• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 80 subscribers
  • Views 1377 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Spy BiWire problem with MSP430FR5969

John Wygonski
Intellectual 870 points

Other Parts Discussed in Thread: MSP430FR5969, ENERGIA

I am using a MSP-FET430UIF to debug a custom PCB based on MSP430FR5969.   I have the GND, SBWTCK, and SBWTDIO, and VCC leads connected via 40cm of 28 ga. wire to the target, with the target supplying 3.3V power.

Running CCS 6.1, I get this error when downloading:

MSP430: Error connecting to the target: Unknown device

Same error running Energia.

If I connect a scope probe lead to the SBWTCK pin on the target, I can download fine without the error.  

I found a post describing a JTAG issue for MSP430 devices with similar symptom concerning edges of SBWTCK signal, and I tried the proposed remedy of a 4.7k pulldown on the TEST pin, but it didn't work.  Problem is also described on p. 25 of SLAU320T. 

Does anyone else have a suggestion as to why the scope probe loading makes Spy BiWire work in this case?

Thanks,

John

  • 0
    TI__Guru 56685 points

    Hello John,

    Have you made sure you have setup the SBW connections as shown in Figure 2-3 in Section 2 of the Hardware User Guide (www.ti.com/lit/slau278 )? This figure should be used for our FRxxx products as well. Please note we will be updating this guide in order to accurately reflect this.

    Regards,
    JH

  • 0 in reply to Jace H
    Intellectual 870 points

    Yes, the signal connections are there because SBW works sometimes.   We do not have the RC in place on *RST because we are using an external reset circuit.

  • 0 in reply to John Wygonski
    TI__Guru 56685 points
    John,

    The RC circuit associated with the RST line is required in order for proper programming via SBW. Please also pay close attention to Note B in the Figure for too high of capacitance can cause connection issues.

    Regards,
    JH
  • 0 in reply to John Wygonski
    Guru 23000 points

    Your connection cable is too long. This can result in crosstalk and noise pick-up from other sources. Adding a (probe) capacitance will work as a filter. The behavior can differ if you connect the probe at the begin or the end of the cable, special with this long cable. The long ground wire act as an inductor and a voltage difference between both devices can occur, adding a (probe) capacitor will filter this partly.

     

    Here are some general rules for a JTAG connection;

     

    JTAG connection 2-Wire, Vcc (sense) from target, MCU is powered by target board PS and not by JTAG:

    • Connect SBWTDIO (/RST) from MCU to JTAG TDO/TDI pin-1.

    • Verify the SBWTDIO (/RST) pull-up resistor is not smaller than 4.7KOhm and the (ground) capacitor not greater than 2.2nF. If any doubt remove both, the JTAG doesn’t (like/) need them.

    • If the Reset is provided from another source, be sure this is an ‘open-collector’ type, is floating and from low capacitance.

    • Connect SBWTCK (/TEST) from MCU to JTAG TCK pin-7. A connection to JTAG pin-8 is normally not necessary and in this case the series resistor is not required.

    • Don’t use long connection cables (>150mm).

    • If using flat-cable use the outer conductors for SBWTDIO & SBWTCK, this reduces their capacitance by ½. Next to both wires Vcc or Vss for shielding.

    • Connect Vcc from MCU to JTAG VCC-TARGET pin-4.

    • Connect Vss from MCU to JTAG GND pin-9.

    • Do not make any other JTAG connection as the above 4 ones.

    When using CCS start with the ‘Slow’ speed option in the ‘Target Configuration’.


  • 0 in reply to Leo Bosch
    Intellectual 870 points
    Leo,
    Thanks for the reply. Reset is provided by a Maxim chip with an open-drain output. My setup meets all the other points you stated except for the cable length. I'll rework it and let you know the outcome.

    John
  • 0 in reply to Leo Bosch
    Intellectual 870 points
    Leo,
    I shortened the cabling to 10 cm plus the ribbon cable and no errors yet while debugging. THanks for the suggestion.

    John
  • 0 in reply to John Wygonski
    Guru 23000 points

    My pleasure John. I’m sure you have chosen the right solution.

    Leo.

**Attention** This is a public forum