• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 100 subscribers
  • Views 379 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Original question:

TMS320VC5416: About the Power Sequence

TMS320VC5416: Pull-down for TRST pin

O.H
Genius 5320 points
Part Number: TMS320VC5416

Hi experts,

This is a continuation of a related question.

A customer of mine is having a problem with the TMS320CV5416 not starting up properly on a board he created. This problem seems to occur frequently at temperatures above 20°C and humidity of about 80%. However, it seems to start normally at temperatures above 30°C and low humidity. It seems that heating up the board where this failure occurred will eliminate the symptoms. So, we cannot do a swap test because the problem is no longer caused by heating when removing the IC from the board. And when we checked the customer's design, the TRST pin was open.

Following the previous thread, they added a 470 ohm pull-down resistor to the TRST pin, but were unable to get it to boot correctly. The JTAG circuit on the board has 5 devices connected as shown below. (From Figure A-7. Multiprocessor Connecti of TMS320C54x DSP CPU Reference Guide)

I know that the support for this device is limited, but if possible, could you please tell me the following questions?

Q1: If five devices are connected at the same time, should I use a pull-down resistor of 470Ω/5=approximately 94Ω?
Or is it better to use a pull-down resistor of 4.7Ω (or 2.2kΩ) as shown in the figure below? (From Figure 3. Single Emulation Header with Multiple Target Devices and Buffered JTAG Signals in XDS Target Connection Guide).For the VC5416, there is a recommended value for the pull-down resistor on the datasheet, but I am confused as to which value I should follow for daisy-chain connection.

Q2: It is difficult to mount the TMS320VC5416 near the TRST pin, so they are mounted a little farther away. Is this okay?

Q3: They don't use a buffer for the TRST pin this time, is it correct that a buffer is not necessary for normal startup?
The TRST pin is marked as "If the TRST pin is connected to multiple DSPs, a buffer is recommended to ensure the VIL and VIH specifications are met." in the annotation, but I understand that the buffer should be used when the emulator is connected.

Q4: Am I right in thinking that if the TRST pin is fixed to Low (= 0.8V or lower) at power-on, the device should operate normally?
We are currently asking the customer to check the waveform of the TRST pin at power-on. If the TRST pin is fixed at Low, we believe that the cause of the device not booting up properly is somewhere else.

Best regards,
O.H

  • 0
    Genius 5320 points

    Hi experts,

    Does anyone have a view on this question?
    My customer is looking for an answer and would appreciate it if you could let me know the situation.

    Best regards,
    O.H

  • 0 in reply to O.H
    TI__Guru* 78245 points

    O.H -san

    As you know , we are not able to support deeper questions on this topic. With discussion with an emulation expert , who has no familiarity with C5000/C5416 and any device specific history - following are the response to the best of our knowledge.

    I am afraid we may not have further guidance on this beyond what  I provide below

    Q1 nTRST is broadcast to all TAPs in the scan chain. Value in newer devices/ datasheet recommends a single pull down (4.7kOhm) is adequate.

    This webpage might be useful:
    software-dl.ti.com/.../emu_xds_target_connection_guide.html

    On 5416 dataheet 470Ohm is called out. I do not have the background on why this is the case.

    Q2 : nTRST is not timing critical, so a longer route is fine.


    Q3 : I wouldn't think a buffer would be required, but it's very curious that the datasheet specifically calls this out.


    Q4 : I would think it would be closer to 0 than 0.8, but yes--if VIL is met then I would think that nTRST isn't the issue here.

  • 0 in reply to Mukul Bhatnagar
    Genius 5320 points

    Hello Bhatnagar-san

    Thank you for involving an emulation expert on this topic. I understood the content of your answer.

    I haven't gotten the signal measurements from the customer yet, but I will ask them to use 4.7kΩ to verify the operation and signal if the VIL cannot be met with 470Ω.

    Once again, thanks for your kind support and cooperation.

    Best regards,
    O.H