Using a MSP430-2272, during LPM3, the current on Vcc of the MSP430 reaches 20uA. This current is caused by I/O P4.5 on which are serially connected a LED and a resistor, the anode LED connected to Vcc.
P4.5 is input, P4REN=0, P4SEL=0.
On the opposite, when I configure P4.5 as an output, level high, this leakage disappears, my whole application consuming only 3.2uA.
I’m satisfied cause the current is sufficiently weak to allow my application (especially RTC) to work long enough during power down.
The problem is that I didn’t precisely identify the leakage origin and I’m afraid of a pending problem.
It appears that your leakage current is due to the pin being configured as input and as such will tend to “float” since the input is not connected to VSS or VDD. This will cause switching currents to flow as the voltage on the pin drifts into the linear region of the input FET. Your best bet is to configure as an output and set the output high as you have seen.
Nevertheless, my surprise as I measure this “huge” leakage current of about 20uA is accentuated by wthat I read on the datasheet dedicated to the MSP430-2272:
"High Impedance leakage current on P1, 2, 3, 4 = 50nA MAX !
assumed:
that the pin is connected to Vss or Vcc
pull up or down resistor disconnected"
I checked those 2 points and also measured 20uA, so…
Perhaps there is some leakage via the LED. One way to determine this is to remove the LED and tie the I/O pin high or low (configured as input, of course!) and see if your current is where it should be.
Another possible cause could be flux on your board.
If I’ve understood correctly at that pin you’ve got a led with a resistor in series, to vcc, is it right?
In this case the pin is not directly pulled up, since the led has a ~1.8V forward voltage. My suspect is that if the device is powered @… let’s say 3.3V, then 3.3-1.8 ~ 1.5, and that’s enough close to VDD/2, which is the voltage at which the digital mos input circuit consume the max current. How to check? add a resistor with a value 100K… 1M directly between the pin and VDD, leaving the led and the other resistor unchanged, and try again. If the leakage current reduces to 3uA the problem was the led
