Create Account
Search
You are here: TDG Forum  >>  Texas Instruments SOMs and Development Kits  >>  OMAP-L138 EVM Development Kit  >>  UART Communication Failures

Technical Discussion Group Forum

This forum is provided for user discussion. While Beacon EmbeddedWorks support staff and engineers participate, Beacon EmbeddedWorks does not guarantee the accuracy of all information within in the Technical Discussion Group (TDG).

The "Articles" forums provide brief Articles written by Beacon EmbeddedWorks engineers that address the most frequently asked technical questions.

To receive email notifications when updates are posted for a Beacon EmbeddedWorks product download, please subscribe to the TDG Forum of interest.

TDG Forum

PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 14 Oct 2013 12:27 PM by  steven.eckhoff
UART Communication Failures
 5 Replies
Sort:
You are not authorized to post a reply.
Author Messages
steven.eckhoff
Basic Member
Basic Member
Posts:192


--
09 Oct 2013 04:02 PM
    We are experiencing frequent failures when using the SOMOMAPL 138 module UART2 serial bus. A failed module is unable to send or receive serial data on TXD/RXD of UART2 (J1 97, J1 99 on the connector). Other elements of the DSP seem to continue to work properly despite the serial communication not working.

    At first I thought it was random/unlucky occurence but with 3-4 similar failures I believe it must be something systematic. Our current communication method involves using a USB to 3.3V converter cable to talk to a PC via a USB port (TTL-232R-3V3 found @ http://www.ftdichip.com/P...s/USBTTLSerial.htm). I have noticed that the LogicPD module has internal pull-ups to 3.3V on the TXD/RXD lines for UART2, and recently I noticed that the converter cable is also internally pulled up to 3.3V. I was wondering if having pull-ups on both sides of the interface could cause the TX/RX pins to exceed their maximum ground sink current when zeros are sent?

    Could this account for the failures I am seeing? Any thoughts or other suggestions on this problem would be greatly appreciated.
    steven.eckhoff
    Basic Member
    Basic Member
    Posts:192


    --
    11 Oct 2013 10:18 AM

    Yes, I have seen issues on an I2C bus where there we pull-ups on both sides. Removing the pull-ups on the non SOM side may fix this. I am not sure why there are pull-ups on the UART TX/RX. On the SOM side they are there to interface with the RS232 transceiver (See the baseboard schematic for details). Did the pull-ups come on the UART to USB device?

    It is hard to say without more information. Are you also able to send a log of the terminal output. I am assuming that you are using Linux. If you are not which OS are you using?
    Matthew Dzikowski
    New Member
    New Member
    Posts:2


    --
    11 Oct 2013 11:35 AM

    I tried removing the pull-ups on three of the failed units.  One unit was once again able to communicate on UART2, the other two were not.  I am not sure what to conclude from this, other than to mention that the two that were unrecoverable are older.  Could continued operation with this double pull up arrangement lead to a permanent failure of the pins?

     

    The OS we are using is generally Windows XP, although occasionally Windows 7 pc's are also used.  I don't think the terminal log would be of much use, as I don't receive corrupted serial data when these failures occur, there is just no received data at all.
    steven.eckhoff
    Basic Member
    Basic Member
    Posts:192


    --
    11 Oct 2013 12:41 PM
    I am not sure why there are pull-ups at all on the TX/RX. Did these come on the product http://www.ftdichip.com/S..._TTL-232R_CABLES.pdf See Pg. 20. There it says that the pull-ups are not fitted. The pull-ups on the SOM side are for the interface to the RS232 transceiver on the EVM and Experimenter baseboard.

    Also, how are you interfacing to the UART2 logic? Are you using the DB9 connector? You should use that one.

    Before interfacing to the transceiver take a look at the datasheet for it. http://www.ti.com/lit/ds/symlink/trs3386e.pdf
    Matthew Dzikowski
    New Member
    New Member
    Posts:2


    --
    11 Oct 2013 02:41 PM
    It may not have been clear in my original question, but I am not using the development board. We have custom designed board that uses the SOMOMAPL138-10-1602QHIR module as a daughterboard. The motherboard simply passes the UART2 pins directly to a connector (no line driver, pull-ups or other logic on the TX/RX lines). This connector is where I have the interface to the cable.

    The datasheet for the cable seems to be a bit misleading. I noticed the lack of pull-ups on page 20 as well, but when I measure the TX/RX pins of the cable connected to a computer but just floating on the other end, they measure 3.3V. I spoke with someone at FTDI, and he said there is still an internal pullup within the converter chip in the cable with about 200k to 3.3V.
    steven.eckhoff
    Basic Member
    Basic Member
    Posts:192


    --
    14 Oct 2013 12:27 PM
    Mathhew,

    In this case you should test the RS232 to USB cable with the SOM on the EVM or Experimenter baseboard. That will verify that the cable works when it receives the signals at RS232 levels.

    The problem on your baseboard exist because there is no RS232 transceiver on board to drive the signals correctly for the cable that you are using. Take a look at the EVM or Experimenter baseboard schematic for a reference.

    I am not sure why they have the pull-ups on there side, but you can test that it functions correctly using the baseboard that came with your kit.

    [1] http://support.logicpd.co...rtalid=0&EntryId=767
    [2] http://support.logicpd.co...rtalid=0&EntryId=913
    You are not authorized to post a reply.

    Forums Texas Instruments SOMs and Development Kits OMAP-L138 EVM Development Kit