Getting ST-LINK-V3 MODS to run after previous used of nucleoboards.

[Tobe]

Hi,

i was hoping, that the Link-V3 would just work, after i used nucleo boards already. But i was dissapointed. I use the GDB server, and it says:

ST-Link enumeration failed

Error in initializing ST-LINK device.
Reason: ST-LINK DLL error.

I just checked a Nucleo board, and there is no problem.

The installation of the driver worked without errors.

[TDK]

Can you connect with STM32CubeProgrammer? If not, at which stage does it fail? Does the ST-link show its serial number? Does the target voltage show up and is it valid?

Since this is custom hardware, was it ever working correctly or is this a first try? Does it have power, are pins connected correctly, is VCAP correct, etc?

[Tobe]

It now works. I think i may have to turn / plugin the cables in a certain order.

But i now have this Error:

Error message from debugger back end:
Error finishing flash operation
Error finishing flash operation

[TDK]

I asked 6 questions, you answered 0 of them. If you want help by me, you're going to have to pitch in a bit more.

[Tobe]

Since the original error message is gone, i did not consider answering the question. But i do so now:

- I dont know i which way i should connect with the programmer.

- The GDB server now works, but running the cube programmer, it reads the ST-LINKV3 serial number and firmewareversion. It shows a voltage of 0.01V, but i confirmed, that the target has its 3.3V.

The current error massage is: "Error: No STM32 target found! If your product embeds Debug Authentication ..."

It is the first connect of this prototype. I was already wondering, why NRST was not on this ST-LINK.

I found out, that it is recognized, when the target is powered off.

What is VCAP?

[TDK]

Okay, it sounds like the computer can see the ST-Link okay and that the problem is between the st-link and the target mcu. Is that wired up correctly?

Since STM32CubeProgrammer reports that the target voltage is 0.01 V, perhaps the target isn't powered. I know you said that it is, but then we need to answer why STM32CubeProgrammer disagrees. Can you provide the schematic? Can you probe the T_VCC pin directly to see if it's reported?

T_NRST is pin 31 on that device and should be connected to NRST on the target chip.

[Tobe]

I have verified, that the labels are connected to the right pin. They also have a 5k resistor before the target.

I could verify, that there is a clock on the clock line, and some data on the DIO line.

One very interesting thing: there is a reset pulse right in the middle of a datapacket! That not so optimal.

When making the programmeradapterboard, i got a bit confused, because there is no NRST pin there. Also i could not find informations about how to connect it.

[Tesla DeLorean]

The physically connectivity to the MCU and the involvement of the resistors would be a lot more helpful here.

The ST-LINK should be able to sample and report the target voltage correctly.

The connectivity to the MCU should be pretty direct and straight forward

NRST / NRESET

SWDIO

SWCLK

VCC to T_VCC to drive the target pins at the correct voltage

GND

[Tobe]

What does the "T_" stand for?

[Tobe]

I have now made more measurements:

The signal have the right levels without a T_VCC connection.