How to kill a microcontroller with one finger

[newxito]

I just wanted to share something that happened to me yesterday. I killed the ESP32 on my nixie board just by touching the adjustable voltage regulator IC with one finger. I’m pretty sure that I touched the feedback pin causing the voltage to increase. 

I already changed the design, and, in the future, I will only use regulators with fixed output voltage. I hope they are less sensitive.

[gregebert]

Or it could have been ESD; winter time is notorious for low humidity (dry air), so it's much easier to accumulate a dangerous charge. All of my larger boards have a banana jack that has a 1 meg resistor to GND, and I either plug my wrist strap into the jack when I'm bagging/unbagging the board (in a conductive ESD-safe bag), or I connect it to my workbench ground. Since I also wear a wrist strap, my body and the board are always at the same potential. 

I also have large resistors between each power supply and GND, for ESD protection. Typically 10Meg. 

I've never zapped an IC from ESD, or any other abuse.

[David Forbes]

The voltage feedback resistor is a low impedance node on every voltage regulator design that I have worked with. Typically 1k ohm. I don't see how that could happen. 

On Wed, Dec 9, 2020, 12:30 PM newxito <axt...@gmail.com> wrote:

I just wanted to share something that happened to me yesterday. I killed the ESP32 on my nixie board just by touching the adjustable voltage regulator IC with one finger. I’m pretty sure that I touched the feedback pin causing the voltage to increase. 

I already changed the design, and, in the future, I will only use regulators with fixed output voltage. I hope they are less sensitive.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/363b5739-77ec-4557-8728-28899df4430an%40googlegroups.com.

[petehand]

Ah, this is not uncommon. A lot of AVR Dragons were slain for the same reason. Touching the regulator affects its feedback and therefore its output voltage. A tiny ESD, not enough to cause any damage, can drive the output high for enough milliseconds to cook a 3.3V IC downstream.

[newxito]

I was using the datasheet schematic of the MP1584 regulator. The FB divider for 3.3V is 124k to Vout and 40.2k to ground. The regulator works great but now is no longer recommended for new designs anyway. 

I switched to AP63203WU-7 for 3.3V and AP63205WU-7 for 5V. The design is now cheaper and uses less components. 2A should be enough for RGB-LEDS and peripherals.

[Dan Hollis]

Are you just using the reference design for AP63205WU-7 ?

-Dan

> --
> You received this message because you are subscribed to the Google Groups "neonixie-l" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

> To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/b005ada3-a0fb-4230-9c3b-a8c3fb70d5d1n%40googlegroups.com.
>

[newxito]

yes, the schematic in the datasheet. I hope it will work as expected, I already ordered some new boards.

[Frank Bemelman]

Or just bad luck. I don't think there is a generic design problem. Usually and in particular when I have been moving around - I touch some ground first to discharge myself just to be on the safe side...

On Wed, Dec 9, 2020 at 8:30 PM newxito <axt...@gmail.com> wrote:

I just wanted to share something that happened to me yesterday. I killed the ESP32 on my nixie board just by touching the adjustable voltage regulator IC with one finger. I’m pretty sure that I touched the feedback pin causing the voltage to increase. 

I already changed the design, and, in the future, I will only use regulators with fixed output voltage. I hope they are less sensitive.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/363b5739-77ec-4557-8728-28899df4430an%40googlegroups.com.

--

Vriendelijke groeten,

Frank Bemelman

Frank Techniek

Vennestraat 11X

2161LE  Lisse

Telefoon +31 252 769 178

[newxito]

I might have to clarify: I'm not saying that the IC has a problem and that I don't recommend it for projects. As I said, it works very well. 

The manufacturer itself has set the status to "not recommended for new designs", they have developed a new version.