Espruino Wifi very hot / broken

Hello,

I have been doing tests to detect 12V voltages with my Espruino Wifi, using a voltage divider with resistors. Since today, as soon as I connect it, it gets very hot. Probably I've done something wrong and broke something inside. It works, but after a few seconds or minutes it resets.

I feed it with a 5V USB adapter, and I have tried another adapters. I have measured the consumption and it's around 350 mA, more than double the normal.

Could it be because I broke some internal component? Is there anything I can do to check or solve it? (apart from buying another, what I will do)

What can I do to avoid it next time? Could it have been through a pin, or through the power supply/GND?

Many thanks!

Are you injecting the 5V on the VUSB or 3.3 pin? Moreover, did you read the warning message on the Espruino Wifi page?
Note: Unlike Espruino Pico and the original Espruino board, Espruno WiFi doesn't contain any battery switchover circuitry. The +/+VUSB pin is connected straight to USB 5V, and shouldn't be used to power the WiFi board while Micro USB is plugged in, unless it is via a diode from 5V.

Could you post your schematic?

Espruino WiFi can draw in the 350mA range for very small periods of time when connecting to WiFi, however if you're running the board from a standard 5v USB voltage source when nothing is connected and no code is running and you're still getting a 350mA power draw then something is very wrong.

Is the WiFi still connected into your circuit? The main culprits would be:

• you're somehow putting more than 3.3v onto the 3.3v line (it could be shorted to 5v?)
  
• you're drawing a lot of power from the 3.3v line so the on-board regulator can't maintain 3.3v (350mA is about the max it's rated for, so that could explain it)
  
• you're putting more than 5v on a GPIO input (all the Wifi's inputs should be 5v tolerant)
  
  

Which part of the board is getting hot? The little 3-legged regulator, or the STM32 chip itself? I have had one chip (on an ST discovery board) that did as you explain - getting very hot - and that was I believe caused by noise on the power supply/GPIO lines. I was using it in the engine of a car, and the electrical noise down the lines was far too great.

Hi Gordon,

Unfortunatelly my Spruino Wifi just died today. I connected it to the 5V USB port of my computer and flashed the firmware to have a clean system and do some tests. It was very hot. After 10 minutes it switched off and never came back.

Voltage of the USB port is normal and I also tryed other 5V wall adaptors.

I probably did something wrong while playing with the voltage divider, maybe I applied 12V directly to a pin, or made a short, don't know. I must say that it has survived many other mistakes I made (as I am learning electronics) so the board is pretty strong and this is absolutelly not a complain.

Just in case it is interesting to you, this is the info I have:

Before dying, I measured 88ºC at the regulator and 50ºC at the chip. The system booted and worked right, maybe a bit slow, but the wifi and my http server worked. Digital pin outputs also worked.

While feeding with 5V, the 350 mA power draw was continuous with Wifi on. I remember that I measured 150 mA before the day it started to get hot, so I think that the heat came from the extra 200 mA.
While feeding it with 3.3V, power draw was 280 mA with Wifi, less temperature but still high.

When it died nothing else was connected to pins and system was clean, no programs, but heat and power consumption was the same.

Thanks for your support!

Fri 2018.09.14

@César, although your voltage divider technique is acceptable, have you considered using opto-isolation to electrically separate the two circuits? May be a bit overkill, but a one dollar part would definitely protect your next treasured Espruino device.

@allObjects wrote a nice overview that may be read under heading 'Optocoupler' that I referenced

Quick Start Best Practices - Connecting a relay using isolation - optocoupler - 2 Relay Board Module

re: 'What can I do to avoid it next time?'

Are you able to post a picture of, and/or provide a schematic as @Jean-Philippe_Rey suggested?

Argh - sorry to hear that. It's interesting about the power usage you mention - my bet would be that at some point 12v got applied directly to some of the chip's IO pins, and that might have physically damaged the chip at that point. While the rest of the chip would limp on, that part of the GPIO would be shorted and would be drawing loads of power.

The voltage regulator itself is thermally protected, so the good news is if you have access to a hot air rework station, probably replacing the STM32F411 chip (and reflashing it with Espruino) would bring everything back to life.

If you're new to electronics that might be a bit difficult - but you may know someone that would be willing to give it a try?

As @Robin says, if you're dealing with digital IO then OptoIsolators are a really neat, safe way of connecting systems together without risking any damage. There's not really any less soldering than if you used a voltage divider.

If you do need to use a voltage divider (eg for analog) then there are some things you can do to be nice and safe:

• Add a 1k resistor between your input and the signal pin of Espruino - that way even if you go connect 12v, there's a limit to the amount of current that can be put into Espruino - which will most likely save it. It works for output pins too.
  
• Put a capactor on the Espruino input pin - it'll reduce the frequency of the signal you can receive, but with the resistor (above) it really gets rid of any short pulses voltage. It also makes your analog signal readings more accurate :)
  
• Use two schottky diodes on the Espruino pin, one to GND and one to 3.3v. This will 'clamp' the voltage to the correct range (but is overkill really unless you actually expect the voltage to go out of range very often)
  
  

Hope that's some help!