• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Subscribers 81 subscribers
  • Views 424 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.

TMS320F28027: Launchpad F28027, FTDI serial to usb fail in high EMI environment

Arief Noor Rahman
Intellectual 360 points
Part Number: TMS320F28027

Hi Everyone,

I am working on digital control 3 phase PFC, the converter use standard 6 switches inverter structure, implemented with Si MOSFET with Fsw = 50kHz, and I use Launchpad F28027 for my controller prototype.

I need to have continuous communication between my PC and the MCU to monitor some operating parameter of the PFC. I use the serial communication (UART) and internal FTDI IC as serial to USB.

My question is, now I use the circuit as inverter to test the basic function and feedback circuit when the input voltage increase above 170V, the communication link always broken. -
->>The inverter was tested and can work properly with input voltage up to 400V

Can anybody suggest me what should I do to improve the communication link?and what is the kind of EMI noise that cause it to fail?dV/dt noise?
- so far I tried to cover the MCU with faraday cage, but still fail

Thanks

  • 0
    TI__Guru 71815 points
    Arief,
    I think the key question is where in the signal path is the SCI/UART signal getting corrupted. Certainly those components closest to the PFC circuit, i.e. the F28027/FTDI/LP as a whole would be thought to be the issue, but it could also be the USB cable itself, or even the PC. It may be as subtle as the clock but not the data line.

    I would start by probing the UART signals coming from the F28027 as close to the device as possible in both the passing and failing conditions. If you see the noise at that point, then we can focus there, but if not I would then look at the FTDI output, then potentially along the cable, etc.

    Another thing to try would just be to slow the baud rate from the C2000 to see if it buys you anything in the way of immunity.

    Another tactic would be to place the UART activity away from the switching points of the FETs, since that is where the excitation is coming from. The placement shouldn't be too hard, since the MCU is controlling the PWMs, what may be challenging is getting enough time to transmit the data before the switching happens again.

    Let me know what you find, and we can look to get more specific with our suggestions.

    Finally, if there are others here on the forum that have dealt EMI immunity prior, your comments are certainly welcome to the discussion.

    Best,
    Matthew
  • 0 in reply to MatthewPate
    Intellectual 360 points

    Hi Matt,

    I have find the solution, I use another ftdi usb2serial IC and put it close to my PC then use fiber optic cable to connect the GPIO29 of the MCU(SCI-TX) to the ftdi RX pin.

    I strongly believe the problem was on the ftdi to PC side comm, because if the data corrupt occur on MCU to ftdi, the data will just corrupt but the PC should remain connected to FTDI IC.

    The possible reason why fiber optic  worked could be the EMI in power converter is so strong that it can corrupt the communication from PC to FTDI IC by noise entering the USB cable or directly attack the FTDI IC itself...thus separating the FTDI IC from the EMI source make keep it work reliably...and use fiber optic to ensure no EMI picked up on MCU to FTDI link