current capacity of PCB traces

InSilico

Hi!

[noparse][[/noparse]This post was moved from another forum because this is probably a more appropriate category]

I will be implementing a light dimmer circuit onto a custom PCB I'll make myself.

Some parts of the board may need to carry up to 10A while some other parts only need to carry up to 4A. Both will carry 120VAC. The traces will be very short (probably no longer than 1-2 inches). Everything will be fused, and safety precautions taken of course!

I'm wondering if anybody has worked with high-amperages and high-voltages on PCBs before? Basically I want to know how wide should I make the traces be, and how hot is it going to get when it's running at full capacity? (I'm using 1 oz./sq. ft. copper clad board) Different sources give me different answers.

Or should I just simply solder 18 AWG wire directly on the pins (a la protoboard) to err on the side of caution? After all, a fire is the last thing I need.

Some advice·is much appreciated!

Post Edited (InSilico) : 8/27/2008 7:06:56 PM GMT

Leon

If it's a one off you could just solder heavy wire to the tracks. I've done that before now.

How different are the answers?

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

Paul Baker

There are alot of "hidden" variables when calaculating the correct minimum trace width such as temperature rise of the trace, if different values are used, different widths can be obtained for an answer.

This should help you http://circuitcalculator.com/wordpress/2006/01/31/pcb-trace-width-calculator/

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Paul Baker
Propeller Applications Engineer

Parallax, Inc.

Post Edited (Paul Baker (Parallax)) : 8/27/2008 11:59:17 PM GMT

InSilico

Thanks for your answers!

@Leon: This is something I will make duplicates of (for different channels) so soldering wires, while a good idea, will become a pain in the butt to solder them individually.

@Paul: That might explain why I'm getting different recommended width sizes.

The PCB Trace width calculator is rather cool...

I'm rather concerned that it assumes a 10 degrees C (50 degrees F) temperature rise, which seems to me rather high... is it alright to let a board get that hot (or am I underestimating the capabilities of PCBs)?

To be on the safe side I'm making my trace widths wider than the recommended amount (I have board room for that).

John Bond

Hey - That's a usefu site Paul. Thanks

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

Dave Paton

Just an aside. I've designed and built dimmers and phase controls for everything from 3V LEDs to 480/3ph motors and outdoor power systems. The thing you'll NEED to account for is creepage distances. The 120V waveform has a peak voltage around 170V, and can arc surprising distances in surprisingly low humidity. UL and CE have distances for AC line creepage, and I'd give them a serious look. Also make sure you can manage those distances on your power components. I've had to create special footprints for TO220s to prevent arcing.

My $0.02, since I hate shoot-through shrapnel.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
This is not a sig. This is a duck. Quack.

Leon

10C rise is only 18F!

Leon

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Amateur radio callsign: G1HSM
Suzuki SV1000S motorcycle

InSilico

@Leon: Whoops, I've converted C to F before subtracting...

@Dave: That's definately something I need to worry about, thanks for telling. For a 120VAC system using FR-4 boards, what do you think is an appropriate distance between two 120VAC·traces? What special footprint did you use for your TO-220s? (This interests me because my triacs are TO-220s)

Post Edited (InSilico) : 8/29/2008 4:56:10 AM GMT