Rule of thumb for trace width

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

boji

Well-known member
Joined
Jan 6, 2010
Messages
2,375
Location
Maryland, USA
Was using calcs today and there's plenty of extra info needed for multilayer boards, etc that mostly don't apply in the basic, old school SS audio world.  I hope you don't mind the unscientific request, but I'd like to know your rules of thumb that lets you decide acceptable ranges of trace widths.

For example, in a small VU/peakmeter buffer board I'm working on,  .3048mm looks soooo tiny to my eyes**, (too small) but is likely perfectly acceptable for passing signal to a vu. 

**might have to do with the tons of 1.27mm widths I've  used in my first real projects.
 
For passing signal anything that won't break during the etching is perfectly fine. Most board houses have a lower limit of around 6mil without paying extra. I use 8-12 for general purpose stuff, depending on the overall design. For actual current there are tables, see attachment.
 

Attachments

  • deu_CMSO001_strombelastbarkeit_de.pdf
    31.1 KB · Views: 21
Sweet! Thanks Volker, quite helpful! 

Bulky buffer so far, using .6096.  No dire need for space saving, this project:
hkThxd7.jpg

 
Trace width is generally not an issue for low power signals...  I am old enough to remember when PCB processes were not as good as now. 

Back at Peavey (last century) I was limited to 20 and 20 (20 mil spaces and 20 mil traces) because of tolerance for silkscreening PCB patterns across 18" wide panels. Modern cheap PCB processes are now better than my old Eagle software can take advantage of.

JR
 
May I add that one 1uF capacitor is backwards and second one is redundant?

Moamps, you may do that kind of thing over my stuff till you are blue in the face!  8)

Let me consult ref. schematic to see whatcha mean...
 
Ok I found one, possibly two reversed (C7&C1), good catch. Thank you!
As far as redundancy, where do you see this, please?  (Apologies for missing part numbers)

One other thing now that you mention polarity, the original DIY schematic has signal input caps in the same direction. Typo?
hfHq1DJ.jpg


Revised caps (C7,C2) :
Akao1gF.jpg




 
I usually start with 25 thou tracks for signals and only reduce this if necessary. I find I often have to drop to 20 thou, sometimes 15 but rarely to 12 or 10 thou.

Cheers

Ian
 
> the original DIY schematic has signal input caps in the same direction. Typo?

They have +5V on one side and presumably zero V on the other side. Same way.
 
20 mil signal,  50 mil power,  larger as needed for ground,  high current,  etc.

Thin traces (8 mil) are not typically needed for audio circuits and often detrimental imo. 
 
They have +5V on one side and presumably zero V on the other side. Same way.
Ahh, makes sense. Thank you PRR!

20 mil signal,  50 mil power,  larger as needed for ground,  high current,  etc.
----
I usually start with 25 thou tracks

Thanks 12ax, and RR this is good to know.  .6096 (24mil) was my goldilocks zone.
 
I've shared this anecdote before but what else is (not) new? Back in the 80s I was paid to kick the tires on a power amp company that an investor was looking at investing in, or buying.

It was a small company with one man engineering department... He showed us the power amp he was proud of, and I noticed some scary skinny traces. In design it is generally better to err in favor of wider than needed, than thinner than needed, so I asked for his rationale.

His answer was brilliant, if not total BS... ::)  He said he made the traces skinny on purpose to keep his PCB house honest.  :eek:

I advised the investor to pass...  8)

I don't think I've heard of that amp company since, or before for that matter. 

JR 
 
abbey road d enfer said:
Should I mention that, the smaller the trace, the lower the capacitance, thus less crosstalk in stereo circuits, less unwanted coupling between signals, better phase margin...

True.  The flip side is you get increased  inductance,  increased resistance,  and potential delamination issues during rework / repair.  So a bit of a trade off depending on the requirements.
 
The smaller the trace, the lower the capacitance, thus less crosstalk in stereo circuits, less unwanted coupling between signals, better phase margin...
--------------
The flip side is you get increased  inductance,  increased resistance,  and potential delamination issues
This console has been one big tightrope walk for me. Not the best choice for a first serious project.
--------------
He said he made the traces skinny on purpose to keep his PCB house honest.

mixer has been one big tightrope walk. Not at all the best idea for a first serious project that requires layouts.

Only reason I went with 1.27mm for signal in the channels was because that's what 1604 cards had, and in my early ignorance, I thought it contributed to the sound.  (fat traces = fat mixes)  ::) ::)

Edit: Forgot to mention you guys have been my balancing pole, to be sure!
zWAv1aO.jpg
 
Back
Top