Yesterday during calibration, I fried my second arduino. I believe that this happened due to a very high inside temperature (around 45°) and then the arduino with the shield on top running hot… I haven’t been smarter the second time, but now I am constantly using a fan.
It seems like the ATMEGA2560 chips next to the USB outlet are what is burnt. The crystal next to it (the long silver thingy) is getting really hot.
Now my question is, does anyone have any experience of soldering a new chip onto the board? I don’t like the idea of just buying a new one very much, I want to try and repair as much as possible. Is it close to impossible?
I have found a chip here:
and instructions on how to write the bootloader on the fresh microcontroller:
unless you have a solder reflow oven, you are better off just buying another board, especially given how cheap they are. a very small piece of aluminum sheet metal heat sink might help. however i have not read any other people having this issue so I’m guessing there is something wrong with your system.
you are only putting the usb cable in the arduino right? the power cord goes into the motor shield on top.
Take a good look at the power plug on the motor shield and the USB plug on the Arduino from the side.
The distance is so tight that a ‘solder-ball’ could pass power down and destroy.
Sorry for the blurred picture, but i guess you get the point.
Can you send us some pictures of how you are set up? Where did you get your parts? I have not killed a single Mega on a Maslow and I tested over 100 shields.
If it were running hot for some reason you could add a heat sink. Shouldn’t need it.
there is a closeup of my setting, I made the board by myself, but luckily I have another original one with me for testing (its a friends, who got the whole kit)
Now I am observing that also the third arduino, which is the friends keyestudio mega2560 is also running hot at the same place at the bottom of the arduino… Even when I plug it into the computer via USB its running hot real quick. Now I am not plugging it anywhere, because I am afraid to fry also this one…
after some research, I have found that it must be the voltage regulator on the arduino that burns first, then some other parts start overheating and the board fries… Now why that is happening in my case, I dont know… I’ve got my USB cable connected to a raspberry Pi, set up with WebControl. I am using the original power supply and had it only plugged into the shield, not the arduino itself.
Small edit: I’ve got motors that are not original. Could it be, that they are drawing too much power - which then makes the voltage regulator on the arduino fail, which then results in a fried arduino chip?!
My guess would be that somehow your controller board is back feeding the voltage from the power supply to the Arduino. The 12 volt power-supply should be completely independent from the Arduino.
Shouldn’t be. The motors draw the power from the power supply plugged into the motor shield, not from the Arduino. Downwards there should be no power passed from the shield down to the Arduino. Upwards only Hi/Lo signals and the PWM for the motors should go up.
I agree with Bar that something from the shield is feeding back. For periods of 30 - 45 minutes (sometimes longer) i’m running Arduinos at 50° - 55°C (122° - 131° F) ambient temperature without ever killing one.
Measure from GND to all pins going down to the Arduino. You should not find 12V on any of them.
Thank you for your replies so far. I will go ahead and solder a new voltage regulator on the third arduino which is running hot, but at least still recognized by the computer.
How would that manifest itself? Would it mean that there is a piece of solder thats touching the arduino somewhere? Could an unstable power supply, such as a slack joint be the reason?
I will do try that, as soon as I am getting back to my space by tomorrow!
I couldn’t find any pin going down to the arduino with more 9V on them when measuring from GND.
Could it be that one of the motor drivers on the shield is faulty?
I also replaced the voltage regulators (on all three arduinos). Two of them seem ok, the “L” and “On” LEDs are both lighting up. But most likely the chips are burned… (I went through the whole process of ICSP, using an arduino UNO, connecting it to the faulty mega2560 and checking with the AtMega_Board_Detector sketch. → GitHub - nickgammon/arduino_sketches: Publicly-released sketches for the Arduino microprocessor )
I have been testing my shield, also comparing it with an original shield. Both shields show more than 5V on the pins going down to the arduino (some are around 7V). Are you sure that 5V is the maximum output of the shield to the arduino?
I am still struggling to find the mistake on the shield… Maybe it is one of the transistors thats bad?
One more edit:
The voltage seems to be changing. I measured up to 18V (?) on my self-assembled board. But the numbers are not consistent… It seems to be the pins that are connected to the H-Bridges
I have got it connected to the original 12V 5A power supply
I have singled out 4 pins, that carry 5V on the original board, that carry between 14 ~ 15V on my self-assembled board.
These are the pins that connect into the arduino on the POWER line:
IOREF
5V
GND
GND
The 5V pin is very near one of the motor power pins on the 6-pin motor connectors. A short-circuit there might cause that. Do you still see that when you plug the 12v into your board without a Mega attached?
Edit – be very careful with your meter probes testing in the motor connectors; I’ve fried a Mega by accidentally causing the short I mentioned with a probe-tip. Best to make the measurement without the Mega attached, or before the fourth cup of coffee…
Thanks for your reply. I haven’t got the mega attached (also because I don’t have a working one anymore)
Can you tell me how to trace the short circuit? There are no pins or soldering blobs connected with each other
Which board are we troubleshooting? Can you give a link to the files you used?
Best to do the initial testing without power, using the meter’s continuity beeper. Is there a short between 5V and Gnd? Between 12V and 5V?
