An Arduino Uno (R2) I'm working with does not respond to the computer (no serial communication, no USB notification)
What do I make of it?
I notice that when I check for response from the atmega8, avrdude responds saying that the device signature is 0x0000
Usually, when you plug in the ISP to the atmega8 it would respond.
So my tx and rx LEDS are constantly dimly lit (see attached).
From these I conclude that the chip is burnt. I know that in the past this board has been (unfortunately) submerged in water.
What should I do about this QFN that is possibly (hopefully not) burnt? I do not want to throw out the board... everything else is working fine but the converter.
Water on its own does not damage the electronics - they actually wash boards after assembly in some cases.
Salt or other debris might be a problem. Inspect it, and if there seems to be debris around the chips, wash the board and use a toothbrush to scrub around the chips. If you have an air blast, use it to remove the water around the chip(s) (but be careful of large components such as the capacitors) and dry it in the sun for a few hours.
I have not scrubbed it yet, because I do not have compressed air. I'll try that first before I scrub it XD
However, if shorts are not the problem, what is? The chip does not even respond to the USBTinyISP. Surely a fizzling short between two IO pins cannot make the chip unresponsive. If the short was between Vcc and GND the whole board would have overheated, so that is not the case either.
twang:
I have not scrubbed it yet, because I do not have compressed air. I'll try that first before I scrub it XD
That's not what I advised.
twang:
However, if shorts are not the problem, what is? The chip does not even respond to the USBTinyISP. Surely a fizzling short between two IO pins cannot make the chip unresponsive. If the short was between Vcc and GND the whole board would have overheated, so that is not the case either.
We are not talking about short circuits, we are talking about corrosion resulting in resistive paths.
If for example, you have a low resistance path between reset and ground, well, it will never go out of reset. This could easily be caused by water and corrosion in the reset button itself - check the voltage on the reset line with a multimeter. If the voltage is anything under three volts, de-solder the reset button and see how you go.
Come to think of it, we are talking about the 8U2/ 16U2 aren't we? Check the reset voltage on this (at the ICSP1 header) rather than the reset button - but check that also.
Stray resistance could also cause the crystal not to oscillate properly.
Yea I was just replying to someone else. I don't think I'll do that anyways
If the voltage is anything under three volts, de-solder the reset button and see how you go.
That actually makes a lot of sense. I just checked the voltage and it's a steady 4.87 volts from the RESET pin to the GND pin.
And, once I buy some compressed air I'll clean it up. Honestly I see nothing to clean, but I do understand that even minute particles could be a cause.
twang:
Yea I was just replying to someone else. I don't think I'll do that anyways
What I advised was scrubbing with a toothbrush and clean water. That is going to be much more effective than compressed air, particularly in dissolving and washing away any salt corrosion. The compressed air is just to complete the process and help with drying.
twang:
That actually makes a lot of sense. I just checked the voltage and it's a steady 4.87 volts from the RESET pin to the GND pin.
And that was on the 16U2/ 8U2 ICSP1 header rather than the reset on the 328 (which should probably be checked as well)?
