Power up the stm32f4 discovery with bare minimum components

[michaelmccartyeng]

Posted on November 11, 2013 at 20:46

I have fried 2 of my 3 stm32F4 discovery boards. Before I even touch the third one I want to understand more about the power subsystem in these devices.

I've been reading the datasheets (for the chip, the ref manual and the discovery). They stay

The device requires a 1.8-to-3.6 V operating voltage supply (V

DD

). An embedded linear

voltage regulator is used to supply the internal 1.2 V digital power.

So that makes me think I can simply wire up 3.3v to the Vdd line and the chip will start working ? Is that a correct assumption ? I'm pretty sure what I did to my existing boards was destroy U1 (3.0 ldo regulator). I've ordered some replacement parts for U1. Theoretically if I only burnt up that component I could connect 3.0v to the 3V pin on the board and it should start right ? I'm not sure but I think what happened was I connected 8v to the 5v line and burnt up U1. U1 does not output 3.0v as it should but more like .6v. And I did see smoke. So If I want to salvage the f4 on my board and simply power it up then program it via another stlink anyone have any ideas ? Done this before ? Thanks ! #stm32f4 #ldo #smoke

[Tesla DeLorean]

Posted on November 11, 2013 at 23:34

You should be able to pull off the regulator and push 3V in there. That should then route to the analogue and digital supply inputs. The analogue power has the power-on-reset circuit internally. The internal regulator needs the bulk capacitors on the VCAP pins.

[michaelmccartyeng]

Posted on November 12, 2013 at 00:16

Thanks Clive.

I used the remaining stm32F4 board and disconnected jp1, then I took off both jumpers on the ''stlink-discovery'' header. Then I connected the 3v and the Gnd from the good board to the bad (P2 header on either). Then I connected board 1 (working) board two cn2:2 -- Pa14 cn2:4 -- Pa13 cn2:5 -- cn2:5 cn2:6 -- cn2:6 Then trying to program the chip on board two I got ''board is held in reset'' Nrst was 1.15v (on board2) Led 8 and 7 Were off on board2

[michaelmccartyeng]

Posted on November 12, 2013 at 00:27

I also tried it on my 3rd board. That board actually blinks ld2 and is somehow resetting my external regulator making the whole circuit blink. Its a programmatic 1s blink. I'm pretty sure on board 3 I connected 8v to the 3v rail.

Well I guess they are shot. I really wanted to tell how to actually do SWD so I can design the correct header into any circuits I create that use the stm32f4 chip. Figures I had the genius idea of soldering the parts to the board instead of creating another board that the stm32F4 could plug into. It seemed quicker, now I have to re-solder everything. I'm going to create another board this time, modularity is awesome.  

[Tesla DeLorean]

Posted on November 12, 2013 at 02:04

You might want to break solder bridges that connect SWO and NRST to the F103 (ST-LINK) part. The DISCO boards aren't really designed to inject SWD connectivity in via CN2. I'd pull the jumpers from CN3 and inject SWDIO and SWCLK on the target side. Then you have SWO (PB3) good for debug console, and NRST to get control of the core.

For my boards I've been using the 10-pin Cortex Debug connector.

http://infocenter.arm.com/help/topic/com.arm.doc.faqs/attached/13634/cortex_debug_connectors.pdf

I think we used a Samtec SMT shrouded header. I'd have to look up the part#, but it was pretty small, about equivalent in area to the 6-pin non-standard thing on the DISCO board.