Lm3s3739 usb host mode not work well

Token,

Is this a custom board or an evaluation kit?  Please send us the date code on the third line of the device marking.  Also, please send us the USB clock setting.

Thanks,

Sean de la Haye

Hi,

 Thanks for you replies,

Stellaris Sean

Is this a custom board or an evaluation kit?

It is custom board, 

Stellaris Sean

the date code on the third line of the device marking

The third line is  "S9-04P006H" , the whole marking is "LM3S3739 IQC50A0 SD  S9-04P006H"

Stellaris Sean

the USB clock setting.

I can't find the code about usb clock setting,  the below code is the system clock and usb pll setting. the all about usb host  is the same as lm3s3748 evaluation kit's code.

The system clock setting (the hardware crystal is 8Mhz). :

       "SysCtlClockSet(SYSCTL_SYSDIV_4 | SYSCTL_USE_PLL | SYSCTL_OSC_MAIN | SYSCTL_XTAL_8MHZ);"

the usb clock setting is:

    MAP_SysCtlPeripheralReset(SYSCTL_PERIPH_USB0);
    MAP_SysCtlPeripheralEnable(SYSCTL_PERIPH_USB0);
    MAP_SysCtlUSBPLLEnable();

Thanks

   Token.

Token,

From your description, it sounds like you are not having any problems with the clocking errata.  I talked with one of the USB experts here and they recommended removing your pullup to VBUS on PB1 (Replace it with a 0 Ohm resistor or short the connection). They have told me that if you connected a oscilloscope on the pin side of the resistor, you would see the voltage droop by ~0.3V when you connect.  The resistor is causing enough of a power draw that when you connect it is going below the 4.7V lower limit.

Thanks,

Sean

Hi Sean,

     I have connected the PB1 to VBUS with 0 Ohm resistor, The VBUS voltage can reach the 4.97v , and the rise time is 380us. The system has generate the VBUS ERROR before connect the usb disk.

   Thanks,

   Token.

Token,

My colleague has suggested that you try driving 5V on PB1 before the device is powered up. He said that this is a requirement listed in section 12.3 in the LM3S3739 errata document.

Thanks,

Sean de la Haye

Hi Sean,

   I  have a test according your advice, driving +5V on PB1 before the device is powered up,  and connect PB0 to GND through 0 Ohm resistor,  But the methods can't resolve the problem. In the key,  some is ok among our productions, i have some switch test,but the vbus error follow the lm3s3739.

    Thanks,

    Token.

  Hi sean,

       This problem is related to the version and code date of the chip? I have test two pcb with "LM3S3739-IQC50A0-$9-06P048",the both can't work on usb host,  then , test two pcb with "lm3s3739-ICQ50A0-$9-21P072H", the both work well?

    Thanks,

   Token.

Token,

Thank you for the information on the part numbers -- this should help the debug.  I will consult my colleagues and see if there are any other ideas.

Sean

Token,

I have talked with my colleagues and we can only conclude that the parts with the LM3S3739-IQC50A0-$9-06P048 date code are damaged. The next step is to contact your local TI Sales office to get them replaced with the parts with the newer date code.  It was also recommended that a 4.7 Ohm resistor be connected between the +5V and PB1.

Thanks,

Sean

Sean,

   Thanks a lot,  Do you mean the chips is okay, only  damaged in using ? To say simply, I only want  to know what result in the problem,  and how to avoid it.  your advice adding 4.7 ohm resistor  i will have a test, but to get the effect it isn't a short time.

  thanks

  Token.

Token,

They are most likely defective if you have newer devices that are functioning correctly.  Since the newer devices you have are working, our feeling is that the replacement chips should also work.

Sean

Sean,

      Sorry, I probably did not describe clearly.　In all products using "LM3S3739-IQC50A0-$9-06P048H', most of it is normal.There is a small part of products is not working properly. but We still have a lot of inventory,and not only the date code. if we apply replacement, We must spend long time in the test. All of these will affect the progress of our products. Oh , My GOD!

    Token.