Queries regarding Power Planes

[cks3976]

Dear Gurus,

We are designing a 12-layers PCB that should carry very high current (70A) from a 48V source. I have reserved 8 layers for Power and Ground signals. Mounting holes and thru-hole pins on the board have caused significant void areas on all these power planes. Dimension of the CB is 17.0" X 1.80".

A minimum of 900 mils of conductor width is needed on 2 ounce copper to carry a total of 76 Amperes of current as per some calculator. But due to these voids I am getting a minimum of 200 mils of conductor width at some places of the PCB. We have multiple power planes on the PCB with almost similar copper area on multiple layers that are stitched using vias (16 mils drill and 32 mils pad).

Am I doing right, or are there anything that I am missing here?

Any help in this is greatly appreciated.

 

[praveenlb]

due to void u are getting fear.some places it is 200 mil trace width ....can you able to increase 2 to 3 ounce copper thickness ..please make sure thermal related issue of your card..
Regards
Praveen Bhat

 

[marce]

It can get complicated when trying to transfer so much current. The current carrying capacity of copper is based on its thermal charecteristics, ie how much current the copper can carry without significant temp rise, and thus resistance rise, then more temprise etc etc then a very smelly PCB with black scorched bits, but if the necked area of copper is small then he temperature rise at that point will not be so great as copper is a good thermal conductor. In situations like this simulation software is useful, I use outside expert help as the software is expensive and has a learning curve, if that is not available, I would have the board built and test it at the extremes of the temperature range using a thermal camera and/ortemp probes fastened to the board.
If this product is going to be sold in the USA you will have to get UL certification, for current of this value they will measure the PCB temp rise.
Sorry there is no easy answer when you start moving this level of current, other than simulation and/or testing.

 

[cks3976]

That was very useful Marce. I think I will have to ring-up the simulation guy. Will do that quickly. Thank you verymuch

 

[marce]

Its the safest way with high current designs, have fun and good luck.