• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 30 subscribers
  • Views 865 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.

TMS570LS1227: Power Issues On the 1.2V Rail

Gabriel Soares
Prodigy 180 points
Part Number: TMS570LS1227
Other Parts Discussed in Thread: HALCOGEN

Hi,

I'm having some weird power issues on the 1.2V rail of my TMS570. I have a custom PCB with the IC on it, and have 5V, 3V3, and 1V2 regulators on board to provide power for each of the rails. 

The schematics for the board are here: https://github.com/sfuphantom/vcu-hw/blob/rev2/PHANTOM_VCU/PHANTOM_VCU.pdf

I have been using the board reliably for a few months now, but a few days ago I was working on it and I could see some LEDs connected to my TMS570 GIO pins start flickering, and then go black all of a sudden, with the error LED also lighting up faintly (I had no contact with the board here, I was just watching it on my bed.. ESD?). The board failed to flash code after this and would not reset. Probing with a multimeter I found that the 1.2V rail was down to 0.4 Volts and a continuity check showed the rail was shorted to GND. 

I replaced the 1V2 regulator but still had the same problem. With the regulator off in between swaps, I again checked for continuity and the rail was still shorted to GND (the only thing connected to this rail is the TMS570). When powered, the hottest thing on the board is definitely the 1V2 regulator which burns to the touch. The MCU is not noticeably hot. 


I took off the 1V2 regulator again and soldered a wire to GND and to the 1V2 rail, and while powering up the 5V and 3V3 rails I supplied 1.2V externally from a power supply. This weirdly seemed to fix the situation and code would again run on the TMS570 with the rail operating normally (even though it was shorted before, now continuity tests would show no short). I put the 1V2 regulator back on and the board worked fine. 15 minutes later I nudged one of my JTAG connectors and the 1V2 rail died again.. 

I had kept my wires soldered to the 1V2 rail so I again supplied 1V2 from the supply and it almost acts like a jump start. As soon as I provide that power the GPIO lights up, the error LED turns off, and the board operates again.. This cycle keeps happening. 

What could be the problem here? I am pretty lost, especially considering continuity shows the rail shorted until I provide external power, then even after I remove external power the board operates fine again with no short until it dies randomly. 

  • 0
    TI__Guru**** 197006 points

    Hello Gabrief,

    The max output of the 1.2V regulator (MIC5504) is 300mA. The maximum current required for TMS570LS12x is much higher than 300mA. 

  • 0 in reply to QJ Wang
    Prodigy 180 points

    Hi QJ,

    Thanks for your reply, so do you think the TMS570 is browning out the regulator? Would this show the short circuit behaviour on the 1.2V rail with the regulator depopulated? 

    I'm guessing that I haven't seen this problem yet in the few months of use because I've been lucky to not flash code with high current draw? 

    Thanks,


    Gabe

  • 0 in reply to Gabriel Soares
    Prodigy 180 points

    Hi QJ,

    I flashed code that only sends a mibSPI command in a for loop and the problem still occurs. I've never had this problem before now, and even providing an external 1.2V from the supply is no longer helping. I have a GPIO LED connected to one of the pins and it is flickering like crazy now without any code configured to that pin. 

    It seems like the current draw isn't the only issue, do you know what else could be causing this?

    Thanks,

    Gabe

  • 0
    TI__Guru**** 197006 points

    Hello Gabe,

    The DAC (U8) is 5V device. TMS570LS12x device doesn't tolerate 5V signals and power supply. 

    "I took off the 1V2 regulator again and soldered a wire to GND and to the 1V2 rail, and while powering up the 5V and 3V3 rails I supplied 1.2V externally from a power supply." 

    How does this work?

  • 0 in reply to QJ Wang
    Prodigy 180 points

    Hi QJ,

    The DAC (MCP48FV)'s voltage input high threshold is low enough to read the TMS570's 3V3 logic. For the MISO pin I put a voltage divider to drop the 5V line to 3V3 for the MCU to read. This then allows the 570 to control the DAC's 5V analog output with just 3V3 logic. 

    Without the 1.2V regulator depopulated no 1.2V is generated onboard. To supply the 1.2V I use a benchtop power supply connected to the 1.2V pad, the MCU draws current from there for the VCC rail. 

    Every once in a while the MCU error LED turns on and draws around 300mA from the supply. After I hold reset for a while and cycle power the MCU then returns to normal operation. Very confused.

    Thanks,

    Gabe

  • 0 in reply to Gabriel Soares
    TI__Guru**** 197006 points

    Hi Gabe,

    The maximum allowable power dissipation of any TA (ambient temperature) is PD(max) = (TJ(max) – TA)/θJA. Exceeding the maximum allowable power dissipation will result in excessive die temperature, and the regulator will go into thermal shutdown.

    At room temperature (Ta=25C), PD(max)=(125-25)/250=0.4W.

    The actual power consumption is P=(3.3V-1.2V)*300mA= 0.63W which exceed the maximum allowable power consumption.

  • 0 in reply to QJ Wang
    Prodigy 180 points

    Hi QJ,

    You are absolutely right, the linear regulator is going into thermal shutdown. I soldered on a second one in parallel to share the load, and for the most part it has helped. The regulators do shut down occasionally but not as often as before. For my next revision of the board I will add a much higher capacity IC. 

    The main problem I'm having though is that the TMS570 is drawing too much current. Even when paused while debugging (i.e. no code is running) the microcontroller is drawing 350mA+. The code flashed is even just a simple while loop blinking an LED. What could be the cause for such high current draw? Only the GIO driver and SCI drivers are enabled from halcogen, clock is left default at 80MHz, everything else is default. 

    Also when I hold the nPORRST button down on my board, I am still measuring 160mA draw on the 1.2V rail. What do you think could cause such demand for current? Is the IC damaged somehow?

    Thanks,

    Gabe

  • 0 in reply to Gabriel Soares
    TI__Guru**** 197006 points

    Hello Gabe,

    Have you solved the power issue on your board?

  • 0 in reply to QJ Wang
    TI__Guru**** 197006 points

    I assume you have solved the issue. I will close this thread.