• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 82 subscribers
  • Views 935 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.

TMS320F28377S: 28377S launchpad heatsink

Anatoliy Gordovskiy
Prodigy 110 points
Part Number: TMS320F28377S


Hi all.

I by matlab simulink embedded make code for F28377S launcpad.

Just read 4 ADC values at 100khz ratio and send to SCI-A at 12mbit, flashing D12 at 100khz ratio. D20 (input mode) by jumper wire was connectad to 3.3V. All works fine, I left board overnight connected only to lab signal generator and oscilloscope (It RIGOL lab equipment). Launchpad was selfpowered from USB. Board work as expected, CPU maybe was around +50C, ambient + 25C.

At morning i want to continue evaluate code, but found CPU are dead, shortclose 3.3V power rail.3.3V power converter at FTDI side are OK, FTDI keep working.

When I evaluate resistanses, I found normally GPIO relative ground has resistance around 1-5 megaoms, just GPIO84 (12mbit TX) 500 ohm.

Maybe it absolutelly unexpacted fault. Or has some limitations for using launchpad? Can`t sent data at 12mbit ratio to onboard isolator infinity time? Need care about CPU temperature?

I confused, purchase new or change to another solution.

  • 0
    TI__Guru 60865 points
    Hi Anatoliy,

    You shouldn't normally need to be concerned about thermal power dissipation for this device when operating it under ambient temperature conditions; even at maximum CPU/peripheral utilization, there should be no way to self-heat the device to anywhere near the maximum rated die temperature.

    Usually if we see that the device is heating up, this is due to accidentally applying an over/under voltage condition to one or more of the device pins. Are you sure that the ADC input pins and the GPIOs were all within 0 to 3.3V nominal input range?
  • 0 in reply to Devin Cottier
    Prodigy 110 points

    Hi Devin,

    Right now i think it possible problem of exact board.

    With bad mood i connect 3.3V from lab power supply (disconnect ISOpower), increase voltage to 4.5V, consuption of 3.3V bus stay 2.5A, I expect saw smoke or self unwelding  CPU. But happens miracle, CPU working (read correct data from ADC and send to computer) under this conditions, over 2 minutes, CPU heat up to 150-200C and current falling to 0.4A.

    I set up back 3.3V (restore ISOpower, remove lab power supply), at reset condition consuption from USB 130ma, when working my code 230ma. By thermal vision temperature of top cover 50C, temperature of chip 55C. Ithink it burn out damaged sub-system.

    But along testing consumption slowly increase, when it stay 660ma onboard  3.3V converter began "flashing" CPU with big heatsink heat up to 66C. If let cool down to 25C and power on again consuption from USB at reset state stay 300ma. If repeat test sessions initial current at reset state unexpected change 100-500ma, each session begin.

    I think it some melting at temperature near 55C inside IC and short close 3.3V power bus. It haven`t affect on A0, A1, B0, B1, GPIO84-86, 20, 12, 13, JTAG which i using in my application.

    Right now i continue evaluate code, but board can`t working more than one hour, even with big heatsink. All time need make pauses for cool down.

    When CPU totally can`t work (due low voltage, caused by overconsuption) picture from thermal vision are same, at first heat up this square, close to Pin1 of CPU.

  • 0 in reply to Anatoliy Gordovskiy
    TI__Guru 60865 points

    Hi Anatoliy,

    How much current consumption are you seeing with external 3.3V supply?  You can see the typical/maximum current consumption in the datasheet here:

    http://www.ti.com/product/TMS320F28377D/datasheet/specifications#SPRS8808739

    I don't think there is anything else on the board that will draw significant current from the 3.3V rail, so if you are seeing more than 500mA of consumption, there is definitely a short/failure in the device or a component on the board.  

    Usually this type of failure is caused by an over/under-voltage event on the power rails or one of the IO pins (digital or analog). 

    I definitely would not expect the device to survive 4.5V on the power rail.   

    I think you need to get a new board.

  • 0 in reply to Devin Cottier
    Prodigy 110 points

    Hi Devin.

    I already book new board, hope it was my mistake like static discharge to GPIO pin.

    Device keep alive even after 5v at 3.3V bus, where declared absolutelly maximum 4.6V. I right now ressurect CPU by this way each time after it out of work. set 5V, wait when current stay normal, power off, cool down, set 3.3V, start test session :)

    Along dubbugging at early stages of evaluation, i missconnect one analog input to onboard 5V, it was jest few seconds. after it that input works well. it can initiate this problem? In the real application will be used amplifiers powered from +/-15V for analog signals, possible unexpected signal magnification if input of amplifier float, for safety reason need add strict limiters to 3.3V or enough just resistors for limit input current of pin less 10-20ma in case under/over voltage?

  • 0 in reply to Anatoliy Gordovskiy
    TI__Guru 60865 points
    Hi Anatoliy,

    5V to the analog input for a few seconds could definitely damage the device.

    In your application you will probably want some means of protecting the input from over/under voltage. To prevent damage, it should be possible to add series R to limit the total clamping current to 20mA. This is for all pins, so if multiple pins could go over/under voltage simultaneously, you'll need to take that into account. This may be difficult with +/-15V drivers; R to limit -15V to 2mA (assuming 10 pins) would be (-15V - -0.3V)/-0.002 = 7350 ohms, which is probably too high to drive the ADC well. You might need to add external clamping diodes or drive the ADC inputs with amplifiers with a 0 to 3.3V range.
  • 0 in reply to Devin Cottier
    Prodigy 110 points
    Hi Devin

    Thanks a lot, will try when arrive new board.