EK-TM4C123GXL: Boards blowing up!

Hi Vito,

Can you elaborate what do you mean by 'Boards Blowing Up'?

Can the boards still power up?

Do you see any LED lit up on the board?

Can you do a resistance check on these GPIO pins you use? If you do a resistance check between these pins and if the resistance is low then there is a short somewhere.

Vito Ignazzi said:

I've managed to blow up three EK-TM4C123GXL development boards and my stock is running low. I'm applying 5V to various GPIO pins, namely PA2, PA3, PA4, PA5, PA6, PA7, PC4, PC5, PC6, PC7, PD6 and PD7. All the GPIO pins in question are declared 5 V tolerant. (I'm aware that PD4, PD5, PB0 and PB1 are not 5 V tolerant and have avoided these pins.)

Yes, these named pins are 5V tolerant.

Vito Ignazzi said:

At first I thought, from the "GPIO Pins With Special Considerations" section on p650 of https://www.ti.com/lit/ds/spms376e/spms376e.pdf that perhaps PA2 to PA5 are set as SSI0 functionality at startup. Accordingly I've added the following lines of code to belt-and-braces ensure that all the GPIO pins are set to input:

    GPIODirModeSet(GPIO_PORTB_BASE, GPIO_PIN_3, GPIO_DIR_MODE_IN);

    GPIOPinTypeGPIOInput(GPIO_PORTB_BASE, GPIO_PIN_3);

    GPIOPadConfigSet(GPIO_PORTB_BASE, GPIO_PIN_3, GPIO_STRENGTH_2MA, GPIO_PIN_TYPE_STD_WPD);

I don't see a problem with your configuration. 

Vito Ignazzi said:

I wonder if the SSI0 is still in the equation.

Vito Ignazzi said:

Have I done everything to ensure the pins are inputs?

Have you applied 5V to these pins PA[5:2] before you change these pins to GPIO input mode? That is something that comes to my mind that may potentially create drive conflict if you drive 5V to these pins before they become GPIO inputs. 

Can you make sure no voltage is supplied to any GPIO pins before they are configured to inputs. Please use a scope to confirm. 

Vito Ignazzi said:

Is there a known problems with any of the listed 12 GPIO pins and 5 V tolerance?

Based on my current understanding of your problem description, I don't think GPIO#10. You should be able to recover from this errata through a power up and the suggested workaround. However, you seem to suggest the board is now dead. 

By board blowing up I mean the GPIO pins in question cause the board to stop working when 5V is applied to the pins. Stop working means my terminal connection is lost and the RGB LED goes off. My 5V comes from a BCD switch. When the switch is moved off the zero position 5V is applied to various GPIO pins. That's when the board shuts down. 

If I don't apply 5V to the pins in question, the board works OK, but does run hot (one very hot). I suspect that the GPIO pins in question have been damaged and are causing the board to run hot. 

My main concern is that I cannot read the BDCD switches because applying 5V to the GPIO pins stops the board from working. I don't think the BCD switches would introduce fast transient, but who knows.

Vito Ignazzi said:

By board blowing up I mean the GPIO pins in question cause the board to stop working when 5V is applied to the pins. Stop working means my terminal connection is lost and the RGB LED goes off. My 5V comes from a BCD switch. When the switch is moved off the zero position 5V is applied to various GPIO pins.

Don't you have series resistor between the BCD switch to GPIO? Can you show your schematic?

Vito Ignazzi said:

If I don't apply 5V to the pins in question, the board works OK, but does run hot (one very hot). I suspect that the GPIO pins in question have been damaged and are causing the board to run hot. 

Have you done resistance check as suggested?

I have done the resistance check and, with the EK-TM4C123GXL unplugged, all lines to the BCD switch are open circuit.

I do not have any series resistance.

The BCD switch is a simple BCD switch https://www.farnell.com/datasheets/3675103.pdf .

I suspect with the 12 GPIO pins listed connected directly to 5V (or just one set, say PC4, PC5, PC6 and PC7), then each and every time the EK-TM4C123GXL will "blow up". I will performing this check when I get more EK-TM4C123GXL boards to sacrifice.

In the mean time, I will disable three switches' functionality and get the product to test without this capability.

Hi Vito,

  Glad your issue is resolved. 

Vito Ignazzi said:

I'm going to declare that at least one GPIO pin from each of the three sets below are not 5V tolerant.

1) PC4, PC5, PC6, PC7

2) PA4, PA5, PA6, PA7

3) PA2, PA3, PD6, PD7

Not according to the datasheet. 

All GPIO signals are 5-V tolerant when configured as inputs except for PD4, PD5, PB0 and
PB1, which are limited to 3.6 V. See “Signal Description” on page 649 for more information
on GPIO configuration

I think what might have happened is that when you are in position 0, you are driving 5V to all four pin inputs. If you have three BCD switches, you could be driving12 inputs directly with 5V. Without current limit resistors on these inputs, you can exceed the current limit. 

Once I get my extra EK-TM4C123GXL boards I will do an experiment.

1) After power up, apply 5V to each of the pins PC4, PC5, PC6, PC7, PA4, PA5, PA6, PA7,  PA2, PA3, PD6, PD7 in turn. That is, one pin at a time. Stop after a board failure.

2) While starting up, apply 5V o each of the pins PC4, PC5, PC6, PC7, PA4, PA5, PA6, PA7,  PA2, PA3, PD6, PD7 in turn. That is, one pin at a time. Stop after a board failure.

I will report back the findings.

I finally got around to testing the LaunchPad. I did the tests detailed in my previous post as well as resetting with 5V applied and tying all four pins in each bank to 5V while powered up, after rest and power cycle. The problem I observed can not be reproduced. However, I have a couple of LaunchPads that the microcontroller runs warm/hot!

In the end, the problem cannot be reproduced (on the Tiva LaunchPad).

Hi Vito,

  Thank you for the update.