CAN problem in reception with latest v1.08 WinCE 6 kernel

  • Hello,
    We are recently having CAN frames problems in reception with NetDCU14 and latest WinCE6 kernel v1.08.
    More in detail, we have communication on both CAN ports simultaneously in our system:
    CAN1: configured with 29 bits header and 100kbps speed and with about 40 messages/second bus load.
    CAN2: configured with 11 bits header and 500kbps speed and with about 50 messages/second (20 for TX and 30 for RX in the NetDCU14).


    What we have detected is that randomly a message with all 8 bytes = 0 appears on NetDCU14 in CID2 in reception rutine (we use F&S CAN dll and .Net in VS2005). Sometimes it takes one minute and sametimes it takes more minutes in the communication proccess, but it finally always occurrs.


    That means, from structs:


    public struct CanEvent
    {
    public int arbitration;
    public CanPort.CanTransmitData data;
    public CanPort.CanEventFlags event_type;
    public uint lost;
    public CanPort.CanTime time;
    }


    and
    public struct CanTransmitData
    {
    public byte dlc;
    public uint id;
    public byte[] msg;
    public byte rtr;
    }


    we obtein data.msg[]=0 for all 8 bytes received in this message on CID2 with the configuration described.


    With a CAN monitor connected to the network we are sure that no CAN mesages are sended by any netwpork node with all bytes set to 0, so we assume it can be a problem with CAN dll driver of this kernel release of NEtDCU14.


    Can you confirm something about or try to reproduce ?


    Thanks.

  • Hello,
    which ID has this stange message? I never saw this! Is the CAN driver version de facto 2.15 - as ith should be.

    F&S Elektronik Systeme GmbH
    As this is an international forum, please try to post in English.
    Da dies ein internationales Forum ist, bitten wir darum, Beiträge möglichst in Englisch zu verfassen.

  • Hello,
    3 different messages with IDs 0x100, 0x101 and 0x102 are received by NetDCU14 10 times per second.
    Any of them can be received with this error randomly in our tests (the content of the data.msg[] is 0 in all 8 bytes received) and it can occurrs in any of the 3 IDs received.
    Before we use NetDCU14 we have been using NetDCU10 during 3 years and never have seem this problem on NetDCU10 CPU.

  • Hello, Version will be found in the serial boot output:
    ...
    CIDx Version XY
    ...


    Thanks

    F&S Elektronik Systeme GmbH
    As this is an international forum, please try to post in English.
    Da dies ein internationales Forum ist, bitten wir darum, Beiträge möglichst in Englisch zu verfassen.

  • Hello,


    ..........
    CID: Version 2.14, ActiveKey = Drivers\Active\23
    CID: Version 2.14, ActiveKey = Drivers\Active\24
    ............


    So, if we are flashing the latest v1.08 kernel WinCE6 with CF2 available on ftp ... how is possible that CAN driver is not exactly 2.15 ?
    Any improvement to consider from 2.14 to 2.15 ?
    Could be that 2.15 was packaged only in WinCE 7 kernel images ? how can we obtein it ?

  • Hello,
    it was my fault CAN 2.14 is correct for kernel 1.08. I attached you driver 2.15. This driver has a better performance. But i have no idea where the wrong ID comes from. If the problem persists - can i see it using "CAN Test Suit"?


    Register the atteched driver by copy it to ffsdisk and change regiters name from <drivername> to \ffsdisk\<drivername>.

    Files

    • fs_candrv.zip

      (18.12 kB, downloaded 167 times, last: )

    F&S Elektronik Systeme GmbH
    As this is an international forum, please try to post in English.
    Da dies ein internationales Forum ist, bitten wir darum, Beiträge möglichst in Englisch zu verfassen.