8
\$\begingroup\$

Ok, you can call me ridiculous.

But, When I make PCB's, I learned myself to get into a habit into making thick PCB traces for every wire including those that others would use maybe a 5 - 8mil width.

I also noticed that thicker traces allow me to use a junky printer to print the artwork to create decent PCB's with the photoresist method. With thin tracks, the printer sometimes prints a hole the size of the trace itself (5-8 mils). and No, I don't always have the time to use the office store for printouts.

Currently my minimums I set for my boards are 12mil width for traces with 12mil clearance. For power (except ground), I aim for at least 40 or 50mils to cover the width of a standard IC pad. For ground, I use 24mil max so it fits in-between pins.

Most of my circuits are powered with 5VDC but I have some powered with 3VDC and I also have part of my circuit that outputs loud audio as well as a radio module.

I used a 7805 voltage regulator without issues, so I can conclude my entire circuit uses less than 1.5A current total.

So my question then is, am I going overkill with my large trace widths? If so, what should I use as absolute maximums for trace widths and why?

and yes, I will use a ground plane, and all my PCB's are single-sided.

Added note

I should mention that the maximum speed passing through the majorit of my board at any one time is 24Mhz courtesy of a crystal, however those two traces (where crystal connects to micro) are less than 1/2 inch long and roughly 40 mils wide with a max of two 45 degree bends.

\$\endgroup\$
  • 2
    \$\begingroup\$ For the kind of low speed circuits you'd be making with these techniques, going for fat traces you can produce reliably makes all sorts of sense. You can often do 8 mil traces with toner transfer, but little reason to depend on that in areas of the circuit that don't require the density. Were you designing higher speed circuits you'd need to start thinking about impedance matching - but then you'd probably also need a more sophisticated process. \$\endgroup\$ – Chris Stratton Jan 15 '18 at 1:16
  • 2
    \$\begingroup\$ Beware though that if you work with small 2-lead surface mount parts, having different widths of copper on each side can cause "tombstoning" during cooling. \$\endgroup\$ – Chris Stratton Jan 15 '18 at 1:19
  • 1
    \$\begingroup\$ I don't use SMD so comment 2 will not apply to me. \$\endgroup\$ – user152879 Jan 15 '18 at 1:54
  • 1
    \$\begingroup\$ thats a pro but the con is the need to fine-solder the component in place with smaller spacing between pins. \$\endgroup\$ – user152879 Jan 15 '18 at 2:40
  • 2
    \$\begingroup\$ @teslajin and edit approvers, mil is not millimetre, mil is thousandths of an inch. Don't edit a question to change units without first understanding what the units are. \$\endgroup\$ – Tom Carpenter Jan 15 '18 at 16:36
8
\$\begingroup\$

If you're not designing for high speeds where the trace capacitance becomes an issue, I don't know of any reason to worry about traces being too wide. Carry on with your 12/12 default dimensions if you're able to fit your designs on your boards with them.

\$\endgroup\$
  • \$\begingroup\$ Can you please provide an example where big traces can screw up in HF/RF applications,? I am just wondering \$\endgroup\$ – Sean87 Jan 16 '18 at 18:02
  • \$\begingroup\$ @Sean87, anything above 1 GHz, you should probably use controlled impedance and then the trace width needs to be what it needs to be. Too narrow or too wide are both wrong. I also have had a 10 MHz bus where the traces were long enough that the capacitance was too much even with 6 mil traces, leading to faults. \$\endgroup\$ – The Photon Jan 16 '18 at 18:05
6
\$\begingroup\$

For home made PCBs, and to some extent professionally produces PCBs, the bigger the traces the higher the probability that the board will make it through the various manufacturing steps and come out with all the connections you intended.

So in your case you want to make them larger than the minimum size they can be in order to carry the currents involved.

However, on the flip-side, larger traces can make the board a lot harder to route especially between pads on through hole parts. (SMT too, but I would not even think about making boards that fine at home.)

As such, like most things in life, it is a trade of.

\$\endgroup\$
  • 3
    \$\begingroup\$ Just a comment, it's so much easier (I find) to do SMD boards at home than through hole. I've had plenty of success with SSOP (0.5 mm pitch). Certainly better than drilling 500 holes by hand (that was a bad day)..! \$\endgroup\$ – awjlogan Jan 15 '18 at 11:19
  • 1
    \$\begingroup\$ @awjlogan oh I agree, but transferring and etching very thin traces with kitchen equipment tends to less than reliable. \$\endgroup\$ – Trevor_G Jan 15 '18 at 14:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy