Pin F0 strange behavior due to NMI on TM4C?

Lee Shaver

I already solved one problem with tricky pin F0, getting some help here to disable NMI. Now it seems there is another problem. I am using this pin to operate a relay, and whenever my program stops, NMI is re-enabled, F0 goes high (I think - going to test with a scope tomorrow) and the relay is closed.

Here's the code I'm using:

	ROM_SysCtlPeripheralEnable(SYSCTL_PERIPH_GPIOF);	// Pins: RGB LED, Relays
	
	ROM_GPIOPinTypeGPIOOutput(GPIO_PORTF_BASE, GPIO_PIN_1|GPIO_PIN_2|GPIO_PIN_3);	// RGB LED

	// Port F unlock procedure:
	HWREG(GPIO_PORTF_BASE + GPIO_O_LOCK) = GPIO_LOCK_KEY;		// Unlock the port
	HWREG(GPIO_PORTF_BASE + GPIO_O_CR) |= GPIO_PIN_0;		// Unlock the pin
	HWREG(GPIO_PORTF_BASE + GPIO_O_AFSEL) &= ~GPIO_PIN_0;  
	HWREG(GPIO_PORTF_BASE + GPIO_O_DEN) |= GPIO_PIN_0;
	HWREG(GPIO_PORTF_BASE + GPIO_O_LOCK) = 0;			// Lock the port

	// Then turn it on:
	ROM_GPIOPinTypeGPIOOutput(GPIO_PORTF_BASE, GPIO_PIN_0);		// Rel1 (conflict with USR SW2)

Am I doing something wrong? Is there something else I should be doing to keep my relay from closing each time the program runs?

over 9 years ago

0 Robert Adsett over 9 years ago

Guru 27665 points

Do you turn the relay on or off when you stop?

Robert

0 Amit Ashara over 9 years ago

TI__Guru**** 244400 points

Hello Lee

NMI is an input and not an output. So how do you know it is re-enabled?

Regards
Amit

0 Robert Adsett over 9 years ago in reply to Amit Ashara

Guru 27665 points

Change from driven to floating could allow detection of change from output to input.

Robert

0 cb1 over 9 years ago in reply to Robert Adsett

Guru 47900 points

Simpler, easier, FASTER - AVOID use of such (potentially) troublesome pins!

KISS - simplest, most direct ALWAYS "Fastest, Easiest!"

0 Amit Ashara over 9 years ago in reply to Robert Adsett

TI__Guru**** 244400 points

Hello Robert,

Poster's illustration of the case is not clear.

Regards
Amit

0 cb1 over 9 years ago in reply to Amit Ashara

Guru 47900 points

And - as always - he who, "Opens the door to difficulty" should express, "No surprise" when it enters...
So many pins available - why choose a challenging one?

0 Lee Shaver over 9 years ago in reply to cb1

Intellectual 390 points

cb1 - Of course, not using this pin would be my first choice - but the circuit is made and the boards are printed, so until the next rev I have to live with it.

Amit - I assume NMI is re-enabled because if I run the unlock code once, I can use the pin. If I stop the program and run a different program without the NMI un-lock code, I can't use the pin. However, perhaps stopping and re-starting the code this way wouldn't be a good test? I could set a flag in EEPROM to run the unlock code on the first pass, and not on the second, to see if this makes a difference (rather than compiling/uploading different code).

Robert - I don't change the relays when I stop the program. They are latching so should stay in whatever state they were last set until the program starts again. Eventually I will store this state in EEPROM and retrieve it on the next run, but during my testing I'm avoiding use of the EEPROM.

One other note - on my dev board - TM4C123GXL - PF0 is wired to user switch 2. I'm not sure if this would have an effect or not.

0 Amit Ashara over 9 years ago in reply to Lee Shaver

TI__Guru**** 244400 points

Hello Lee,

Doing a Soft Reset of the core will reset the peripheral to the power on reset values. By Core I mean using something like System Reset and not CPU reset

Regards
Amit

0 Robert Adsett72 over 9 years ago in reply to Lee Shaver

Guru 10570 points

Probably not relay kickback then.

Can you tell if the program restarted? If not, can you wiggle a pin or something on start to see if you are resetting?

Robert

0 Lee Shaver over 9 years ago in reply to Robert Adsett72

Intellectual 390 points

OK, so I made some discoveries. I was having the problem on another relay also, on pins F4 and E0. I fixed the problem on F0 by removing the 0 ohm resistor that connects it to switch 2 on the launchpad.

Is there anything about pin F4 I should know about? I know it's also attached to a switch, switch 1, but taking out the 0 ohm resistor there doesn't make a difference. The relay still toggles on whenever I re-start my program.

0 Lee Shaver over 9 years ago in reply to Lee Shaver

Intellectual 390 points

I fixed the problem. On PF0, I needed to remove R1.

On PF4, there was enough noise going to the MOSFET I used to drive the relay that it would switch on. This was solved by adding a resistor (25kohm - what I had handy) from PF4 to ground.

In the next version of the hardware design I will include pull down resistors on all relay signal pins.

0 Amit Ashara over 9 years ago in reply to Lee Shaver

TI__Guru**** 244400 points

Hello Lee,

The launchpad is not a perfect platform for development as it contains its nuances (so much recorded in the forum)

Regards
Amit

Arm-based microcontrollers

Arm-based microcontrollers forum

Pin F0 strange behavior due to NMI on TM4C?