Fried Moteino - What went wrong?

[Stark]

Hi there.
Disaster struck a Moteino!

I require a non battery supply for an always awake Moteino. Well I thought why not apply 5v from a 5v 2a charger applied to Vin and Gnd. Simple solution, apparently not, as a small puff of black smoke from the Moteino was the result.

The Moteino (RIP) was connected to an RTC, 433tx for RF light socket switching,  a lux sensor and a token buzzer. RTC and buzzer powered via digital pin the rest via 3.3v. All grounded back to Gnd pin.

RTC powered via digital pin to avoid it constantly drawing current. Lux sensor figured always on as it is a low drain. 433tx was always on as a work in progress.

This setup worked fine with a 3.7v lipo.

The result now is that all peripherals seem fine when tested with another Moteino. But Moteino RIP;
Sending lux readings ok.
Switching 433tx not working, either not receiving radio command or not sending data via digital pin output to the 433tx.
Cannot be reprogrammed via FTDI Moteino did one reprogramme now bricked.
RTC seems ok and reported time via serial.

Any ideas what went wrong?
Is this a case of excessive current draw?

I can live and learn with this but still need to find a safe solution for the USB power (nonbattery) option.

[Felix]

I power Moteinos from 5V wall warts all the time, the Vin will take up to 16V without smoke. Something else must have gone wrong.
Sounds like the radio is fried. It could be that just part of one of the silicon chips is fried and that causes the Moteino to act weird and some things work some don't.
Powering very low power devices from the Moteino GPIO is ok, like 10-20mA tops. Using 2 pins you can even turn on a small signal relay like the one I use in various projects (garage mote, doorbell mote etc).
A power surge at power up might kill the regulator and everything along its path.

[Stark]

Thanks Felix for the comment.

I have a feeling the radio is still working given that I could still send the lux sensor readings to the gateway.

The switching didn't work but I get the impression the digital data pin to the separate 433 TX was the culprit and that the RFM69 was still receiving the switch commands.

I tested the wall wart at 5.2v so nothing out of the ordinary there.

The peripherals current consumption are all within 10mA maximum and below.

433 TX 10mA
RTC (DS3231) 200uA
Lux (TSL2561) 0.6mA

The RTC starts immediately on setup via a pin set to Output High and was set this way to save power when running on battery. I suspect a spike was the issue here.

I am a bit apprehensive about trying again with a new Moteino. Is there any safeguard I could introduce, would resistors on the digital power lines help with any spikes?

[Felix]

Hard to say or guess. Have you checked everything is wired correctly and there are no shorts or anything like that?
What is the "433 TX" after all? It's not a RFM69 right?
To find stuff like that you have to isolate the problem. You can start by hooking up piece by piece until it won't work any longer, looking on a scope also helps to see what's going on in the circuit.

[Stark]

Sounds a sensible approach will build up and monitor as I go.
Will report back if I find anything of interest or only as to serve as a warning to other novices like myself 

As an of topic background....
The 433 TX refers to a cheap 433mhz radio transmitter module its being used to replicate the RF remote control for switchable sockets I have around the home. All the user control commands are sent via the RFM69 as network messages with the actual switch codes left to the 433TX module. Works a charm. Now looking at the integration of a receiver module into the Moteino network for monitoring of on off state when the standard remote for the switches is used.

Hence I have a great need to get away from the battery power but without destruction!

[Felix]

Another random thought - avoid ground loops, in case your wall wart is not galvanically isolated from mains.