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.

MSP-GANG programming with BSL: overall result ok but actually not working (for the most part)

Other Parts Discussed in Thread: MSP430F2274, MSP-GANG

HI,

I'm flashing MSP430F2274 with MSP-GANG over BSL. For the most part it is not working (only a few times it worked). I get an overall result  OK (with the GO command). Then i try to read back and i get:

ERROR 68: FBSL initialization error
Failed

Then i try to read again, and it reads empty memory.

I'm using the latest GUI and i updated the firmware several times.

Any help would be appreciated.

Gorazd

  • Hi Gorazd,

    We are looking into the issue and hope to have an answer to you soon.

    Regards,
    Akash Patel
  • Hi Gorazd,

    This error you are receiving is regarding the Fast BSL initialization. Communication through BSL is at 9.6kb/s while Fast BSL is at 150 kb/s. Do you have any RC filters or suppressors on the BSL TX/RX data lines that may be limiting the communication speed?

    If you can provide us with a schematic that would be great.

    Regards,
    Akash Patel
  • Hi,

    the weird thing is i get a PASS but the MCU is not working? Flash is actually empty if i read back.

  • Hi Gorazd,

    Generally we suggest that you have GND between RX and TX to avoid crosstalk, especially if with a long cable. Also you are reading empty flash because you have the password set to "All 0xFF." You should program the MCU with the correct firmware and provide the password from the Code File.

    Regards,
    Akash Patel
  • "Generally we suggest that you have GND between RX and TX to avoid crosstalk, especially if with a long cable"

    but if there is a crosstalk, the verify would FAIL, wouldnt it?

    "You should program the MCU with the correct firmware"

    The firmware is 100% correct. I can program it without problem with SBW. However, after programming a device "successfully" with BSL, the device seems to be corrupted. It can still be programmed "succesfully" (with both BSL and SBW), but it doesnt work. A device works only if programmed with SBW and was intact beforehand.

    Do you think it is possible for a device to get corrupted by programming it with BSL somehow?


    Another question regarding SBW: is SBW communication error prone if the signal wires are of different length?


    "and provide the password from the Code File"
    I tried this option, it didnt work. How do i provide a password correctly in the Code file? And what password?
  • Hi Gorazd,

    It's important that SBW signals are clear at both ends and we recommend that you either using a flat ribbon cable or twisted pair. If using a twisted pair, each of them should have one wire as the signal line and the other connected to GND on both ends.

    For more information regarding the password please see #5 (on page 15) under Section 2.1.1 Programming Using Interactive Mode: www.ti.com/.../slau358j.pdf

    Regards,
    Akash Patel
  • gorazd kovacic said:

    The firmware is 100% correct. I can program it without problem with SBW. However, after programming a device "successfully" with BSL, the device seems to be corrupted. It can still be programmed "succesfully" (with both BSL and SBW), but it doesnt work. A device works only if programmed with SBW and was intact beforehand.

    Do you think it is possible for a device to get corrupted by programming it with BSL somehow?

    If I remember right, for BSL on MSP430F2274 must be specified password (vector table from last download). Without password BSL will mass erase device including info segment A with DCO constants. And in this case downloaded program will not work if DCO constants are used.

    gorazd kovacic said:

    Another question regarding SBW: is SBW communication error prone if the signal wires are of different length?

    SBW should have same length of CLK / DATA (TEST / RESET) lines. Problem for SBW is cap on RESET pin. I tested SBW without cap on RESET pin (5xx flash target device) with direct connection from master (pins) over 1 m length ribbon cable (without drivers / termination) on 6 MHz (200 KByte/s write / verify). It was working just fine on 1.8V and 3.3V.

  • And why you want to use BSL at all, when your device support SBW with 2 less wires than BSL?

  • Hi, tnx for your answers, i didnt read the user manual carefully enough ( SLAU358 pdf).

    We experienced BSL is faster. Also we thought BSL was more reliable when considering long wires (50 cm). But yes, you're right, SBW should do the work just fine.
  • gorazd kovacic said:

    We experienced BSL is faster. Also we thought BSL was more reliable when considering long wires (50 cm). But yes, you're right, SBW should do the work just fine.

    My SBW+ (MSP-GANG like) programmer is faster than MSP-GANG, and without cap on target device RESET pin it can write up to 48 KByte/s on 2xx and 200 KByte/s on 5xx. Connection between master and target devices can be longer than 50 cm. With 1 nF cap on target RESET pin it is limited to slow mode (still faster than MSP-GANG). It is without fuse blow option for 2xx.

    http://forum.43oh.com/topic/2972-sbw-msp430f550x-based-programmer/?p=48762

**Attention** This is a public forum