BOOTING NUCLEO BOARDS at 4 Volts 150mA max

[rich.g.williams]

When I started working with NUCLEO boards I was very cautious about powering the board up from 5V let alone 12V even powering the board via STLINK USB at 5V (maybe as much as 250mA)

So I did some reading. On my NUCLEO-H5E5ZJ :-

1. The buck regulator ST1S40IPUR  (U16) becomes transparent at under 5V - it just conducts
2. The linear regulator for main 3V3 is LD39200PU33R (U15) and has very low dropout
3.  Most of the STM32H5E5 pins have Absolute Max tolerance rated at 4V
4. If not being used I could put the STLINK into reset via a jumper (less current to board)
5. Avoid initial connection of Ethernet cable or any USB

I tried feeding just under 4V limited to 150mA max into the buck regulator and still got 3.3 Volts out of U15.  The MCU runs fine and I can boot it using USART1. I haven’t learned how to use STLINK just yet :)

For my board with STLINK in reset the Ethernet PHY current about 45mA and total current was 95mA on wakeup and 115mA in bootstrap.

Why do this, well if I short more than 4V to any port pin I risk destroying the STM32H5E5 and in development shorting pins happens.

 

[gbm]

I don’t understand the nature of your problem. All Nucleo boards are designed to be powered from 5V via USB cable. Most of pins of STM32 MCUs are specified for 5 V input operation in digital mode; this is also true about H5E5.

[rich.g.williams]

​@gbm 

 

This is one of the several threads I have read ,

                           MCU overheat

this is the kind of problem some people have.

 

I only connect USB when absolutely needed during development, all my work is done at 4V and a 150mA current limit.

 

You say  “Most of pins of STM32 MCUs are specified for 5 V input operation in digital mode” 

Not so the STM32H5E5ZJ  absolute maximum voltage for most pins is 4V.

[gbm]

Check the datasheet. Less than 8 pins are 3.3 V only, the others are FT. See Table 16. Of course, normally the chip is powered from 3.3 V, as all the standard STM32s but this is provided on every board, Nucleo or custom design.

[TDK]

Most of the STM32H5E5 pins have Absolute Max tolerance rated at 4V

This is misleading at best. The absolute max on most pins is up to 4 V only when it is unpowered. If the chip is powered from USB and sees 3.3V, the absolute max on most pins is 6 V.

[David Littell]

This seems like a lot of hand-wringing over nothing.  Do what the rest of us do: don’t hook things up hot, be careful when you’re hooking things up (i.e., don’t make a rat’s nest), and check, check, check before powering.  Keep a spare Nucleo (cheap!) for the eventuality of blowing 78VAC into an ADC channel - it happens to everyone sooner or later.

[rich.g.williams]

Perhaps you missed the point ​@David Littell 

If you power your NUCLEO development board from 4V at 150mA max current limit you can short pins out to your heart’s content during 99% of your development work.

Try advising people like ​@Winston Smith who maybe has a blown NUCLEO board 

       MCU overheat and LD4 blinking

[rich.g.williams]

​@TDK 

You will certainly loose this argument.

1. I have seen several posts where visitors have damaged their NUCLEO development boards and I reference the most recent above.
2.  STM32H5E5 datasheet “Absolute maximum ratings” (DS14971 Rev 1, page 162, Table 16: "Voltage characteristics") shows that there are pins that cannot tolerate more than 4V when the device is powered

 

3. From the official schematics for NUCLEO-H5E5ZJ they show beyond doubt that ST are encouraging users to connect between 7V and 12V to pin 24 of header CN7 and nearby pins 23 and 25 and 26 are GPIO pins, an accidental short between VIN on pin 23 and any GPIO will destroy the MCU if VIN is between the official range of 7V < VIN < 12V.  

 

4. A youngster learning about STM32 reading your data and schematic for NUCLEO-H5E5ZJ (and other NUCLEO boards) trusts ST specification for VIN and may not realise that he or she is only one header pin slip away from destroying the PCB and tears. 
5. ​@TDK  if you read my original post I clearly said:- “When I started working with NUCLEO boards I was very cautious about powering the board up from 5V let alone 12V even powering the board via STLINK USB at 5V (maybe as much as 250mA)”
6. ST - you should be ashamed. Sorry to say this. I will write it up fully and post on my website Bill’s Notes  billsnotes.co.uk

[gbm]

Powering Nucleo via VIN is definitely possible but it’s rarely used in practice. In almost every circuit we design there is some voltage on board, which, when shorted with some IC pin, will destroy some circuitry. That’s how prototyping works. You are not obliged to use any voltage above 5 V for powering Nucleo. Nobody forces you to short any GPIO pin to any “forbidden” voltage. Nucleo is a PROTOTYPING board. It’s not supposed to be totally foolproof and nobody expects it to be.

BTW do you realize what these “other” pins are and how many of them there are. I guess maybe 1..3.

BTW2: in practice, most of 3.3 V MCUs are able to survive 6..7 V power supply. Just don’t say it loudly and don’t rely on it - it’s definitely out of spec. Some Chinese STM32-alikes, officially 3.3 V, are not very unhappy even at 8 V,

[rich.g.williams]

​@gbm 

“BTW do you realize what these “other” pins are” it only needs to be one.

“Powering Nucleo via VIN is definitely possible but it’s rarely used in practice” I am using VIN now at 3.95V and 150mA max.  If I followed ST’s official  “Operating range: 7V < VIN < 12V”  and connected a 12V Lithium Ion to VIN I would expect smoke and a burning smell.

Sorry you loose this argument.

[TDK]

 

@TDK You will certainly loose this argument.

Uh huh.

Let’s keep in mind this is what you claimed:

Most of the STM32H5E5 pins have Absolute Max tolerance rated at 4V

Now look at the datasheet. What category do most pins on the STM32H5E5 fall into? It’s not the “other” category. It’s the FT_* category.

 

 

an accidental short between VIN on pin 23 and any GPIO will destroy the MCU if VIN is between the official range of 7V < VIN < 12V

Friend, an accidental short between two random pins on a board will kill 99% of all electronics out there. This is not some ST-specific thing.

[rich.g.williams]

Ok I agree with some of the STM32H5E5 pins can only tolerate 4V.

No ST's or ST subcontractors placement of VIN adjacent to GPIO pin on header CN7 is asking for disaster.

Sorry you loose the argument.