Internal Short in MCU?

[OpenCircuit]

MCU is a STM32F103C8T6-LQFP48. Trying to ascertain if there is an internal short. Removed from circuit.

Supply coming in from voltage regulator via pin 24 (VDD_1). I have absolute continuity from 24 to 8 (VSSA), 23 (VSS_1), 35(VSS_2), 36 (VDD_2), 47 (VSS_3) and 48 (VDD_3).

I do not know what many of these stand for. Tried searching and it seems I should NOT have continuity from any VDD to any VSS. I would like to ascertain form someone who knows for certain as opposed to myself.

Page 26: http://www.st.com/content/ccc/resource/technical/document/datasheet/33/d4/6f/1d/df/0b/4c/6d/CD00161566.pdf/files/CD00161566.pdf/jcr:content/translations/en.CD00161566.pdf

Thanks for any help.

[Rerouter]

when you say continuity, do you mean just with the beep mode on the multimeter, or you measure 0 ohms?

if its just the beep, that comes down to the meter your measuring with,

if its 3 digits or less ohms, then you likely have had the chip latch up destructivly, and short out, or over driven a protection diode between the rails and shorted a diode (generally only happens when abused with a current limited supply to stop it fusing)

[OpenCircuit]

Yes, just the beep. It is a Fluke 114. Very tiny and hard to get resistance readings on this little chip. "Continuity beeper turns on at < 20 ohms and turns off at >250 ohms."

https://www.eevblog.com/forum/testgear/ds0138-diy-kits/

[David Chamberlain]

So you have removed the chip from the board WHY? It is not functioning on the board and you suspect the chip is dead? From my experience MCU's tend to go open circuit and create smoke when used incorrectly - other experiences may differ.

Perhaps you could explain more about your setup here and the board the chip came off. Is it your own board?

More common would be a solder bridge on the PCB between pins creating a short. Or getting the supply voltage around the wrong way.

[rstofer]

From Section 2.3.9 on page 15

VDDA and VSSA must be connected to VDD and VSS, respectively.

The xSSx in any of these specs means SubStrate - ground.
xDDx in any of these specs means Drain supply - whatever the power supply is supposed to be.  Same with VCC...
xEEx means Emitter supply and probably doesn't apply in this case.  Usually a negative supply

Given all the diodes in the device, I wouldn't be surprised if you got a reading from VDD (red probe) to VSS (ground probe).  If you swap the probes, you might not get anything.

[JPortici]

So you have removed the chip from the board WHY? It is not functioning on the board and you suspect the chip is dead? From my experience MCU's tend to go open circuit and create smoke when used incorrectly - other experiences may differ.

Perhaps you could explain more about your setup here and the board the chip came off. Is it your own board?

More common would be a solder bridge on the PCB between pins creating a short. Or getting the supply voltage around the wrong way.

every time i saw a MCU die, it had shorts between VDD and VSS. Usually by overvoltage

this, unless the short wasn't the result of bad soldering. in that case polishing and/or removing the MCU and cleaning the pins solved the problem.

[OpenCircuit]

So you have removed the chip from the board WHY?

https://www.eevblog.com/forum/testgear/ds0138-diy-kits/

Given all the diodes in the device, I wouldn't be surprised if you got a reading from VDD (red probe) to VSS (ground probe).  If you swap the probes, you might not get anything.

Same result.

[OpenCircuit]

every time i saw a MCU die, it had shorts between VDD and VSS. Usually by overvoltage

this, unless the short wasn't the result of bad soldering. in that case polishing and/or removing the MCU and cleaning the pins solved the problem.

I looked closely and did not see bridged pins. Even tried wick and flux in the event a short existed under the chip where I cold not see so well.  MCUs are presoldered before shipment.

Resistance is 23.5ohms -.2 (probes) =23.2 from pin 24 (VDD_1) to 35 (VSS_2).

The first voltage regulator is rated to 12v (78L05) and reduces to 5v. Then to a 3.3 (LM1117) supplying the MCU on pin 24.

78L05 was getting really hot while 5v supply was stable throughout circuitry. Problem was on 3.3v side where a significant drop was found-wish I would have recorded the drop.

Fun to play with.

[OpenCircuit]

Doing some more thinking on this....

If resistance from VSS_2 to VDD_1 is 23.2 ohms- as measured out of circuit- and voltage is 5v, then current is .21 amps or 210mA.

Going through the 78L05 datasheet it appears Ipk or peak output current is 140mA. Does peak output mean the max current throughput?

Current consumption from oscilloscope manufacturer is 120ma with a tiny display, so 210 mA is way out of range.

Have about three more weeks to think about this until the new MCUs arrive.   

[OpenCircuit]

New MCUs arrived. Replaced, no short present. Voltages correct now.  Waiting on flash cable.  Fun stuff.