Dead Nano - Did it deserve the Wiss Shears?

:slight_smile:

It’s beating a dead horse probably… but I wish to academically explore the symptoms.

The video of me actually snapping it is a google photos and is not cut, but if you are bored…
Snapping

  1. I felt the Nano was at one time shorted by means including, but not limited to: output pins, ground + vcc, aref… the period of faults may of been trivial.

  2. Pins 0, 1, 2, & 3 began all to report low regardless the pinmode or digitialWrite()s;.

  3. On a rather large test sketch (i can post it if you wish), it cascaded pins that were pulled up but otherwise open to input. Say one went low, then several would follow.

  4. Lastly, on same sketch, I began to get spurious readings from pins 6, 9, a1 and a5 I believe. I could not find any cause. They were pulled up, open and would go low at nearly random intervals. The complexity of intervals suggested a cause outside the sketch. Oddly, another “lighter” sketch only exhibits the dead pins and cascading symptoms.

  5. I can rule out, reasonably, excessive RF interference, PC & power supply malfunction, lead wire faults or qualities borne out of capacitance, induction, or impedance. Likewise ESD, it isn’t that sensitive and I treated it like a memory module.

  6. I conclude that several failures occurred due to very poor handling of device. One of which had to be in memory register, precipitating the errors on longer sketches. Other damage may of been to pull-up resistors or the switches thereof.

Any other thoughts?

If behaviour gets worse over time, it's probably an overvoltage (e.g. ESD). I've seen it in my first job where a micro-controller had one blown pin.

The production of this micro-controller was near perfect; usually when testing we got maybe 1 in 10000 failures.

Suddenly we however picked up higher failure rates related to input leakage on one pin. Running tests comparing leakage results with some golden samples showed that the input leakage of the batches was higher than the reference samples.

And running the tests a few days later again, they were again higher, etc. Till it became so high that it was out of spec and the device was / devices were rejected.

After some finger pointing (supplier blaming us, we blaming the supplier), it was eventually determined that the suppliers packaging line had a grounding problem, hence static electricity did build up which in turn damaged the pin.

And I have seen the same behaviour when my wife's computer and embroidery machine (connected via RS232) were hit by lighting via the phone line.

After fixing the computer, it could occasionally not communicate with the embroidery machine. This 'occasionally' became more regular and eventually there was no communication at all possible. Tried other computers, still no communication.

ESD is a slow killer.

I've treated it well in regards to ESD and poor in terms of hookups, soldering, general electronics. I had a aluminum foil-clad paperboard holder that was connected to the ground. I'm poor. However, it is winter and absolute humidity is low. I have to take item sometimes from bathroom (bright as broadway), bedroom (computer), and basement (workbench). Wires protruded at times.

I like the idea of ESD, as the bugs were hard to pick up. The fact it didn't take longer code for example, could of been a half-nuked cluster of junctions somewhere in the register that just was pushed over the limit once utilized.

I've been handed a Ipod only to hook it up to PC and blow out the southbridge.

I feel like taking next board and hooking up 220Ohm in series and 1M in parallel to pins.