1
\$\begingroup\$

I'm designing a board that should handle a reasonably high current (30A) I would like to keep this on 2 layers (70um). For routing reason I need to apply the ground on the bottom layer (while positive is applied on top layer along with the 'load') in the bottom layer I have also shunt resistors. Is it possibile and 'safe' to handle such large current trough via stitching? There's some rule of thumb to 'size' vias in such cases (number and diameter)?.

PS I'm working in DC (up to 50V).

\$\endgroup\$
  • 2
    \$\begingroup\$ You can work out the voltage drop but this method demands a lot of solder fil. A better approach is a thruhole bus bar, \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Mar 23 '18 at 0:23
  • \$\begingroup\$ There is a track width calculator here but given your positive load current needs a return path why do you think via stitching will help? If you have positive and ground on both sides of the board it can but I would be considering either thicker copper or supplementing it in some way. Soldering copper braid in parallel may be an option for small quantities or mechanically mounted copper bars for larger ones. A single layer of 70um copper requires about 60mm wide tracks \$\endgroup\$ – Warren Hill Mar 23 '18 at 0:23
  • \$\begingroup\$ One square of copper foil of thickness 1.4 mils (the standard) has 1/2 milliohm resistance. At 30 amps, the power in that square (whatever size square) is IIR or 0.45 watts. In 10cm square, this is OK. In 1cm square, this is already too hot. In via with 1mm depth and 1mm periphery (also a perfect square, so is 0.5 milliohm), you'll have risk of a charred or discolored PCB due to heat. Can you extract 0.45 watts from a tiny plated-thru hole? \$\endgroup\$ – analogsystemsrf Mar 23 '18 at 6:31
1
\$\begingroup\$

The current capacity of a via is complex. The first factor is the amount of copper in the via which is determined by the plating thickness and via hole diameter. The typical plating thickness is 1 mil which equates to 0.7 oz copper. If we assume a via diameter of 10 mils, then the circumference of the via is 31.4 mils. Using a PCB trace temperature rise calculator this via could carry 1.6A with a temperature rise of 10 degrees C. However if the vias are connected to copper planes, there will be significant heat sinking and the vias could carry more current than you would expect. To make matters worse, components on the board that dissipate power will contribute to the temperature rise of the board. The most accurate way to estimate the temperature rise would be a finite element thermal analysis. I have not seen any published papers on this but I am sure it can be done but takes a lot of effort and expensive software. Another method would be to build a board and use a thermal imaging camera to measure the temperatures.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.