I am using the G400D and its built in Ethernet interface for a prototype project. Everything seems work fine except that the ATMEL controller on the G400D gets very hot. The G400D takes 250 - 300 mA with Ethernet enabled and data sending back and forth. Is it normal that G400D gets very hot? I feel like I can grill a cheese burger on it…
@ bbcat -
if you can grill a cheese burger then I don’t think it is normal…
1 Like
Well, the program still runs fine. Nothing seems wrong with the code or my printed board design.
I got another G400D yesterday, updated its tiny loader and firmware, deploy my program, put it on the board. Guess what, this one is not hot at all. I am confused with the variation of the G400D. Basically, the same version of tiny loader, firmware and code, and I put them on the same board (just by swapping the G400D). One is extremely hot, the other is totally normal. I don’t know why one of them is on the edge of frying, some manufacturing defects?
Disconnect everything beside power and make sure power is correct. Do not forget that G400 is not 5v tolerant.
Yep, the power is 3.3V. The thing confused me is that I have two G400Ds, I just swapped them on the DDR socket in my printed board, all the other conditions are the same. Both of them execute the program exactly as I want (even the hot one still works fine). The only difference is one gets very hot, the other is totally normal.
I would suggest you inspect the board with a magnifier, you might have a bad solder somewhere… a reflow might help.
As shown the attached image, I just swap two G400D modules on the DDR socket. The two G400D have the same tiny loader and firmware version, which is 4.3.6. They are running the same program and using the exactly same IOs. If there is any soldering or connection mess up on the printed board, both of them should have been affected with the same manner. But interestingly, one is very hot, the other is completely normal. I wish I have more G400D to make further comparison. But I suspect that one of the G400D module might have some manufacturing defects on it. I could not confirm that, just a guess.
I was talking about inspecting the G400D that gets hot 
not the printer :whistle:
And with your great detective work here, I’d approach GHI formally and see what they suggest you do…
Do you remember what phase the moon was in when you first powered up each board?