Stupid question about blowing up Teensy3.2

[queenidog]

Since most of the questions here are quite intelligent, let me ask a stupid one.

I've been working with electronics for ...50 years at least, yet still managed to blow up a Teensy by applying 12 volts to a digital input, D15, during a rewire.

Question: Would current be able to propagate OUT of the chip (Teensy 3.2) simultaneous with the 12v application, damaging other outside connections? I'm thinking the small pcb tracks inside the chip should blow so fast that there wouldn't be time enough to damage anything outside of the chip. I ask because after I replaced the chip the display it is attached to is not diplaying properly, using a known good program. (I don't know about the quality of the replacement Teensy: it has been used before).


For your enjoyment I'm submitting a photo of the damaged part.

[defragster]

Found this on Arduino boards: rugged-circuits.com/10-ways-to-destroy-an-arduino/

perhaps this or other things apply - given 'some signs' of catastrophe in the package seems all safety margins and components violated:

Method #3: Apply Overvoltage to I/O Pins
...
If the internal protection diode fails open, then the overvoltage destroys the I/O pin. If the protection diode fails by shorting out, it’s even worse because now the overvoltage is applied to the entire +5V supply on the Arduino. This means it will reach other components, like the USB interface chip, and destroy them too.

[queenidog]

Found this on Arduino boards: rugged-circuits.com/10-ways-to-destroy-an-arduino/

perhaps this or other things apply - given 'some signs' of catastrophe in the package seems all safety margins and components violated:

Arrrggghh! Thanks defragster. I was hoping it would be the opposite. I'll start checking downstream of the logic chips. Better pull out the schematic.

Two photos: pcb layout and board with Teensy at upper left.

[liudr]

Write a test code to test your electronics. We don't know how the 12V may have damaged everything else. The very least, find a way for 12V to not be accidentally wired to a Teensy pin. This could involve optimizing your design, wiring and/or build process.

[queenidog]

Write a test code to test your electronics. We don't know how the 12V may have damaged everything else. The very least, find a way for 12V to not be accidentally wired to a Teensy pin. This could involve optimizing your design, wiring and/or build process.

I only put the 12 volts on D15, nothing else. Had I put it on Vin...I would have blown everything on the board...all TTL 5 volt devices. The only way I think of ensuring my spare Teensy is working is by using it in another circuit. Right now I don't know if it's the spare Teensy or the board that it is on is at fault. I have others on order, but they could take some time to arrive here. Thanks for your insight.

[PaulStoffregen]

It's very difficult to predict. 5V tolerant chips like MK20 on Teensy 3.2 have more complex ESD protection than just simple diodes. NXP and other companies almost never share the design of those circuits. But as a general rule of thumb, overvoltage usually causes shorts within semiconductors, by 2 types of failure mechanisms. The simple one is breakdown of thin oxide insulation between layers. The other is reverse biased P-N junctions. All over the chip, mosfet source & drain terminals are N or P material implanted in a base of other opposite material. All those locations where opposite materials meet form "parasitic" diodes which are normally all reverse bias. Either type of breakdown means failure which conducts the too-high voltage to other stuff inside the chip. But the ESD protection circuit may have actually "worked" and shunted the massive current to GND. Very difficult to anticipate, since we don't even know what circuitry is really inside the chip.

[queenidog]

Well the data are in. I tested my Teensy in another circuit with different program...and it works fine.

So...when I blew up Teensy #1, it took out some chips downstream. Thankfully I put all my chips in sockets so replacement should not be difficult. I'll start close to the Teensy, then move outwards based on connections.

Thanks all for your insight.

[ratnin]

From my experience you CAN damage other on-board components when applying excess voltage to a Teensy GPIO when the voltage is large enough (~9-10V+), the circuit path to the pin does not have an inline resistor to limit current and the other PCB components are supplied 3.3V thru the Teensy on-board regulator.

When the voltage exceeds the protection circuitry on the pin it appears to feed back to the 3.3V regulator output which them damages other components feeding on this same 3.3v rail. You'll typically find that the Teensy board becomes hot above the Vreg when powered up after this occurs.