Atmega1284P and FTDI Programmer burnt out?

Hey everyone! New here, but had a bit of experience in electronics.

Working on a school project at the moment, and just received and soldered in my custom PCB.
It uses two microcontrollers - Atmega1284P and Atmega328 - to avoid some interrupt timing issues with some neopixels. Anyway, since moving the project over from a breadboard onto the PCB, it worked for a couple hours until the Atmega1284 stopped running it's program and became warm to the touch.

This happened when I connected a large LED Neopixel panel to 5V and powered the data connection from the Atmega1284. The power to the panel was connected to the power supply and the grounds of the circuit connected together, the chip shouldn't have been driving any current.
I'm using this lab power supply from Jaycar:

The chip alone drawing around 120mA which seems very high for this chip. The computer also doesn't recognise the FTDI programmer anymore, and all LEDs on the programmer come on when they didn't before. So I think I've busted the Atmega1284 and the FTDI programmer.

The resistance on the PCB between 5V and Ground without the Atmega is about 200 ohms, with the chip in its socket is 60 ohms. So it sounds like the chip is internally shorted. Maybe a voltage spike from the power supply?

Happy to provide more information to help diagnose the problem.


Schematic_CycleHubV3_2020-06-12_13-04-08.pdf (120 KB)

did you connect the neo pixels directly ?

The Data pin of the neopixels are connected to the Atmega through a 300 ohm resistor. The 5V and GND of the NeoPixels are connected to the power supply output, which share common GND and 5V with the Atmega (i.e. in parallel)

My guess is that you are missing a resistor and much current has been drawn someplace. BL_SIG is connected through a mosfet to GND without a resistor. Is it ok or does BL_sig source current and therefore needs one res? You use your analog output as a digital one in your scetch?(to drive the mosfet)