Big Old Fat Traces (High Current PCB)

[the_enginerd]

I'm designing a PCB for an audio amplifier that has the potential to put out 40A (instantaneous), and honestly, I'm out of my comfort zone. The connectors are rated to a min 35A and I'm basically going for big fat traces to take the current load.

All components in the assembly are pass-thru connectors, nothing to overheat besides the board itself.

• The narrowest neck is 5.5mm thick (besides certain spots where I have to taper down by the molex)
• I plan on using 4oz copper
• I read about doing a waves solder to help as well, same goes for buss bars
• Trace/Pad clearances are 50mil (plz don't harass me for using mm and mil)

What I'm looking for answers to
1) Is this just dumb? Am I crazy?
2) Any suggestions
3) General high current PCB design suggestions

Proto-types for testing will be done, but I'd rather not spend the money if this won't work out of the box.

Thanks for the feedback

[T3sl4co1l]

Audio?  The RMS level over thermal time scales (seconds) is a tiny fraction of the peak.  10A would be overkill, I would think.  Don't worry about it!

Tim

[DerekG]

I'm designing a PCB for an audio amplifier that has the potential to put out 40A (instantaneous) .... The connectors are rated to a min 35A and I'm basically going for big fat traces to take the current load.

So you need to ask yourself - Under a fault condition, can the 40A be continuous? If so, perhaps add in a slow blow fuse/circuit breaker rated at 35A to prevent connector/pcb track failure.

You will need to choose the "slow blow" fuse/circuit breaker with care.

Open up the solder mask as you have suggested to coat the current tracks with a bit more tin (tin/Pb) during wave soldering. Perhaps look at which way the board should travel through the wave solderer, remembering the leading edges of the thick tracks will receive a "normal" amount of solder & the trailing edges of the thick tracks will often receive a "thicker" amount of solder due to the meniscus effect.

Also look at the maximum temperature the board will rise to when being pushed hard.

[the_enginerd]

Audio?  The RMS level over thermal time scales (seconds) is a tiny fraction of the peak.  10A would be overkill, I would think.  Don't worry about it!

Tim

This is what been pondering in my mind. We are setting up a test to nail this down tomorrow but this is what we are hoping for. This also makes sense as to why when I look at the same manufacturer's speakers and components in the amplifier that I see parts rated to a more reasonable 15A.

Would you think if we find that it's at a much more reasonable level that I could employ my fat traces but go down to 2 oz CU?

[LapTop006]

Would you think if we find that it's at a much more reasonable level that I could employ my fat traces but go down to 2 oz CU?

You could also get a test board with standard 1oz copper and measure what the actual trace resistance is to see if you really need to care.

[free_electron]

even though your board may be 4 oz copper : the plating in the holes won't be ! So whenever you experience a layer-hop : shoot a ring of via's around the holes.

For example : your power arrives on bottom layer to a capacitor pin , then leaves on top layer. Don't bank on the solder and the pin ! it may not flow up far enough in the hole to guarantee a nice solder fillet under the component. It wouldn't be the first board where the plating burns out...