Hi!
I'm in the middle of designing a PCB for an audio power amplifier. However I don't know how to estimate how wide the tracks have to be for the more high-current parts, like output, ground connections, power rails etc. Also, I'm not very familiar with calculating the maximum currents that can be expected in these lines.
Any input much appreciated!!
Thanks!
Per
I'm in the middle of designing a PCB for an audio power amplifier. However I don't know how to estimate how wide the tracks have to be for the more high-current parts, like output, ground connections, power rails etc. Also, I'm not very familiar with calculating the maximum currents that can be expected in these lines.
Any input much appreciated!!
Thanks!
Per
Last edited:
For 10 A you'll need approx. 5 mm traces at 70 um and signal traces 0.5 mm.
PCB Trace Width Calculator
http://circuitcalculator.com/wordpress/2006/01/31/pcb-trace-width-calculator
Trace width is very dependent if the trace is at the edge of the board, if the pcb is mounted horizontally or vertical etc.
PCB Trace Width Calculator
http://circuitcalculator.com/wordpress/2006/01/31/pcb-trace-width-calculator
Trace width is very dependent if the trace is at the edge of the board, if the pcb is mounted horizontally or vertical etc.
Last edited:
Remember also that when it comes to grounds you MUST consider what is actually going to be flowing in the conductors too. You HAVE to design so that all signal grounds are uncontaminated. Increasing the print size 10 fold will not eliminate this kind of issue.
As a classic example of that only today I was looking at an old Rotel amp and the 1 inch or so square copper print that connects the two smoothing caps together has a clearly visible ripple component across it of the order of several millivolts.
You must treat each conductor as a resistance... if it helps put nominal values in of say 1 ohm and ask yourself "would that cause a volt drop that would influence the signal ? " the answer must be NO 🙂
As a classic example of that only today I was looking at an old Rotel amp and the 1 inch or so square copper print that connects the two smoothing caps together has a clearly visible ripple component across it of the order of several millivolts.
You must treat each conductor as a resistance... if it helps put nominal values in of say 1 ohm and ask yourself "would that cause a volt drop that would influence the signal ? " the answer must be NO 🙂
Ground is very critical in good analog pcb's so it can be a good idea to let us review your work, free of charge 
That would be very kind of you!!🙂
So, as I understand it, the best thing you could do is separate all the signal grounds from all the other grounds, and then run two individual cabels to the star ground?
So, as I understand it, the best thing you could do is separate all the signal grounds from all the other grounds, and then run two individual cabels to the star ground?
Tinning the track manually, with a good thickness of solder, can up its current rating significantly.
Tin is by a factor of 10 a worse conductor than copper, any solder ontop of a trace maybe looks good but won't lower resistance much.
Anyway, with 70mum boards currents as high as 15A are not a big deal anyway.
Have fun, Hannes
Anyway, with 70mum boards currents as high as 15A are not a big deal anyway.
Have fun, Hannes
This is a useful reference.
The CircuitCalculator.com Blog » PCB Trace Width Calculator
There's other stuff that might be helpful on that site including a calculor to give you the thermal resistance of different sizes and plating thickness of via holes. It proved very useful for me when I had to design a board with surface-mount power transistors on one side, but the heatsink was fitted on the other side of the pcb. Multiple thermal vias from the transistor pads on one side, to similar pads on the other, (which then passed the heat to the heatsink/radiator) worked very well indeed.
The CircuitCalculator.com Blog » PCB Trace Width Calculator
There's other stuff that might be helpful on that site including a calculor to give you the thermal resistance of different sizes and plating thickness of via holes. It proved very useful for me when I had to design a board with surface-mount power transistors on one side, but the heatsink was fitted on the other side of the pcb. Multiple thermal vias from the transistor pads on one side, to similar pads on the other, (which then passed the heat to the heatsink/radiator) worked very well indeed.
- Status
- Not open for further replies.