TM4C123GH6PM: HIB Won't go low

Hello Joseph,

On the TM4C123 LaunchPad the VBAT, VDD and VDDA all are connected to the same PCB trace for supply. So by using the HIB pin to control VDD and VDDA, the VBAT is also dipping. This is what I suspect may be causing the Hibernate Module to reset and wake up the device.

I may be unclear on the connections you may have on the Power Supply and how the switch feedbacks to the LaunchPad itself, so you may want to draw a circuit to show the /HIB and TM4C123 power supply is getting connected?

Regards

Amit

Amit,

Thank you for your reply.

I am aware of how the VBAT pad is connected on the LaunchPad.  I lifted pin 37 (VBAT) and this is connected to the battery supply.  This supply is a switch that allows main power to power the VBAT pin via a low-Iq regulator, or a battery.  The VBAT pin is steady at 3.3V and that part of the circuit is not affected by the HIB pin (this has been verified with a DMM).

The schematic for the TMC123 Launch pad indicates that GNDX is not connected to the board ground (however, it appears to be connected internally to the part).  I measured the voltage at GNDX and it is solidly 0V with no disturbance when hibernate is activated.  A comparison of the datasheet pin listing showed no other VDD or GND differences between what I would have connected and what the LaunchPad has connected.

I have replaced the TMC123 part twice and the issue remains.  At the moment, the HIB pin is not connected to anything (the pin is lifted off of the PCB pad).  The MCU software is set up to configure the HIB module at reset, then delay for 10 seconds before initiating the hibernate mode.  During the 10 second wait period, HIB is at 3.3V.  When the hibernate mode is activated, the HIB pin only drops to 2.8V.  My initialize code is as follows:

	// if not already done, init the HIB module first thing...
    SYSCTL_RCGCHIB_R |= 0x01;
    ui32Loop = HIB_CTL_R;
    if((ui32Loop & HIB_CTL_WRC) == 0) ipl = 0xff;	// set HIB error
    else{
    	if(!(HIB_CTL_R & HIB_CTL_CLK32EN)){
    		HIB_IM_R = HIB_IM_WC;						//
    		while(!(HIB_CTL_R & HIB_CTL_WRC)) ui32Loop = HIB_CTL_R;
    	    ui32Loop = HIB_CTL_R;
    		HIB_CTL_R = HIB_CTL_CLK32EN;				// enable HIB clock
    		ipl = 1;									// set initial power on
    	}
        ui32Loop = HIB_CTL_R;
    	while(!(HIB_CTL_R & HIB_CTL_WRC)) ui32Loop = HIB_CTL_R;
    	HIB_CTL_R = HIB_CTL_PINWEN | HIB_CTL_CLK32EN;	// enable /WAKE control
    	while(!(HIB_CTL_R & HIB_CTL_WRC)) ui32Loop = HIB_CTL_R;
        ui32Loop = HIB_CTL_R;
    }

The many ui32Loop queries are for debug purposes.

The HIB initiate code is as follows:

		if(hfon_timer == 0){
			puts0("Power-off..");
			wait(10);									// give the UART time to finish
			ii = HIB_CTL_R;
			HIB_CTL_R |= HIB_CTL_HIBREQ;				// initiate hibernate
			for(ii = 545 * 10000;ii!=0;ii--);			// delay abt 2 sec to give PS time to discharge
			puts0("..failed.");							// if we get here, it didn't work
		}

hfon_timer is a uint16_t variable that is decremented by Timer0_A (this is verified as working).  The #defines for the register names and bit-masks are from the tm4c123gh6pm.h header file.

Other than replace the TMC123 chip yet again, I am at a loss...

Hello Joseph,

Indeed, get a scope, as I was thinking all this time how could the /HIB not go to 0V. I can start re-re-working my LaunchPad as well.

Regards

Amit