Vdd being held at 0.7V with backup battery fitted

[paulyoung9]

Posted on March 08, 2012 at 14:07

Hi,

When I switch off the power to my STM32F417 with a backup battery connected, the Vdd rail doesn't drop below 0.7V and Vbat falls rapidly. It looks like Vbat is connected to Vdd and is powering the rest of the circuit. If I briefly short Vdd to ground, it seems to sort itself out.

I tried adding a resistor to load up the Vdd rail to see if this would help, but I just got lower voltage (~0.4V) and faster discharge of the backup battery. Seems like it needs a fairly hard short to ground to make it switch off.

Has anyone got the battery backup working on STM32F4?

Do I need to do some configuration in software before this will work correctly?

Thanks.

#battery-backup-rtc #inmates-running-asylum #battery-backup #stm32f4 #vbat-stm32f4 #vbat-battery-backup #vbat-stm32f4 #0.7v-rtc-vbat-battery-backup

[bolmsted]

Posted on June 02, 2012 at 03:48

Does anyone else have this problem or know of a fix for it?

I have been able to confirm that I am seeing this exact same behaviour.

It has the appearance of a hardware problem with the STM32F4 and I checked the errata and it is not in there.

Thanks!

[paul-arthur]

Posted on June 07, 2012 at 10:25

Hi,

I've got the exact same problem,

[mm_it]

Posted on June 13, 2012 at 17:28

Hello,

I have the same problem.

Dear ST, when you need a demonstration board, I can send you two pcb.

But, please help us.

My Demo-Program:

void main(void)

{

while(1);

}

without startup-code.

Regards, Markus.

[frankmeyer9]

Posted on June 13, 2012 at 21:07

It has sometimes been mentioned here - this forum is NOT necessarily monitored by ST people.

Better contact your local representative / distributor directly in this case...

[bolmsted]

Posted on July 04, 2012 at 21:18

An update on my experience with VBAT...

My original board had a LTC3973 switching power supply feeding 3.3V. I disconnected and replaced with a linear supply at 3.3V. No change.

I am always able to ''fix'' the problem by shorting. I have also tried this on two different boards.

I am using a 100pin STM32F4-417 with silicon revision ''A''.

I will not use this part in future designs requiring battery backup.

[zaurozavr]

Posted on July 05, 2012 at 00:16

Have you played with the backup initialization?

/* Enable the PWR APB1 Clock Interface */

RCC_APB1PeriphClockCmd(RCC_APB1Periph_PWR,

ENABLE

);

/* Allow access to BKP Domain */

PWR_BackupAccessCmd(

ENABLE

);

[bolmsted]

Posted on July 05, 2012 at 01:24

I verified those bits are enabled.

I have not had problems with the RTC or backup SRAM. They just work. My (and others) specific problems are that the backup battery drains very quickly because the VBAT switch is ''half switched'' powering the 3.3V line with 0.7V. Shorting the 3.3V line to ground always fixes the problem. The switch then fully switches and the 3.3V line is at 0V.

While I have a workaround for testing. Installing a shorting switch is not an option.

[Tesla DeLorean]

Posted on July 05, 2012 at 02:47

Do you happen to have anything going on with PC13, PC14, PC15 and PI8?

[bolmsted]

Posted on July 05, 2012 at 14:53

PC13 - unconnected - input/2M/pullup

            I also tried floating on this pin.

PC14 - ABS25 32.768kHz crystal - input/2M/floating

PC15 - other side of crystal - input/2M/floating

PI8 - this device does not have this pin - registers are set to defaults of zero

The crystal works. I have left the device for weeks sitting and when I restarted the time was within seconds of the correct time.

[Tesla DeLorean]

Posted on July 05, 2012 at 15:42

Ok, just trying to ascertain if there is an accidental conduction path between the backup domain and primary supply or IO ring.

A number of people have been reporting issues here, at least enough that would worry me about a potential problem, surprised ST hasn't commented on this at all, and it's not in the errata.

ST: A lot of us would appreciate it if you participated in your own forum.