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.

My Stellaris Launchpads LM4f 120 chip suddenly started getting really heated

Other Parts Discussed in Thread: LM3S8962, LM3S6965, LM3S3748

Since yesterday, My stellaris Launchpads LM4f120H5QR chips is getting extremely heated.

I checked changed the board to my friends board, that board doesn't get heated for the same code.  I tried to check if there are any short connections, there are none. 
I dont undestand why it gets heated.

its a new board. I cant seem to find why my board gets heated. I have not connected anything to that its pins. Its simply running a simple code with a couple of lines..

 
I noticed when i remove the code and empty the code memory and switched it on, the heat it less. But its still more than my friends board.

What should i do?

Should i continue to use it even though it gets heated?

  • According to document "Getting Started with the Stellaris EK-LM4F120XL LaunchPad Workshop" you will find on the page 1-3, that LM4F12x is under the "sampling state". So it is not in the "production state", which means that malfunctions may appear on the sampling phase of the production.

    And if you look your board with magnifying glass, you will notice, that your microcontroller is typed as : LX4F12x.... Not as the production and final version will be labeled as LM4F12x...

  • Dear Pentti,

    I've been using TI's ICs for a really long time. I do not think this is a malfunction of an the ARM chip. it cant be.  but it seems like i cannot use it any longer. I dont want t draw so much current from my USB it might damage my computer. right?

  • Makes sense to find your board's schematic - and then confirm that the MCU's 3V3 Regulator is working correctly.  While the voltmeter is out - you should confirm that the 5V from the USB cable is correct.

    Note that there have been reports of bad USB cables being supplied.  Wise to substitute a high quality cable - just in case.

  • I've been engineering all kind of electronics and sofwares for 30 years. But if your lm4f board is "under the development phase" how could wait for full functionalities for the stelallaris lm4f microcontroller board product.

    It is normal life that products under the development phase do not behave as final products. And as i mentioned on my previous message, check your product's type - is it  LX... or LM labeled. LX is development version and LM is the final version when product is ready for sale. I am sorry for my poor english.

  • I checked on my board. It says LX. not LM . I do not understand. TI said it was giving a LM4F chip when I bought the development board. Can anyone explain as to why i have received a LMX soldered to my board?

    And I also checked the Vin to the chip. it is exactly 3.3V. So I think the voltage input to the chip is correct. I double, tripple checked. I cannot seem to find a place of S/C or a crack. 

    What should I do?

  • I also checked with another USB cable. the same problem remains. my friends board seems to work with no problem for the same code though.

  • I bought two of the Launchpad units. I had two intermittent cables. (Both units)

    I also had to re-solder the connections on all USB connectors. Luckily I have a large Inspection unit -- Magnifying Glass with Fluorescent light.

    No problems now -- but I did not notice overheating.

  • isnt that bad for TI? I mean, I got my self and even recommended thinking TI stuff is quality. Im starting to think otherwise... Im seriously confused. 

  • There usually are multiple Vdd and Gnd pins on ARM MCUs.  You must be careful - it may be that one or more "lifted" from their pcb trace.  I'd list each/every one - and then carefully probe for continuity to Vdd or Gnd.  The VddA (analog) pin also must be connected to 3V3 - and so the GndA pin to Gnd. 

    Suppose its possible that one of your xtal caps or the main Osc. xtal trace opened - causing MCU to over-clock - which would draw more current.

    Unlikely that TI assembled these boards - their charge for these is small enough that most would do well to purchase several - just in case such incidents occur.

    Update: You do realize that ESD is an unfriendly/unwanted "killer" of such fine-geometry, modern ICs.  From your writing - it would appear that your finger is your, "heat sensor" and if your habit is regularly to "touch" the ICs - ESD rises much higher in the diagnostic que! 

    While you state that "nothing" is connected to your board now - perhaps the (however brief) introduction of some voltage outside the clearly stated limits of this ARM MCU was the cause.  (only you know)

  • isnt that bad for TI? I mean, I got my self and even recommended thinking TI stuff is quality. Im starting to think otherwise... Im seriously confused.

    Welllllll Yes, no, hmmm... I guess, maybe so.

    On the other hand -- that is what a "test product" is all about.

    Very few people or design teams get it right the first time.

    That is why you build a prototype or run a model whenever possible.

    Just remember that hard data wins over models any day -- this is not climate science... this is hard science...

  • While you state that "nothing" is connected to your board now - perhaps the (however brief) introduction of some voltage outside the clearly stated limits of this ARM MCU was the cause.  (only you know)

    Lucky me -- I sweat easily.... I blow very few IC's... Others with dry skin -- or who love silk are truly dangerous... give up your Jimmy Choos!

    Cottons, leather shoes, eschew cheap polyester suits, be nervous, sweat -- it works for me... (roflmao -- even if it is true...)

  • 1) Program the board with some example code such as project0.

    2) Remove the VDD jumper and use a multi-meter in current measure mode to measure the current into the LM4F device.

    If it is more than about 40 milliamps then something is conducting either internally or externally to the LM4F.  If no visible external shorts are found after extensive search then likely the short is internal and caused by ESD or EOS (electrical over stress).  In that case some pin or pins is probably no longer functioning.

    If indeed you have an ESD or EOS damage then if the loss of function is acceptable you can continue to use the device.  If not you will have to replace the board or the chip.

    Dexter

  • @ Stellaris Dexter,

    Earlier - suggested ESD as likely cause - if "out of spec" external sig. was not introduced...

    That said - is there merit to such MCU current measurement with the MCU fully erased?  (one would think that this condition would be more "repeatable" chip to chip - and may provide more consistent insight...)

  • Doing the current measurement with a fully erased device should be very similar to running an example code.  

    On a fully erased Stellaris the core will be executing ROM code in the bootloader waiting for packets to come in from any of the available bootloader interfaces.  In this case the expected current is probably slightly lower since ROM access should be lower power than flash access.

    Dexter

  • Thank you everyone for participating in this forum to help me. 

    I have no intention to replace the board or the chip. Indeed i dont have the tools to do that. I mean, to get my self a LM4F chip to SL the shpping cost would cost me 4 times! more than getting a development board. Yes believe me!

    so im going to live with this. I accept that I touched the IC. but i didnt know it would burn it.  I can assure you I didnt give it an ' out of the specification' input. I understand that I should not connect a direct voltage onto a i/o pin.

    Now im planning to get some usefull work out of this board. 

    I appreciate the advices given to me to check for S/C and removed paths but i cannot find any even with a magnifying glass. 

    When I remove the Jumper, the board stops heating. 

    That means there is a chance that the programmer circuit is working. So im planning to do this now.

    Im going to remove the jumper and keep it forever and try to use this programmer (top part) to program my LM3S chips I have at home. Namely 

    1. Stellaris® lm3s8962 Microcontroller

    2. Stellaris® lm3s6965 Microcontroller

    3. Stellaris® LM3S3748 Microcontroller

    this board has allowed 8 pins to be taken out of the board through pads( i can solder wires to these) from the programmer circuit module. (the top cct)

    namely,

    1. TCK

    2.TMS

    3. TDO

    4. TDI

    5. EXT

    6. TXD

    7.RXD

    8. Reset

    do you think I could use this board now to 'act' like a JTAG. because i dont have one to my self now. I would like to make some useful work out of this board.

    I am talking about exactly the process explained in here.

    http://processors.wiki.ti.com/index.php/Stellaris_LM4F120_LaunchPad_Debug_How_To

    thank you.

  • Amalinda said:
    When I remove the Jumper, the board stops heating. 

    This is to be expected - as Dexter explained - you've removed the power path to your MCU by so doing.  He thought (and most here agree) that making a simple current measurement - would give better insight into the, "degree of damage" to your MCU.

    Danger in "touching an IC" is the electro-static discharge.   Ground straps are mandatory for personnel working on/around such boards - we've measured many thousands of volts from the human body.  (low humidity is greatest concern - surely you've noted "shock" - seen discharge when you touch a metal door knob)  Such interaction with modern, fine geometry ICs is never good.

    Sorry - I am w/out comment as to the ability of your board to serve the broader debug/program purpose you seek... 

    BTW - doubtful that too many "recognize" - know just where "SL" is.  (from your far earlier post, Sri Lanka?)

  • Yes, you should be able to use the Launchpad as a debugger for those other Stellaris MCU's.  The wiki link you have already mentioned is the correct place to start.