STM32H573 / STM32H563 max pin voltage / pins driven in unpowered state / current injection on pins

[bernhardb]

Hello ST,

1. ds14258 Rev.6 Table 57 on p.163 specifies a positive I_inj of "N/A" for pins other than PA4, PA5, PB2, PB12,
PC14, PC15, PD8, and PH2 (as opposed to a positive I_inj of "0 mA" in the row above for pins PA4, PA5, PB2, PB12, PC14, PC15, PD8, and PH2)

 

Does "N/A" mean that it is *impossible* to cause any positive injection (on pins other than PA4, PA5, PB2, PB12,
PC14, PC15, PD8, and PH2) with pin voltages between 0V and 4V when VDD is unpowered or at 0V?

i.e. Do pins other than PA4, PA5, PB2, PB12, PC14, PC15, PD8, and PH2 have some kind of "nice" input structure that allows powering them when VDD is unpowered or at 0V?

 

 

2. ds14258 Rev.6 states in Note 3 on p.128: "Positive injection (when VIN > VDDIOx) is not possible on these I/Os and does not occur for input voltages lower than the specified maximum value."

 

Does "not possible" in the above sentence mean:

a. that externally raising the voltage on those pins to between 0V and 4V when VDD is unpowered or at 0V *will not cause* any positive injection currents to flow into the pin, OR

b. that the voltage on these pins should *never* be raised above VDD when VDD is unpowered or at 0V, since inappropriate positive injection currents would result?

[AScha.3]

You have to see the basic condition: chip is powered, typ. at 3v3 .

>Positive injection (when VIN > VDDIOx) is not possible on these I/Os

So its no injection possible = happening , if you stay below the max. allowed voltage.

You can just imagine the 5V-tolerant pins as not connected to VDD with a "standard back diode" , but with something like a zener diode. So a current is not flowing , if below the limit. Above...might just destroy the pin.

 

>Does "not possible" in the above sentence mean:

-> b. that the voltage on these pins should *never* be raised above VDD when VDD is unpowered or at 0V, since inappropriate positive injection currents would result?

right. No injection allowed.

[bernhardb]

Hello AScha.3,

Thank you for the quick response. To clarify: You seem to be implying that the "N/A" in ds14258 Rev.6 Table 57 on p.163 therefore should rather have read "0mA" - is that correct?

 

AN4899 Rev.3 Section 5.1.2 says on p.13:

Positive-injection current is the current induced when VIN > VDD. For STM32 devices, the maximum positive-injection current on TT and FT GPIO is defined as N/A or 0 mA.
• N/A means that, as long as the input voltage is within the AMR range, and due to the internal design of the GPIO, no current injection occurs. As a result, no corruption of the GPIO and STM32 device operation is observed in such a case.
• 0 mA means that current injection can damage the GPIO and induce STM32 malfunction.

 

Now ds14258 Rev.6 in Table 57 on p.163 uses "N/A" for most pins, but the same document seemingly specifies "0 mA" the same pins in Table 18 on p.128.

 

So I think we still have confusion.

 

This message was hoping to clarify this issue for the STM32H5 family, since there is a lot of inconsistent information on this forum, e.g:

https://community.st.com/t5/stm32-mcus-products/5v-ft-pins-vs-injection-current-on-stm32g0/m-p/797595

https://community.st.com/t5/stm32-mcus-products/stm32-gpios-do-not-have-partial-power-down-ioff-back-drive/td-p/850254

https://community.st.com/t5/stm32-mcus-products/input-signal-when-mcu-is-powered-off/m-p/591270

Super User "TDK" seems quite adamant that it is possible to power the pins lets say to 3V3 when VDD is unpowered or held at 0V.  Is "TDK" therefore not correct, although this seems to be corroberated by AN4899 for FT (and I assume FT_* ?) pins?

 

Sorry to be pedantic about this, but it makes quite a difference to a design if all of these STM pins have to protected by analog switches or not in cases where a user might drive externally-facing pins before powering a module with on-board STM.

[AScha.3]

So at first : "N/A" or "0 mA" (effectively N/A) means you cannot use the feature in that configuration. For instance, positive current injection might be N/A on certain 5V-tolerant (FT) pins.

So any current injection not allowed.

+ on FT-pins : Table 58 , note 5 :

VIN must be lower than Max(VDDXXX) + 3.6 V.

So if VDD = 0 , unpowered, Vin < 3.6V should not kill the chip, if no current flowing .

5.3.14:

As a general rule, avoid current injection to the I/O pins, due to external voltage below VSS or above VDD (for standard, 3.3 V-capable I/O pins) during the normal product operation. 

 

So for a hobby-design ...try it. On FT pins with lets say 100kohm in series, apply 3v3 and test : < 1uA ?, ok.

For industrial design, as always:

- all inputs on optocouplers = no problem with external voltage, powered or not.

- analog inputs : resistor + protection diodes to GND + VDD .

- outputs with opto , transistor buffer or relay.

 

[TDK]

@bernhardb Did you get the answer you were looking for here? If so, can you mark a solution?