Nucleo-432KC Repair

[waymond91]

Hi All!
I have been using a lot of MBed boards for development lately because they are so cheap and have so many of the embedded setting pre-configured.
However, I have been having a problem with some of the smaller Nucleo boards (the ones designed to fit the Arduino Nano footprint), primarily the Nucleo-L432KC.
It seems that I have a power problem, and I don't really understand the design.

I powered the board through the "VIN" pin via 2-cell LiFe battery. The board uses a LD1117S50 (U6 in the schematic) as a regulator and it should be OK for voltages up to 25V. It is possible that I plugged in the power and ground backwards?

However after powering the board, the high-side power switch (U1 - ST890CDR) ran very hot. Now when I plugin via USB, the device shows up on my computer and appears to work, however the 3.3V bus seems to be around 1.9V and the ST890 is very hot.

I have setup LDO regulators before but have never used a high-side switch before. According to the datasheet this is there to protect everything in case of a short circuit, as best I can tell.

When I measured the resistance between the 3.3V bus and Ground its about 390 Ohms, so clearly there was a short.

I figured that maybe I had blown the "VIN" regulator. So I removed it from the board, but it exhibited the same behavior as before (1.9V on the 3.3V bus). Likewise, I did the same thing with the 3.3V regulator and had the same result.

Does anybody have any insight as to how this circuit is supposed to work and what component may have failed?

I have included png's of the schematic. All the design files can be downloaded here:

http://www.st.com/en/evaluation-tools/nucleo-l432kc.html

[singapol]

You can measure resistance of pins like Vcc if it is low or desolder U1  and know the answer.

[HackedFridgeMagnet]

...
I powered the board through the "VIN" pin via 2-cell LiFe battery. The board uses a LD1117S50 (U6 in the schematic) as a regulator and it should be OK for voltages up to 25V. It is possible that I plugged in the power and ground backwards?
...

When I measured the resistance between the 3.3V bus and Ground its about 390 Ohms, so clearly there was a short.

I figured that maybe I had blown the "VIN" regulator. So I removed it from the board, but it exhibited the same behavior as before (1.9V on the 3.3V bus). Likewise, I did the same thing with the 3.3V regulator and had the same result.

Does anybody have any insight as to how this circuit is supposed to work and what component may have failed?
...

Bad idea reverse polarity while using a battery.  I think you may have blown your Nucleo processor.

The 390 ohms could just be the feedback network of one of the regulators.

I would power the Nucleo with a current limited power supply at maybe 7 volts, disconnect any jumpers and measure the outputs of the regulators. (5 and 3.3v)
If it is drawing more than 0.8 A or less than the specified voltages then something is short.
See what is hot, see what you can disconnect.

also, separately, power another good nucleo board with your STLink and check the STLink is ok.


good luck

[waymond91]

Ok, I did remove U1 and that seemed to fix the short as it were, and I can power it directly from my bench supply directly no problems.
I have designed and assembled a few custom AVR boards before, and I never included any mosfets or any form of switching in the 5V or 3.3V rail before.
I assume that it is safe to design the supply with a simple LDO regulator and few decoupling capacitors for something low power like an ARM MCU, and that these extra features are "guilding the lily," and are supposed to help protect the components on the board against dumb end users, such as myself?  

[HackedFridgeMagnet]

Good work.
I guess U1 is there to give the st-link utility the capability to force a hard reset of the main processor. The utility can then call this if the software you are writing is locked up.

[Gyro]

A bit picky, but you might find it useful for future usage... The battery is Lithium Ion, not Lithium Iron (iirc, the last battery that used Iron in its chemistry was NiFe which preceded NiCd).

[HackedFridgeMagnet]

LiFePO4 probably.

could be referred to as Lithium Iron Ion.

[waymond91]

I am in fact using Lithium Iron.
They are a lot more safer and I believe they are rated for more discharge cycles (but I would have to check).

I am using another 303K8 nucleo board (a different board than earleir) and I powered it up with my external supply both from Vin and 3.3V (with appropriate voltages) with current limitation. Nothing bizarre happened at all that I was able to perceive.

Now when I got to power it up via USB no LEDs come on and nothing shows up on my computer. Nothing on the board gets hot. Although I cannot reprogram it. What am I doing wrong here??!!
 

[HackedFridgeMagnet]

You probably know this but they have a different cell voltage than other Li Ion batteries, so make sure your charger supports the correct voltage.

I am in fact using Lithium Iron.
They are a lot more safer and I believe they are rated for more discharge cycles (but I would have to check).

I am using another 303K8 nucleo board (a different board than earleir) and I powered it up with my external supply both from Vin and 3.3V (with appropriate voltages) with current limitation. Nothing bizarre happened at all that I was able to perceive.

Now when I got to power it up via USB no LEDs come on and nothing shows up on my computer. Nothing on the board gets hot. Although I cannot reprogram it. What am I doing wrong here??!!
 

Trace the power rails. ie. Make sure you have the appropriate voltages on the path through to the power pins on the chip. USB lines too if they are ok, and the _Reset pin ( I assume it has something like that), I can't be more specific.