How big does my power plane need to be?

Discussion in 'General Electronics Chat' started by NGinuity, Jul 31, 2012.

  1. NGinuity

    Thread Starter New Member

    Jun 23, 2012
    Hi Everyone,

    I am designing a power distribution circuit that will be capable of supplying 40 amps total to the power plane at 12 volts. I have found a lot of literature on trace widths in comparison to current handling and copper thickness, but I was wondering if anyone knew the rule of thumb for the size of a power plane? I am looking to make the positive plane as an external layer with no mask over it.

    I plugged the following values into the calculator at

    45 Amps
    3oz copper
    10 degree Celsius rise
    65 degree Celsius ambient temperature (It could get pretty hot in the environment I want this in)

    This yields a trace width of roughly 2 centimeters. Can anyone confirm that this is in the ballpark of where I want to be or is the calculator rubbish? If it makes a difference for any of this, the boards will be plated with silver.

    Thanks for any assistance!
  2. Sensacell


    Jun 19, 2012
    Talk to your board house about what they can do for you as far as copper weight and plating options.

    Work backwards from "how big can it be?" look at your design and see what could possibly fit.

    @ 45 Amps, make no compromise.
  3. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    With a 65° ambient and just a 10°C rise your board will only get to 75°C... still pretty cool. You may run it higher; FR4 is speced to run up to 140°C. I mention this since I design lots of military stuff that needs to work in a 125°C ambient so 75°C seems chilly.

    That said, 45 amps is a LOT of current. You may want to supplement the copper with some nice bus bars. I just see traces flaming and vaporizing off the board when anything goes wrong.
  4. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    A few corrections.. Depending on your supplier FR4 is rated 105 deg C or 130 deg C. (UL maximum operating temperature..that's what really matters as far as this is concerned in the US or any other country really not the Tg..aka glass transition temperature) Very few suppliers (trust me) in the world can do FR4 with a 140 deg or more MOT.. I can count them on one hand.

    I run 100 Amps through PCB traces all the time (5/8" trace/double sided/3oz copper). The biggest issue is where that trace is located in relation to any components that are dissipating the heat. The trace (talking about the one I just mentioned) will run cool if the heat producing components are far away.. If they are close..It will run much hotter obviously..

    For a single sided trace 2 cm wide with 3oz base copper "should" be sufficient for 45 Amps.. Of course I have no idea how much heat your components are dissipating locally or anything else..
    In general the PCB trace calculators on the internet do a decent job and get fairly close and seem to have a slight safety factor from my own testing..
  5. NGinuity

    Thread Starter New Member

    Jun 23, 2012
    Thanks for all of the great information everyone! I'll try to respond to all of these as best I can.

    The board house will give me 5oz copper but the cost is prohibitive enough that I'd rather pay for the added trace real estate (I've got the real estate in the enclosure, might as well use it). I was going to take the figure the calculator gave me and derate it by 50% just to be on the safe side, so I would have 3cm traces instead of 2cm. According to the calculator, at 5oz, it only requires about 11.5mm. That said, I feel a little uneasy about 45 amps being on even a 15mm power plane because of inconsistencies that might exist when the board house etches it.

    I'm really trying to get away from a bus bar approach, but I understand the concern. On the input to the bus, it will have the appropriate fuse to mitigate that anything goes wrong scenario.

    That's actually comforting to know that I'm not out of the ballpark. The heat dissipating components won't even be in the same unit. I'm designing a power distribution unit and the power plane will be feeding a relay bank which then immediately feeds a block of connectors. The "switch" side of the relay is being driven by a remote unit. The "consumers" will be generating all of the heat in other units, so I figured a 10 degree rise is extremely adequate.