• State
  • Locked Locked
  • Replies 5 replies
  • Subscribers 30 subscribers
  • Views 181 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.

TM4C1294KCPDT: Low Voltage UART Pulses

Garret Holt
Intellectual 1050 points
Part Number: TM4C1294KCPDT
Other Parts Discussed in Thread: ISO7521

Hi Team,

We are interfacing a Tiva(TM4C1294KCPDTI3) processor with C2000 via ISO7521. We are observing a dwarf pulse that is only around 960mV tall on the UART line. The tiva is powered from an isolated winding of the flyback while the C2000 is powered by another isolated winding. There is Y-cap (2.2nF) that links the positive DC bus voltage on the flyback primary winding to iso GND of the TIVA powered isolated winding. With just the Y cap they observe the 960mV pulse but when they add damping resistors (2x 100Ohm) in series with that Y-capacitor the pulse is transmitted properly at 3.3V. When observing the UART lines with and without the resistors there is a clear difference in the amount of added noise. The low voltage pulse is an issue because it is not transmitted by the digital isolator and causes data loss.

We suspect the Y cap is providing a feedback path that is adding noise to the Tiva lines. The resistors potentially dampen this noise which is why the issue is removed when they are added. We are trying to understand why the presence of noise causes the device to act in this manner. Have you ever observed an issue with a Tiva UART line outputting pulses of lesser magnitude? If so, could this be caused by noise and why does this manifest as lower magnitude pulses? Thank you for any guidance you can share.

Regards,

Garret

  • 0
    TI__Guru*** 128155 points

    Hello Garret,

    Have they done any probing to see the power signals for the TM4C device on waveforms? If there is noise present, then it would be beneficial for us to understand more about how much noise is present with and without being dampened by the resistors.

    Garret Holt said:
    Have you ever observed an issue with a Tiva UART line outputting pulses of lesser magnitude? If so, could this be caused by noise and why does this manifest as lower magnitude pulses?

    That is not something I have seen on my end, but I don't recall having assessed a system configuration quite like this one either. More information here about what noise may be entering the device would be beneficial.

  • 0 in reply to Ralph Jacobi
    TI__Intellectual 1050 points

    Hi Ralph,

    On TM4C1294KCPDTI3 we probed Vdd pins 90 and 69. RX is on PB0(pin 95) while TX is on PB1 (pin 96). Both signals look clean. Is there a particular VDD that supplies these pins that we should be looking at specifically?


    Regards,


    Garret

  • 0 in reply to Garret Holt
    TI__Guru**** 159085 points

    Hi Garret,

      Not really sure what is happening as I have not seen this either. But you seem to have a workaround solution using the resistor that filters out the noise band. I have a couple of questions.

       - Is the dwarf pulse seen on the UARTTX or UARTRX?

       - Is the dwarf pulse see on during transmission or during idle?

       - What does the VDD power look like when the dwarf pulse happens? A scope cap will be good.  

       - Is the noise you see before the ISO7521 or after the ISO7521 isolator?

       - If possible, what if you somehow bypass/unpower the ISO751, do you see the dwarf pulse? 

       - Is this a phenomenon only seen on one board or all the boards?

      

  • 0 in reply to Charles Tsai
    TI__Intellectual 1050 points

    Thanks Charles, I have included responses below.

       - Is the dwarf pulse seen only on the UARTTX or also UARTRX?

    Only seen it on the UARTTX line; the RX line hasn’t shown the issue

      - Is the dwarf pulse see on during transmission or during idle?

    Only during transmission

       - What does the VDD power look like when the dwarf pulse happens? A scope cap will be good.  

    The power rail on VDD pins 90 and 69 both look very clean.

     - Is the noise you see before the ISO7521 or after the ISO7521 isolator?

    Before the digital isolator as this is only observed on the TX of the Tiva device.

       - Are there any pulses when idle or dwarf pulses identified that correspond to a real known bit?

    Yes each dwarf pulse seems to be a real intended pulse from the UART, just at reduced amplitude.  We have also seen the UART finish a transmission with the dwarf state; meaning when it tries to generate the stop bit (a high) it outputs the dwarf and remains in that state until the next transmission; so it actually idles in dwarf state between UART transmissions.

       - If possible, what if you somehow bypass/unpower the ISO751, do you still see the dwarf pulse?

    Haven’t tried this yet; as it would require a lot of alterations, like the application (motor output) can’t run if the two boards aren’t communicating. We don't understand how a buffer input (on the ISO751) could cause the dwarf pulse on the UARTTX 

       - Is this a phenomenon only seen on one board or all the boards?

    This condition is not unique to a specific board; we’ve seen it on many if not all boards

  • 0 in reply to Garret Holt
    TI__Guru**** 159085 points

    Hi Garret,

     I actually don't know why without the resistor the signal cannot be pulled up to the VDD rail. However, I think it is something unique to your board as you mentioned many if not all the boards have the same problem. 

      - Can you show how you configure the UART pins?

      - What other peripherals are operating at the same time as the UART? For experimental purpose, does it make a difference if UART is the only thing operating at the time? The best way to solve problem is to start with a simple system. 

      - Does the baud rate matter?

      - Can you try with the LaunchPad? Unless you can replicate on a launchPad, it is not going to be easy to solve a problem like this. Yet, you may not see it on the LaunchPad as it has a different regulator on the board. 

      - Do you have common ground for all components on the board?

      - Regarding my earlier question about bypassing the ISO751, the purpose is to see if the Tiva board is affected in any way with/without the presence of the ISO751. The C2000 board needs not be functional. 

      - Can you send the schematic around the Tiva device? There is no guarantee I can spot anything wrong unless it is something that is obvious.