3d Printer ATX power distribution and control board.

[TbirdMan]

Hello there. I am an Engineering Physics grad student in the USA. I have decided to take on a fairly simple project for my first PCB design. The design is intended to allow the adaptation of a standard ATX PSU to supply a 3D printer. the ATX plug is broken out to a header for the PS_ON and PWR_GOOD pins. there is a USB pass through that has a select-able (via jumper) +5v source between 5v standby, the main 5v from the PSU, or the USB input 5V bus. this will allow the electronics package to be powered independently of the ATX power supply so the controller can turn the ATX supply off when not needed. I also decided to breakout some 5v and 3v3 general purpose pins (they're there so why not?). the main rails are powered by two 6 pin PCIe power plugs and are intended to provide up to 15A each with a fuse in line.

what I am hoping for help with is the main rail design. I think the GND should be ok for 30A, but the 12V lines are what concerns me. I have left the mask off the traces to allow for adding solder to beef up the capacity, but I would prefer not to have to do that. the width at the narrowest is 4mm but the traces are doubled up top an bottom so it is more like 8mm but I understand that top and bottom is not exactly the same as doubling the capacity. I am including images of the top and bottom for reference, let me know if you need anything else.

[prasimix]

...but the traces are doubled up top an bottom so it is more like 8mm but I understand that top and bottom is not exactly the same as doubling the capacity.

Why is that? I know that internal and external layers is not the same when we talk about temperature rising for the same current but top and bottom external layers should be the same. 8 mm can deliver almost 11 A with 10oC increase on 36 um cooper (KiCad calculator, see picture in attachment).

[TbirdMan]

You may be right, again I am new to this, but approaching this from a physics standpoint, one path for the trace to dissipate heat (into the PCB) is shared by the top and bottom. I would assume this would result in less efficient dissipation by the PCB since you are putting twice the heat into the same effective sinking area. I think I am gonna make the board a bit wider and add a couple mm to each trace. that should allow the traces to carry the current without the need of extra solder. when I get the boards made I will run some static load tests at 15A to see what kind of heat rise I get. Again, this is new to me, so any input is welcome.