am quite confident.
Thank you very much for helping.

Hi Alan,

  First thanks to Chester for pointing MEM#04 as a reason for the locked out device. In your earlier posts, I never came across the MEM#04 errata as you didn't mention about having a EEprom write cycle prior shutdown. If your system indeed was writing to EEprom with a abrupt shutdown or reset then it can render the device bricked for the rev6 silicon. If you cannot avoid abrupt shutdown or reset during a write cycle then I will suggest you use an external EEprom if possible.