32F413 Port Pin States During Deep Sleep?

[JOLSO.11]

I'm the hardware guy on a 32F413 project where we have a lot of impulse noise nearby on the PCB.  Because of this I've been on a campaign to make certain that all microcontroller I/O pins are equipped with pull-up or pull-down resistors.

The project has reached the stage where our quality assurance folks get cranky whenever I try to add a resistor to this PCB.  So when I find a 32F port pin that is missing a pull up/pull down resistor (as I did this week) it becomes a crisis of meetings and memos.

Remind me to never work again in a medical equipment company.

So I approached our firmware guy and asked him to alter his code to make certain the offending port pin is equipped with either an internal pull-up or a pull-down during his sleep mode,

Firmware guy replies:  "No can do, hardware guy.  The micro is in its deepest sleep state and the port pins flap in the breeze during deep sleep.  You must fix that undefined I/O state with an external resistor. Luv ya, big guy..."

Is this true?  Do all port pins go to tri-state mode during deep sleep?

[AScha.3 · Accepted solution]

>The micro is in its deepest sleep state and the port pins flap

Depends on the sleep mode , if its Standby ...then, from ds:

and from rm:

So some pins still working (to wake up cpu) , but all other ... hi Z , off.

So better put a pulldown/-up on every output pin, that would float otherwise.

I do this always on "important" output signals, just because on every start/powerup the cpu is in reset and all (almost all) pins floating anyway. 

(Input pins should anyway have some fixed state, from the circuit the signal is coming from.)