Concerning ST-Link on Nucleo Board

[Zt Liu]

Posted on June 08, 2017 at 17:33

hi,  fellows!

I have some questions about ST-Link, especially those on Nucleo boards.

First, I often use the on board st-link link to debug/program some other custom boards, it works fine.

But someday, I'd noticed that the debugger's AIN_1(target VCC) is never  really connected to the MCU VCC

of the application board, for R9 is not lying there. See below:

From the UM1075 User manual of ST-Link, it says:

The power supply from the application board is connected to the ST-LINK/V2 debugging and programming board to ensure signal compatibility between both boards.

So is this really unnecessary? (For me, the debugger/programmer works fine so far.)

Or is this risky for some reason?

I really need to know about this!

For we are making a project with the whole nucleo board as a part,

 with st-link on it.

And I've already removed R23 for saving power.

See below:

Thanks for reading the post!

Zt

#debugger #st-link #programmer #nucleo

[Tesla DeLorean]

Posted on June 08, 2017 at 20:00

The stand-alone ST-LINK has buffers so that the debugger and the target board can run at different voltages. The NUCLEO/DISCO boards don't have these buffers, and just use/assume the use of 3.0V or 3.3V signalling.

Most of the pins are 5V tolerant, and the VIL/VIH vs VOL/VOH provide a lot of leeway.

The idea would be to avoid back-feeding current into a processor via the GPIO pins and skewing the current measurements.

If the target STM32 were running at 1.25V I'd probably want the buffers

[Zt Liu]

Posted on June 09, 2017 at 01:14

Thanks Clive, as always!

So since my target MCU is running at 3.0/3.3V, there is a lot of leeway.

Guess I can sleep well with the target VCC pin left open! 

🙂

[Uwe Bonnes]

Posted on June 09, 2017 at 11:15

For the STLink on Nucleo and Disco boards, nearly _none_ of the pins used is 5-Volt tolerant! So if you only power the target and not the stlink, the stlink will mostly be back-powered by some pin and target current will increase substantial!

[Zt Liu]

Posted on June 09, 2017 at 11:36

the stlink will mostly be back-powered by some pin and target current will increase substantial!

It's true, and I'd discovered that :

Even the  target MCU is in stop mode, the WHOLE Nucleo board is sinking almost 2mA!

This almost drives me mad, so I disconnect to Vcc Target, Rest pin, and  took off LD3 too. 

Now the whole board, in stop mode, only takes around 20uA with rtc on. 

[Tesla DeLorean]

Posted on June 09, 2017 at 20:54

>>

nearly _none_ of the pins used is 5-Volt tolerant!

I see where you're going with that, the F1 host side is problematic. The target side (SWDIO, SWCLK, SWO) all the parts I looked at are FT pins