Problem with VBAT and RTC

[roby]

Posted on February 05, 2009 at 13:00

Problem with VBAT and RTC

[roby]

Posted on May 17, 2011 at 13:01

Hello,

I'm working with STM32-SK from IAR. I have attached a CR2032 battery to the board, then I have measured the current of battery when the board is off. I measured 1.25 uA. Then I have repeated the measure and switched ON and OFF the board through connect and disconnect the JLink from USB. Now I measured 36 mA! The chip is quite hot but the firmware and hardware is still working. If I test the resistance (without battery) on VBAT pin I measure around 50 ohm and when I switch off the board it lost data from backup domain. What could be happened?

[relaxe]

Posted on May 17, 2011 at 13:01

Simple things first:

When I blow up something on my board, it always tend to be the STM32. A good old overvoltage: bang! STM32 smells and draw all current the board have. Invalid ground loop on the serial port: Bang! STM32 smells and start to heat instead of working. I currently have 3 dead square (pins cut) STM32F103ZET on top of my desk to remind me my errors of the past, plus 2 others I dremeled into oblivion. .... Since the STM32-IAR/SK does not have a CR2032 socket, I suspect you did something wrong, something very wrong. Notice that the LARGE end of the CR2032 is positive, and SMALL end is negative. un-intuitive. Could you have just reverse-connected it? Also, the IAR/SK board have VBAT connected to VCC (3.3V). Maybe: -You actually power the whole board from your battery (since VBAT = VCC) -You actually try to ''regulate'' 3.3V on VBAT, and your battery try to force 3.0V, hence ''recharging your non-rechargeable battery'' for 36ma For more clarity: Could you mesure the TOTAL current your board use from USB(or other source) AND from the battery, running and sleep with and without battery? Current: ===USB=== ===Battery=== USB+BAT_Running USBnobat Running USB+BAT_Sleep USBnobat Sleep -Relaxe

[roby]

Posted on May 17, 2011 at 13:01

Hello Relaxe,

IAR/SK board have VBAT connected to VCC (3.3V) only if you close BAT_E jumper. In my board BAT_E jumper is open. The polarity of battery was OK, since the board work many times without problems. Only now the backup domain switch seem to burn out. It seems to be a power-up problem, but how can I protect STM32 from this strange damage? The STM32 is still working but the backup domain don't work... Best regard, Roby

[relaxe]

Posted on May 17, 2011 at 13:01

Bang! Busted STM32. :-]

Fear not, it is easy to rework. If you need protection from accidental but ''shit happens'' connection, wire a resistor in serie with your positive, limiting forward current. Since MAX current to backup domain is about 2.3uA, I woul limit the current to... 500uA? Making around 6K for a 3.0V CR2032 500uA should not bust any transistor, or so I hope. Wire a capacitor between VBAT an GND to prevent very small over voltage and glitches.

[relaxe]

Posted on May 17, 2011 at 13:01

Also, you DO know there is a procedure required to enable any acess to the backup domain????

If you don't know what I'm talking about, better revise the backup domain part of RM0008.