cancel
Showing results for 
Search instead for 
Did you mean: 

STM32F407 MCU Maximum Voltage Input

IKhan.2
Associate II

Hi All

Maximum input voltage for this series of MCU per Table II of the datasheet is specified to be 4.0 VDC for all pins other than five-volt tolerant. Table 12 specifies the maximum injected current for all pins other than five-volt tolerant to be +/- 5 mA.

Given the above specifications would it be permissible to exceed the maximum input voltage if the injected current is limited to lets say 0.1 mA by a resistor in series with the non five-volt tolerant I/O pin? If not what would be the reason?

Regards

ik

1 ACCEPTED SOLUTION

Accepted Solutions

Ok, this is not over max. ratings , so cpu should be ok with this.

But anyway you go over limits of "normal" working conditions and use the protection diodes.

And you get "side effects" : any injected current might have strong evidence on other pins on same port or internal multiplexer, maybe spoiling all ADC values on all other pins, when injection happens. So its just : not killing the cpu, but dont expect it working "normal" .

 

Why not simply stay in recommended area, it is not difficult:

AScha3_0-1719554023862.png

So with 3k3 signal at cpu pin is 0...3 V (2.97V at 12.0V in ),

and protection is for any "accident" , like 15V coming to the input.

If you feel a post has answered your question, please click "Accept as Solution".

View solution in original post

5 REPLIES 5
AScha.3
Chief

Hi,

sure, you can try this - it will be an avalanche breakdown in the port/mosfet/protection diode , maybe at 6 or 7 V (i tried...with a cheap chip, some limit at 12V - and survive !); if the device is still ok, it will work , but maybe its "sensitive" , you kill it with just going 0.3V ( at 60 uA ) over the stated max limit (i tried...also on some), it will never work again, as it should. If this is just a "not important" pin, disable it and can use the remaining pins. If it is a pin you need - exchange the chip and be more to the specs next time. 🙂

So:

A: want some game, see the limits and the result...try everything, maybe funny.

B: you have to build a reliable thing, with some years warranty ... keep always below max. ratings. 

If you feel a post has answered your question, please click "Accept as Solution".
IKhan.2
Associate II

Hi

To be more precise review circuit below and provide comment:

 

IKhan2_0-1719528315747.png

Regards

ik

If the pulse is of very short duration it may be ok. I have seen zero crossing detection in microchip app notes. It's best to use an opto isolators  or zener diode . 

Ok, this is not over max. ratings , so cpu should be ok with this.

But anyway you go over limits of "normal" working conditions and use the protection diodes.

And you get "side effects" : any injected current might have strong evidence on other pins on same port or internal multiplexer, maybe spoiling all ADC values on all other pins, when injection happens. So its just : not killing the cpu, but dont expect it working "normal" .

 

Why not simply stay in recommended area, it is not difficult:

AScha3_0-1719554023862.png

So with 3k3 signal at cpu pin is 0...3 V (2.97V at 12.0V in ),

and protection is for any "accident" , like 15V coming to the input.

If you feel a post has answered your question, please click "Accept as Solution".
IKhan.2
Associate II

Hi

Thank you all for good input.

Regards

ik