Homebrew_PCBs

Been checking the net and I've found a few trace width calculators.
I'm having trouble believing what I'm seeing though! I have two 7
segment displays (man71A). Each one takes about 80mA to operate. I
had just guessed and used a 20 thou' trace (0.02 inch). According to
the calculators I've found though, 1oz copper carrying 160mA with a
max temp rise of 5C only needs to be 0.0014 thou' (1.4 mills). Is
this correct? Are the trace sizes I see on boards as big as they are
simple due to limitations in the pcb making process and not for
carrying current?

What trace sizes do you generally use for TTL circuits?

And actually, while I'm at it - what pad sizes do you like to use?
PCB123's default pad size is 0.05 inch. Seems pretty small to me
but, I'm just a noob :D Any trouble soldering on pads that small?

In a message dated 12/31/2002 10:03:08 PM Central Standard Time,
ghidera2000@... writes:

> Are the trace sizes I see on boards as big as they are simple due to
> limitations in the pcb making process and not for
> carrying current?<<

Yes! I think I heard, once, that 0.01" wide is about the practical limit for
small. I have SEEN formulae for calculating the minimum copper-width per
amp, but it was quickly clear that I'd never be working with currents
approaching those where calculating copper-width was necessary, so I did not
memorize, so, have forgotten.

>
> What trace sizes do you generally use for TTL circuits?<<

Surely, even 0.01" wide traces would be sufficient for even maximum fan-out
from TTL. However, not only for current-capacity, but to reduce EMI effects,
wider traces than that, say, up to 0.1", might be adviseable for Vcc traces
(+5VDC) for a dozen TTL IC's, particular "board ground-buss traces". But
"one-each" NON-solder-plated (i.e., "copper-only") traces? I'd want at
least, oh, 0.025" wide, as the integrity of the conductor would be tenuous,
if narrowerer, for home-brew!

> And actually, while I'm at it - what pad sizes do you like to use?
> PCB123's default pad size is 0.05 inch. Seems pretty small to me but, I'm
> just a noob :D <<

For my own "one-each" NON-PTH home-brew boards, I'd want a pad at least .065:
dia., as it is usually difficult to get registration just-so, so that the
hole doesn't sever the circumference of the pad when registration is clumsy
(been there, done that!). However, the couple of times in the past when I
have had professionally-made boards made, and I have used Bishop Graphics and
all that, and 2X artwork, and all is on a "guaranteed grid" of 0.025", then
0.05" dia. pads would be fine for something tiny like IC pins or 1/4 W. R's,
etc. There is nothing finer than the "wicking effect" of a solder-plated PTH
board! The pad COULD even be as small as, say, 0.01" larger dia. than the
hole-dia.! BUT, when we are talking about NON-PTH, you need SOME copper
there for the solder to adhere to, to make a "metallurgical bond" which will
conduct electricity! (Yes, surprise! That's what a PCB-trace is FOR!).

Took me some decades of tinkering and breadboarding to learn it, but for more
than two or three IC's and the usual amount of discretes therewith, a
"one-each" PCB is just too much hassle! Thus, the attraction of wire-wrap!
I agree, however, for just one to three IC's, often even a "one-each" PCB is
nicer. Or, if you have to make ten or more of "same board", then PTH
professional is the way to go. Whew, if I were 17, again, and knew all I
know now, of such home-brew hassle, I could be a whiz with a much nicer
personality!

Jan Rowland, old Troll




Any trouble soldering on pads that small?>

>

[Non-text portions of this message have been removed]

Been through this years ago. You will find everyone has trace widths and
pads that they like to use. I am sure the calculations are correct but I
wouldn't use a 1mil trace in homebrew. I made a test board years ago to see
how many amps it takes for a trace to fail. Made each trace from 6mil to
100mil. That's when I found out that the inkjet printer doesn't recreate
trace widths accurately. Only laser makes the traces right. Hooked up the
wires got a 12v motorcycle battery a good Fluke meter with auto peak and
started testing. If my memory is correct the .006 trace failed at 1A. The
.010 3Amps. The .020 5Amps and so on. I found that solder coating not only
protects the trace but almost increases by 2/3 the Amps it can handle.
I like overkill on my circuits the biggest trace I can fit that doesn't take
up too much board space. I use the funnel type of design big end at the
power source or buss supply to the smallest trace at the lowest amp part of
the circuit.
I usually don't go below .020 trace anywhere on the board.
It is too hard to make a board with anything smaller and why do it if you
have the real estate use it, less etching.

As for the pads I like to drill 100mil and for power diodes and wire inputs
150mil.

Derek B.
http://groups.yahoo.com/group/Denver_CNC_Club/

>
> Been checking the net and I've found a few trace width calculators.
> I'm having trouble believing what I'm seeing though! I have two 7
> segment displays (man71A). Each one takes about 80mA to operate. I
> had just guessed and used a 20 thou' trace (0.02 inch). According to
> the calculators I've found though, 1oz copper carrying 160mA with a
> max temp rise of 5C only needs to be 0.0014 thou' (1.4 mills). Is
> this correct? Are the trace sizes I see on boards as big as they are
> simple due to limitations in the pcb making process and not for
> carrying current?
>
> What trace sizes do you generally use for TTL circuits?
>
> And actually, while I'm at it - what pad sizes do you like to use?
> PCB123's default pad size is 0.05 inch. Seems pretty small to me
> but, I'm just a noob :D Any trouble soldering on pads that small?
>

>

> Took me some decades of tinkering and breadboarding to learn it, but
> for more than two or three IC's and the usual amount of discretes
> therewith, a "one-each" PCB is just too much hassle! Thus, the
> attraction of wire-wrap! I agree, however, for just one to three
> IC's, often even a "one-each" PCB is nicer. Or, if you have to make
> ten or more of "same board", then PTH professional is the way to go.
> Whew, if I were 17, again, and knew all I know now, of such home-brew
> hassle, I could be a whiz with a much nicer personality!

If I could find wire-wrap goods (wire and wrapper) in Brazil, I'd be
more than happy :o)))

In a message dated 12/31/2002 11:53:10 PM Central Standard Time,
hightechsystems@... writes:

> As for the pads I like to drill 100mil and for power diodes and wire inputs
> 150mil.
>

Derek: This is kinda BIG for IC's, seems to me! Wouldn't the pad have to be
considerably SMALLER than 100 mill for pins on 100 mil centers????


[Non-text portions of this message have been removed]

Hi Jan,
The pads I am talking about here is the free pads, resistor
small diodes, etc. The IC's are 50x100mil ovals normal size.
These are already in the parts inventory and I didn't have to make a new
part. You can make the pads any size you want even odd sizes 83mil.
I use PC Logic mainly because it has some very nice features that others
don't have and I am use to it.
Like easy to make custom parts while you are drawing the PCB. Step and
Repeat when printing, etc.

Derek

> -----Original Message-----
> From: JanRwl@... [mailto:JanRwl@...]
> Sent: Wednesday, January 01, 2003 8:19 PM
> To: Homebrew_PCBs@yahoogroups.com
> Subject: Re: [Homebrew_PCBs] Choosing a trace width
>
>
> In a message dated 12/31/2002 11:53:10 PM Central Standard Time,
> hightechsystems@... writes:
>
>
> > As for the pads I like to drill 100mil and for power diodes and
> wire inputs
> > 150mil.
> >
>
> Derek: This is kinda BIG for IC's, seems to me! Wouldn't the
> pad have to be
> considerably SMALLER than 100 mill for pins on 100 mil centers????
>
>
> [Non-text portions of this message have been removed]
>

>

In a message dated 1/1/2003 10:19:07 PM Central Standard Time,
hightechsystems@... writes:

> The pads I am talking about here is the free pads, resistor
> small diodes, etc. The IC's are 50x100mil ovals normal size.
>

Clear, now.


[Non-text portions of this message have been removed]