STM32F411 Output pin connection while no VCC is applied

[Nir Alon]

Posted on April 01, 2018 at 13:22

Hello all,

In my application, I'm controlling a

secondary 

device using enable pin which has an internal pull-up resistor.

The output pin is set to be ''0'' or open drain type output, and the STM32F411 VCC1 can be powered down while the

secondary 

device's VCC2 is still active.

Can I connect the output pin directly to the secondary device enable pin or should I place a Tri-state buffer between them?

Thanks!

Nir

[Andrew Neil]

Posted on April 04, 2018 at 11:12

It means that the STM32 must be powered.

So the answer to your original question:

Can I connect the output pin directly to the secondary device

is, 'No'.

[Andrew Neil]

Posted on April 04, 2018 at 11:37

 ,

 ,

No.

See: ,

https://community.st.com/0D70X000006SsgsSAC

 ,

[LMI2]

Posted on April 04, 2018 at 18:11

Where should that link point to? Nothing happens

[Dvorak.Peter]

Posted on April 04, 2018 at 20:12

Simple 

Change the polarity of logic and use input pulldown. For extra safety use 2K series resistor.

Peter

[T J]

Posted on April 05, 2018 at 00:37

then I think the Output will be always on while the processor is asleep,

that's why my original post had a transistor, relieving all issues..

[Simon Peacock]

Posted on April 06, 2018 at 04:23

1/ When the STM32F411 powers down, its input diodes will try to power its internal PSU circuits.  This shouldn't be an issue as the 411 will take more current than is available.  You may however get hick-ups... the resistor charges the capacitors around the 411 to a point where the 411 turns on.. it then starts, drains the capacitors are stops again.. only to charge and start again (this mostly happens at higher temperatures).

2/ It might be wise to use an external pull-up due to internal semiconductor resisters being highly variable.  Even better, when the 411 is powered down, turn the pull up off to avoid any issues.

3/ Option C... external resistor pulled up to the 411's power rail, 3 V will still turn on the 3.3 V micros logic

4/ Option D... no resistor, totem-pole the output 3V into 3.3V is cool, pull down for safety (if required)

Simon

[Daniel O'Neill]

Posted on April 07, 2018 at 09:54

Hi Nir,

I have done things like this and it does meet spec. Note that there is no input current spec with this, no current will flow as pin doesn't have a diode to Vdd.

Note 1 is meet if the supply is at 0V (see Vdd-Vss spec above that) and you are only using a 3.3V signal.

To convince yourself that it is fine try it out. Measure the voltage drop across a 1k, from 3.3V into the pin.

A lot of devices don't allow this but the STM32s are pretty good like this. The real tricky thing is that the FT (5V tolerant) pins are only rated to Vdd+4.0. If your supply is off you can't have 5V on the pin.

Daniel

[Nir Alon]

Posted on April 08, 2018 at 08:10

Hi Daniel,

The external pull-up that is connected to the other MCU's 3.3V is a 47K, so in worst case, there would be 3.3V/47k = 70uA flowing into the local MCU IO.

I've tried building this configuration and measured 3.3V across the external pull-up, it confirms that there is almost no current that the IO in drawing.

(Of course what while doing the measurment the local MCU's VDD was off.)

Thanks,

Nir