Trouble communicating with a new board design - SWD ERROR

[ravix475]

Posted on July 24, 2015 at 01:19

This is my first project with an STM part. I have started development on a NUCLEO board with an STM32F401RE chip and I'm programming/debugging using the onboard STLINKv2. So far firmware development has gone well. 

I designed a new board with the same chip (STM32F401RE) and I bought a second NUCLEO board to harvest the STLINK from. 

I cannot communicate with the new pcb at all -- from either STLINK, and I'm afraid I might have done something wrong in my schematic. 

I am attaching the schematic and a picture of my programming setup. 

Some notes: 

1. I have removed the CN4 jumpers on the STLINK board

2. I have tried with only SWDIO, SCK, GND, and VDD  (also trying with SWO and NRST)

3. I have tried various configurations of BOOT0, BOOT1, although by default BOOT0 = GND. 

4. I have tested continuity all the way between the pins and the STLINK and made the wires shorter. 

No matter what I do, when I plug in the STLINK board, the NUCLEO drive contains a file called fail.txt, which contains the text ''SWD ERROR''. And I am unable to connect or communicate with the target board from the STLINK software. 

[Tesla DeLorean]

Posted on July 24, 2015 at 02:19

Can you attach a better focused/resolution picture, rather than in-line (100kb) one?

Check the voltage on the VCAP pin, and the NRST pin.

Can you use the ST-LINK Utilities rather than the on-board mass-storage method?

The board looks to have it's own supply, if so *don't* connect VDD to the ST-LINK

[raptorhal2]

Posted on July 24, 2015 at 15:54

The functional connections to the SWD are incorrect. E.g., GND is pin 3. Check Section 5.2.4 in the Nucleo reference manual.

Cheers, Hal

[ravix475]

Posted on July 24, 2015 at 19:45

Thank you for putting me on the right track. I checked VCAP1 and saw that the 1.2V rail was failing to come up properly. I tried replacing the capacitor on VCAP1 with the one from my spare NUCLEO board without any change. After reviewing the reference manual I found that VDD - VREF+ cannot be greater than 1.2V. I had VREF+ unconnected. 

After connecting VREF+ to VDD the chip powered up successfully. 

V

DDA

−

V

REF+

< 1.2 V 

[ravix475]

Posted on July 24, 2015 at 19:46

Hi Hal, 

Thanks for the review. You are, of course, correct. But I am already re-arranging the wires in between the boards to correct this. 

[Tesla DeLorean]

Posted on July 24, 2015 at 19:51

After connecting VREF+ to VDD the chip powered up successfully. 

I was looking for VDDA on your design, assuming that's the VREF+. The Power-On-Reset circuit and PLL are run off the analogue supply. The parts won't start without it.

Pin 13 on the 64-pin devices is VDDA, +VREF is connected to it internally.