PCB track current capacity

[Claus Jensen]

What is the rule of thumb for the current carrying capacity of a
tinned copper PCB track on standard substrate?

For example, what width would be required for 1 amp at 100Hz 50% duty
cycle?

Is there any process by which the track can be made less wide, ie. a
thicker deposition of copper on the board prior to etching.

Thank you,

Claus Jensen

 

[[email protected]]

Type "pcb trace calculator" into Google and you'll get plenty of
calculators.
Usual rule of thumb is to calculate width for a 10degC rise in temperature,
but your electrical requirements will always come first of course.
It's not an exact science though, calculations are based on old extrapolated
IPC graphs and can be quite error prone.
It also varies based on solder mask, and whether it's an inner or outer
layer etc.
Good info is here:http://www.ultracad.com/article_temperature.htm


Yes, that is common.
Normal boards have either "0.5oz" or "1oz" copper thickness (weight
actually), but you can specify up to say 4oz copper from your PCB
manufacturer. That allows greater current capacity for a given width.

0.5oz (per square foot) is 18um
1 ounce copper per square foot is 35 um thick and so on.

4 ounce copper would be 0.14mm thick, and really closed-spaced narrow
tracks become impracticable - the gaps between the tracks need to be
wider than they are deep.

 

[David L. Jones]

What is the rule of thumb for the current carrying capacity of a
tinned copper PCB track on standard substrate?

For example, what width would be required for 1 amp at 100Hz 50% duty
cycle?

Type "pcb trace calculator" into Google and you'll get plenty of
calculators.
Usual rule of thumb is to calculate width for a 10degC rise in temperature,
but your electrical requirements will always come first of course.
It's not an exact science though, calculations are based on old extrapolated
IPC graphs and can be quite error prone.
It also varies based on solder mask, and whether it's an inner or outer
layer etc.
Good info is here:
http://www.ultracad.com/article_temperature.htm

Is there any process by which the track can be made less wide, ie. a
thicker deposition of copper on the board prior to etching.

Yes, that is common.
Normal boards have either "0.5oz" or "1oz" copper thickness (weight
actually), but you can specify up to say 4oz copper from your PCB
manufacturer. That allows greater current capacity for a given width.

Dave.

 

[C;aus Jensen]

Type "pcb trace calculator" into Google and you'll get plenty of

calculators.

Dave.

Thank you for your advice. The calculators appear to assume DC
current. How would one adjust the final figure for a lesser duty
cycle? In practice, is the relationship purely linear? For example,
would a 50% average duty cycle mean you can simply use a 50% thinner
trace?

Claus Jensen

 

[C;aus Jensen]

Type "pcb trace calculator" into Google and you'll get plenty of

calculators.

Dave.

SORRY, I left out the key problem in my last post. I need to draw two
spirals, one within the other, with equal spacing between the traces.
IOW a double trace spiral. I can draw one spiral, but simply rescaling
the second one won't make it fit. Any ideas?

Claus

 

[Nobody]

SORRY, I left out the key problem in my last post. I need to draw two
spirals, one within the other, with equal spacing between the traces.
IOW a double trace spiral. I can draw one spiral, but simply rescaling
the second one won't make it fit. Any ideas?

Arithmetic spiral or geometric spiral?

For an arithmetic spiral (radius is proportional to angle), adding
a fixed offset to the angle (i.e. rotation) is equivalent to adding a
fixed offset to the radius.

For a geometric spiral, (radius is exponential to angle), rotation is
equivalent to scaling.

You state that rescaling doesn't work, so I'd assume that it's arithmetic,
in which case you should just need to rotate it.

 

[Eeyore]

What is the rule of thumb for the current carrying capacity of a
tinned copper PCB track on standard substrate?

See IEC / EN 60065.

I'm being mean since I have a copy here. It's all related to acceptable
temp rise. Give me a bit and I'll get back to you.

Graham

 

[Eeyore]

Yes, you can make the track thicker -- in the US this is expressed as
ounces of copper per square foot, with one ounce being pretty standard
fare, and Really Serious Power boards going up to 4 oz.

Dunno about trace widths, though -- someone will chime in.

1 oz copper is 35um IIRC etc. You'd be surprised how much current it CAN take
under surge conditions as well.

I'll see if I can find the chart.

Graham

 

[Eeyore]

Type "pcb trace calculator" into Google and you'll get plenty of
calculators.
Usual rule of thumb is to calculate width for a 10degC rise in temperature,

TEN degrees ? You have to be joking for power apps. 50C at least. Other matters
such as ventilation will affect results.

Graham

 

[Eeyore]

Thank you for your advice. The calculators appear to assume DC
current. How would one adjust the final figure for a lesser duty
cycle? In practice, is the relationship purely linear? For example,
would a 50% average duty cycle mean you can simply use a 50% thinner
trace?

Don't forget P = I ^2 * R. So NO.

Graham

 

[Jasen Betts]

What is the rule of thumb for the current carrying capacity of a
tinned copper PCB track on standard substrate?

For example, what width would be required for 1 amp at 100Hz 50% duty
cycle?

Is there any process by which the track can be made less wide, ie. a
thicker deposition of copper on the board prior to etching.

yeah, that, or plating extra copper on after etching

even a thick layer of solder will help a bit.

if it's localised to a small part of thr board you can "surface mount"
some copper wire on the track.

 

[Ben Bradley]

What is the rule of thumb for the current carrying capacity of a
tinned copper PCB track on standard substrate?

Here's a neat online tool:

http://circuitcalculator.com/wordpress/2006/01/31/pcb-trace-width-calculator/

Of course, while I don't know of any problem with that thing, you
shouldn't trust your career solely to a link you find online.

As always, all applicable disclaimers apply.

 

[Rich Grise]

What is the rule of thumb for the current carrying capacity of a
tinned copper PCB track on standard substrate?

Sure: look up ampacity of a given cross-sectional area of copper,
and find out how thick the copper is (is that spec easily available?)
multiply by the width, and that's the area.

Google "ampacity".

Cheers!
Rich

 

[Rich Grise]

1 oz copper is 35um IIRC

Thirty-five MICROMETERS? That's 0.0013 inches - doesn't sound like very
damn much copper. Where did you get this figure?

Thanks,
Rich

 

[Rich Grise]

Sure. Well, 50% of cross-sectional area. It's purely linear if you're
talking about pulsed DC. (fancy waveforms complicate the calculation.)

Don't forget P = I ^2 * R. So NO.

What the hell does this have to do with anything?

Thanks,
Rich

 

[Rich Grise]

]

SORRY, I left out the key problem in my last post. I need to draw two
spirals, one within the other, with equal spacing between the traces.
IOW a double trace spiral. I can draw one spiral, but simply rescaling
the second one won't make it fit. Any ideas?

Draw a trace that's the width of the sum of the two plus the gap; then
"un-draw" the gap out of the middle. ;-)

Hope This Helps!
Rich

 

[Martin Griffith]

Thirty-five MICROMETERS? That's 0.0013 inches - doesn't sound like very
damn much copper. Where did you get this figure?

Thanks,
Rich

e.g. from
http://www.eurocircuits.com/index.php/Service-overview/Service-Overview.html

Single-sided 35µm/1oz

martin

 

[Eeyore]

Sure. Well, 50% of cross-sectional area. It's purely linear if you're
talking about pulsed DC. (fancy waveforms complicate the calculation.)


What the hell does this have to do with anything?

Try doing the sums ?

Graham