ideal trace width for 2A current???

Try this:

That depends on what ideal means to you!

If you want minimal impedance, than as wide as possible.

If you want to uses as less space as possible, determine the allowable warm-up of the trace and calculate the minimum width for that.

If...

--
Stef    (remove caps, dashes and .invalid from e-mail address to reply by mail)

Some time back I read 1A requires 1mm of trace. It was not mentioned internal or external. Now on board I have space constraint...so at the max I can make it

2mm. But is it really necessary? As I mentioned it is Non impedance controlled board. Signals have been routed in GND and VCC plane. I have to modify this board for 2A. Previous tracks are of 1mm for 1A current.

If you don't care about how much the voltage drops, or how long the trace lasts, you can make it as thin as your fabrication house can reliably etch it.

If not, then what Stef said is valid, and the calculator that Arlet gave you a link to is (if it's what I think it is) appropriate to give you the information you need.

Keep in mind that if you've been told to do this, you're probably being paid to do this: i.e. you're being paid to think. I can't speak for anyone else directly, but while I'm happy to dispense some knowledge on newsgroups for free, I generally like to get paid for thinking, too, so I don't do much actual number-crunching engineering for people on newsgroups.

So this'll be the third response with knowledge, now it's your job to go heat up some brain cells.

There are two reasons for making the trace wide: One, because it'll drop voltage proportional to resistance and current. At high currents, if you don't want too much voltage drop you'll need to bring the resistance down. Two, because that voltage drop times that current will cause resistive heating. At high currents, if you don't want the trace burning off the board then you'll need to bring the resistance down.

There are three factors that you can control: One, you can order the board made with thicker than normal copper, so that each trace, for a given width, has less resistance per unit length. Two, you can make the trace wider, which both decreases the resistance per unit length and increases the ability of the trace to dissipate heat. Three, you can have bare wires soldered on top of the traces, or you can replace the traces with insulated jumpers (I wouldn't do this unless I were desperate, but you never know).

--
Tim Wescott
Control systems and communications consulting
http://www.wescottdesign.com

Need to learn how to apply control theory in your embedded system?
"Applied Control Theory for Embedded Systems" by Tim Wescott
Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html

You might also be able to duplicate the trace on multiple layers, and connect them with vias.

Two things you want to control : track heating, and track voltage drop. In some cases Voltage drop is the more agressive spec.

Some remedies: # Double the copper thickness # use copper pour, to maximise the width, and neck as needed # Run multiple narrow traces # shorten the trace, by moving components, or raising its channel priority

-jg

There is also another option - the track does not need to be of uniform width. If the width constraint is so that it can be routed between other components, then you can have a fairly wide track either side with a narrower section where space is at a premium. This is a good method for controlling voltage drop and overall heat output, although the heat generated locally at the narrower section will be the same or possibly even higher than the narrow-all-along equivalent. How these balance out is ultimately system-specific.

--
Andrew Smallshaw
andrews@sdf.lonestar.org