STM32F429x don't start - Vcap problems...

[kecskesgery]

Posted on October 01, 2015 at 13:33

Hello there!

I've desingned a board for a project with a new 

STM32F429VGT6 controller. My main problem is that I cannot program t

he controller, ST-Link says it ''can not connect to target''. Of course, I've already checked, the controller gets 3.3V, and the SWO pinout is also good.

I went through the datasheet again, and I I've found out, that on the Vcap1,2 pins I should measure around 1.2V. But I'm only getting some mV. I'm using 2.2uF capicators, with low ESR value. I have no idea what causes the problem, and at this point I'm pretty much stucked... :(I've attached my board pinout, maybe it helps. Thanks for the help in advanced! #stm32f4 #stm32f4 #swoproblems #swoproblems #vcap #vcap

[Tesla DeLorean]

Posted on October 01, 2015 at 15:03

Check the part orientation, the designator is a small dimple, not the molding eject pin marks.

[Tesla DeLorean]

Posted on October 01, 2015 at 15:07

Not sure I've got a 100-pin to hand

https://community.st.com/0D50X00009Xka68SAB

 I don't think you have enough pins for the Regulator Bypass to be an issue.

Edit: Fixed DEAD LINK, original post from 1-Oct-2015

[kecskesgery]

Posted on October 01, 2015 at 15:20

Hey,

i've checked the orientation, that was the first thing, and it's OK.Any other clue?

[Tesla DeLorean]

Posted on October 01, 2015 at 18:31

Well you're kind of back to the supplies, you'd want to check all the nets, make sure you don't have any islands, and confirm ground connectivity.

In most case I'd expect the power ring in the die to connect all common external pins, there's usually multiple pins to localize the high current flows. What kind of current draw are you seeing? Pull BOOT0 High, not sure it helps on a blank device, but the System Loader isn't going to put the device into a low power mode that would switch off the internal regulator. Without 1.25V on VCAPs the core of the part isn't going to run

[kecskesgery]

Posted on October 01, 2015 at 23:03

Thanks for the fast response clive!

Well, right know there are only the microkontroller and the power supply circuit on the board. I'm giving 5V to the board, and I get 3.3V via a buck converter. The current drawn is less then 10mA on the board right know, since I'm limiting it when I test the board.Also, here's the attached schematic for the Boot mode selection (boot0 in particular):Selectiong Boot0 hight/low does not change anything...:\ Still not getting 1.2V on VCaps.Any other idea?

[carl2399]

Posted on October 02, 2015 at 00:28

Limiting the current to 10mA? That's probably a bit low. Try increasing the current limit to about 50mA (or even 100mA). I generally see around ~20mA with my custom bootloader running, and around ~30mA during normal operations. That includes a couple of other components and sensors as well, however I'm only clocking (post PLL) at 30MHz.

[Mark Edwards]

Posted on October 03, 2015 at 00:12

What's the state of NRST and what is it connected to?

As this has caught me out in the past.

[kecskesgery]

Posted on October 03, 2015 at 00:22

Hey,

so, nRST is connected to a button, and this is the scematic:This one should work, and it does work, as I have tested it.  Still don!t know what else to check.... :(

[Tesla DeLorean]

Posted on October 03, 2015 at 00:30

Unfortunately I don't have a magic wand.

Have you built up a couple of boards to rule out board/chip issues? Are there components placed as resistors, capacitors or inductors that aren't actually the type and value of part specified in the design?The thing with NRST is to check if it's being clamped low by the POR circuit in the part, ie analogue supplies missing. Not aware it should stop the regulator turning on. The STM32 has other pins, some times not bonded out, which control the regulator or reset behaviour.You could stare at the Gerbers and check for lack of connections there, or shifts in pins in the symbol vs part.