I was trying to make a mini version of the Arduino MEGA. Sort of a MEGA-MINI. I soldered everything together, but I can't get it to enter programming mode to write the bootloader. I connected it to Atmel Studio 6 and it just says "can't enter programming mode." I can't even read the device ID.
I connected my Arduino MEGA to the computer and I can read it fine. So my installation is OK as well as the drivers for my AVRISP mkII. I put two boards together. I don't think I over-soldered/killed either of the chips.
Can someone maybe take a quick look at my schematic and tell me if I screwed something up? Is there a design error?
I have it powered via my FTDI adapter and I get a green light on the AVRISP mkII.
I don't see a problem.
Is that reset label correct ?
Is your crystal running ? Perhaps try another crystal.
Can you measure the 5V ?
To use the ICSP header, don't connect the FTDI header and don't use the jumper on the reset line.
Vcc is good. Atmel Studio shows it as 4.8V with the FTDI adapter.
The crystal was the first thing I swapped out. I thought maybe I fried a cap or something while soldering it, but 2 boards? I doubt it. Both assemblies I put together appear dead as a doornail. So I though I had a design error. I'd hate to spend anther $18 bucks on another Atmega 2560 if the third board is going to end up the same.
I put a logic probe on the reset line and a push of the button sends it low. Normally it's pulled high.
I powered it off my breadboard without the FTDI adapter of the DTR jumper. Studio 6 reads the voltage as 4.8V. Still no dice. "Can't enter programming mode". I reconnected my Arduino MEGA, it reads it instantly. Swapped back to the MEGA-MINI, can't read it. It's like it's dead.
What would cause the chip to not go into programming mode? It's midnight, so maybe tomorrow I'll put a scope on it and make sure that the crystal is indeed running.
Other than that, you think my schematic looks OK? No stupid mistakes?
If the caps on the crystal are really 0.1 uF as indicated on the schematic, that is the problem. They should be around 22 pF, but in accord with the ATmega clock circuitry as well as the crystal specification.
Oh man! I think that's it! Talk about "forrest for the trees"! I know they're supposed to be 22pF. Why did I do that!? I must have used the clone tool on all the caps, and forgot to change the value. And then blindly soldered them in not even realizing the typo. What a dumb mistake. Uhhh, I'm kicking myself right now!
I'll swap the caps out tonight and see if that does it.
Yes, the crystal caps were obviously wrong, but that won't affect ICSP programming
will it ? Without flooding the ground planes its hard to tell what else is bad - C2 has
no connection at one end, so I presume that's ground plane, but I can't see a via that
leads to it.
C1 has its ground end lead to a pin on the package which seems suspicious, they
should both be going straight to ground plane.
Thanks again guys. It was the caps on the crystal. I put .1uF instead of 22pF on the PCB. I used the clone tool on the schematic for all the caps and forgot to change the value to 22pF. I didn't even notice I had made that mistake on the PCB and blindly soldered them in. I should have noticed when I laid my parts out that I didn't have any 22pF caps.
I also screwed up some of the pinouts on the LCD, but no big deal. This is just a prototype. So I just ran some wires across the board to another header. Everything works great now.
I'm working on a GPS clock. There'a s lot of hardware so I wanted to try out a test run with the 2560 just to make sure I could bootload one and upload sketches. This mini board basically qualifies my design. (I'd rather waste a $5 PCB that a $40 one!)
Anyway, I'm linking to a lot of libraries so I can't use a 328 like I usually do. My sketch is about 40K and there's more to do. I was going to use a 644, but I figured I go for broke and make it with a 2560. At least this way I have a crazy amount of code space for fonts, graphics, sounds.
