Track Width for 6 A, 2oz

[PowerIT2020]

Hello,

I will have a max current of 6 A flows in the PCB, i chosed copper thickness: 2 oz
And i calculated it from: https://www.desmith.net/NMdS/Electronics/TraceWidth.html
It give nearly 70 mil track width for external but 186.74 mil for internal, my pcb will have solder mask layer over the copper layer then which one i should use? (I'll use 70, this will result in overheating?)

I'm confused between what internal and external means? I have only one copper layer at bottom and over it solder mask layer. I'm noob in this, just started.

Thank you.

 

[ZASto]

Internal means that the track is located in sandwich of two insulators, i.e. inner layer.
Solder mask could influence a bit, but not much. Your calculations are ok.
Here is from KiCAD's calculator:

 

[PowerIT2020]

Thank you ZASto, so as i told i have only one copper layer at the bottom and over it just the solder mask layer, so it's OK i use the 70 mil?

 

[KlausST]

Hi,

For a more exact estimation you need to take into account:
* what do you call "overheat"? 100°C? 150°C? Fire?
* PCB material
* other heat sources
* air flow
* altitude
* case, material, closed?
...

Klaus

 

[PowerIT2020]

Hi Klaus,

Normal PCB from fiber glass FR-4, other heat sources will the the component themselves but not much, the working temperature maybe 70°C (Including ambient), will case it later and add fans for ventilation.

 

[KlausST]

Hi,

A rule of thumb says that every 9°C of temerature rise will reduce lifetime to 1/2.
18°C --> 0.25
27°C --> 0.125
and so on.
If you want high lifetime, make low temperature. Short, wide, thick traces.
If you want early failure make high temperature.

Klaus

 

[PowerIT2020]

OK, then i must lower the rise time but it will give more thicker tracks.

 

[KlausST]

Hi,

OK, then i must lower the rise time but it will give more thicker tracks.

You speak in riddles.
What has rise time to do with?
And why thicker?

Klaus

 

[PowerIT2020]

Sorry i wrote wrong, i mean if i reduced the temperature rise (not time rise) the tracks became thicker.
Another thing need to know, how the power supply pcb withstand 35 A and more? at same time it's not expensive because total cost of power supply not high.
My understanding is that i'll make the tracks as big as i can, and use highest copper weight.

 

[ZASto]

As you use wider tracks, temperature rise will be lower.
Consider copper pours instead of tracks for high amp current path.

 

[KlausST]

Hi,

Good point.
I don't recommend copper pour as GND plane.... but here copper puor may bring significant benefit.
It may lower the power dissipation, but also improves heat spreading.

how the power supply pcb withstand 35 A and more?

A useless question .... as we don't know what power supply you are talking about.
* It may be a pulse power supply,
* it may use thicker and wider and shorter traces,
* it may use aluminum core
....

Please confirm your value: the given 6A are "6A RMS, low frequency over more than 1 minute"?

Klaus

 

[PowerIT2020]

Hello,

I mean my computer power supply called: switching power supply and it's rated for 35 A, i remember i break one it's on pcb i ask because heavy weight pcb expensive.

My circuit will work on 6A DC, but there is capacitor supply ripple current with 60 kHz switching.

ZASto i don't know much about copper areas, will google it soon.

So guys lastly to say roughly it's safe i use external track width for my case (as said just solder mask over the copper and copper at bottom, at the top the component will exist the FR4 isolate the two)?

- - - Updated - - -

Can't edit the above post, i covered my bottom layer with copper area as ZASto said. But also KlausST said that this will increase heat spreading i agree with you both and confused what to do,
Larger area of copper means good dissipation but at same time the large area the heat will speard and heat up the pcb.

What you recommend?

 

[KlausST]

Hi,

will increase heat spreading i agree with you both and confused what to do,
Larger area of copper means good dissipation but at same time the large area the heat will speard and heat up the pcb.

Heat spreading is good, because it reduces temperature. The same heating power (energy) becomes spread over a larger area.
It reduces hot spots.

Klaus

 

[PowerIT2020]

Though that you mean heat will spread to heat the FR4 layer and the components, OK got you.
I'll make copper area as you guys suggest, and make the tracks as big as i can.

Thank you both ZASto and KlausST :smile:

 

[ZASto]

Here is a piece of PCB that I designed long ago which carries 10A per contact:

 

[KlausST]

Hi,

Those short pieces of traces may carry a lot more current than the calculation says.
It's a game of generated heat power, area to dissipate this heat, ambient temperature, ambient heat sources..

Klaus

 

[PowerIT2020]

Here is a piece of PCB that I designed long ago which carries 10A per contact:

View attachment 158066

The thin tracks carry 10 A? or only the big ones?

 

[ZASto]

Only the fat ones. These are connected to A1 and A2 of Triac.
The thin ones carry almost nothing

 

[PCB Designers]

Hi,

Those short pieces of traces may carry a lot more current than the calculation says.
It's a game of generated heat power, area to dissipate this heat, ambient temperature, ambient heat sources..

Klaus

Once I had to design a board with 20A Current carrying conductors. At that time no info was given about temperature.

I took 2oz Copper and 200 mil track width. (Calculated from some application note)

The board worked fine and no heating problems occurred. (Today, I feel, it was a wild hit and trial approach )

Klaus, your comments are requested.


Regards,
Maqbool

 

[marce]

Download Saturn PCB toolkit (a search using the previous text will find it), most people use this to determine track widths, it is also based on IPC-2152 specification, also worth a look at if your serious about design...