STM32F446RE can no connect to target

[hcls]

Posted on November 17, 2015 at 14:59

Hello, this is my first post.

It is the first time I use an STM32 Device, I am familiar with 32Bit processors of TI so far.

I did a new PCB connected with the ST-Link programmer of a Nucleo board. But I can not connect to my board. I am using the STM32 St-Link Utility.

My configuration:

TQFP64

Pin 19,32,48,64,13,1 on 3V3

Pin 31, 47, 18, 63, 12 on GND

Pin 30 on 4.7uF to GND (VCAP)

25 MHz crystal on Pin 5 & 6

Pin 60 (Boot0) connected to 10k to GND

Pin 46 to ST-Programmer SWD

Pin 49 to ST-Programmer CLK 

Pin  7 to ST-Programmer NRST

The programmer also has GND of the board.

All other Pins are left floating.

Power consumption 0.3 mA! The CPU seems not to run.

Pin 1 is at the lower left if the text on the chip is readable.

The voltage on the VCap_0 is 0V.

It is a bit confusing, I just have seen that the VCAP_1 (Pin 30) capacitor is not populated and has been shorted to ground on the Nucleo board.

Could anyone give me an advice how to handle this situation?

Thanks in advance

Chris

[stm322399]

Posted on November 17, 2015 at 16:07

you'll get more chance to swd into the MCU while BOOT0 is tied to VCC (as per 2606). If your MCU has BOOT1, tie it to the ground.

BTW it is also likely your custom HW being screwed with solder bridges, mirrored footprint or any other kind of misfab.

[hcls]

Posted on November 18, 2015 at 13:36

Hi Laurant,

Thanks for your reply.

I checked the PCB, it is nothing else connected than the uC and some decoupling capacitors along the Vcc and Vss pins.

I tied the Boot0 to Vcc now, Boot1 is not existing on TQFP64.

Could you please tell me what I should do with the VCap1 (VCap 2 is not populated on the TQFP64 as far as I know). 

The Footprint seems fine, I also checked all voltage pins. How much current should the uC use out of the factory if boot0 is high and the system voltage is 3.3V?

Thanks for your help again.

Chris

[stm322399]

Posted on November 18, 2015 at 14:33

Chris,

VCAP1 must be connected to the ground through a low-esr (<1ohm) ceramic capacitance (4.7µF). Resulting voltage sb. around 1.2V

Everything else looks good to me (i am not an HW guy).

I guess that the MCU will draw somewhere between 1 and 10mA.

[Tesla DeLorean]

Posted on November 18, 2015 at 14:53

Well it does seem to be indicative of power not reaching the part. I'd probably look very carefully at VDDA.

Measure the voltage on NRST.

Do you have a schematic or gerber for this board? Have you checked the netlist, especially the power connectivity, and have you checked continuity on an unpopulated, or partially populated board.

I would go with the DM suggestions related to VCAP1, but as indicated you should see 1.25V or so there. That you do not suggests the regulator is not on.  Most often seen where the orientation is wrong, but you indicate you've looked at that.

[hcls]

Posted on November 18, 2015 at 16:07

Juhhu :) You are great!

It is working now, as mentioned the VDDA was not populated yet and open.

Next time I will measure everything twice.

Thanks again for helping me!

Cheers,

Chris

[stm322399]

Posted on November 18, 2015 at 16:38

Hum ... shall we understand that pin 13 was not really connected to 3.3V ?

[Tesla DeLorean]

Posted on November 18, 2015 at 17:29

Hum ... shall we understand that pin 13 was not really connected to 3.3V ?

That would seem to be the inference. When reality defies the stated facts, I tend to question the facts, not the reality.

[hcls]

Posted on November 19, 2015 at 07:34

Yes.

For the VDDA I used a LC Filter. This filter I have not populated then thats because the it was floating.

I have no idear what I have been measured there :)

Cheers and a good rest of the week,

Chris