High Power PCB

[eerocketman]

Hey guys,

I was wondering if anyone had any good ways of carrying high power across a PCB? It is a discharge circuit starting at about 500V at about 100A.

Thanks!

[RJFreeman]

I was going to say that I leave an open area in the solder mask so that you can solder tinned wire along it for high current paths, but 100 Amps is a far whack, so I would be thinking you need a bit more than that.
I would be inclined to use wires rather than a PCB trace for that.

[eerocketman]

It is not 100A for very long. The current and voltage ramp down (don't have numbers till tomorrow).

Instead of wires i was thinking about running long high current shunts on top of the trace.
Here is something similar to what I want to use. But the ones I have are about twice as long as the 1206 http://www.keyelco.com/category.cfm/Test-Points-Tips-Probes-Clips/Zero-ohm-SMT-Jumpers/id/1203

However in doing this I am worried about having non-insulated conductors at a very high voltage.

[digsys]

... However in doing this I am worried about having non-insulated conductors at a very high voltage.  ...

You have some serious thinking to do here. First, forget that SM pads idea. There is no way you'll cover the clearances and creepages required for the HV, and being DC,
makes it much worse. Those are way past lethal voltages !! I would use heavy copper flat bar, and have that section isolated, with insulated connections to whatever you're
monitoring to / with. Even if it only a mS pulse or less !! And even if you do work something out, any contamination / dust / ageing will likely fry in the future.

[daqq]

I was wondering if anyone had any good ways of carrying high power across a PCB? It is a discharge circuit starting at about 500V at about 100A.

Massive, thick, wide traces? Depends on the duration of the current pulses and their repetition rate really. You can also use thicker copper ( http://www.epectec.com/articles/heavy-copper-pcb-design.html ), combine multiple layers in parallel.

However in doing this I am worried about having non-insulated conductors at a very high voltage.

At some point you will have to have non isolated/coated connections either way.

[RJFreeman]

500V is not a drastically high voltage for PCBes, if you do the sums the Filter cap in a switch mode power supply will charge up to 340 Volts (unless you are in a 110V country).

IPC2221A suggests at least 2.5mm spacing for uncoated external conductors on a PCB with 500V between them, or 0.8mm spacing with a Polymer coating

page 53
https://www.google.com.au/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&cad=rja&uact=8&ved=0ahUKEwj2habwjIPVAhULzbwKHf3FC_YQFggoMAA&url=http%3A%2F%2Fwww.sphere.bc.ca%2Fclass%2Fdownloads%2Fipc_2221a-pcb%2520standards.pdf&usg=AFQjCNECilMoxY_eKkNOUqh538Xrmm2bEg

[T3sl4co1l]

What are you discharging?

Tim

[digsys]

500V is not a drastically high voltage for PCBes, if you do the sums the Filter cap in a switch mode power supply will charge up to 340 Volts (unless you are in a 110V country). 

For sure, but without knowing what the OP is doing, I certainly wouldn't specify it for 100A !! No matter how short :-)

[T3sl4co1l]

FYI, camera photoflash circuits are built on thin multilayer PCB or flex, and they do alright switching a hundred amperes or so, pulse widths in the low ms.  (They use a DFN8 size IGBT to turn off the xenon tube, allowing multi-flash and calibrated exposure.)

Tim

[Neomys Sapiens]

You should maybe calculate the energy using i²t and, using a estimate of your trace's mass, the resulting rise in temperature. That will give you an idea whether it vaporizes outright or if it is a matter of minor optimization.
For example, 100A for 1ms over 100mm of 1.5mm² will send any electrician seeing it into convulsions, but the temperature rise is merely around 20K. 10ms will result in 200K, which is not suffered well by most isolating materials.

[RJFreeman]

For sure, but without knowing what the OP is doing, I certainly wouldn't specify it for 100A !! No matter how short :-)

which is why my initial suggestion was to use wires rather than PCB traces....

[eerocketman]

It is a discharge circuit for a massive battery charger. I think i am going to use the shunts that I linked earlier and apply conformal coating. My current drops to about 10 amps in half a second.

[ahbushnell]

My current drops to about 10 volts in half a second.

Volts are not current