Dave,
I worked for a division of Cutler-Hammer when Eaton bought them out in the 1970's.? Eventually, we bought ourselves out of Eaton and became employee owned. Later mergers left no trace of the original company.
toggle quoted message
Show quoted text
On 3/7/2019 10:52 AM, Dave wrote: Hi Harvey #2,
?Thanks for posting those links. I just downloaded them and will start studying them later today. My boards will be simple and home brew but I like to do things correct too. So that helps. And who knows, maybe I will get more involved than I think and actually need a board house to make my boards. My? dad worked for Unit Drop Forge which I believe Eaton bought them out. But they were a large forging plant and made crankshafts and other things for Caterpillar.
On 3/7/2019 11:09 AM, Harvey Altstadter wrote:
Dave,
You have asked a lot of thoughtful questions about layout, pad and trace sizes, board margins, etc., and received a lot of good advice. I thought I would point you at some reference material on standards that various organizations use that is available on the Internet.
1. AN3962 from Freescale (now NXP) has a good summary of information about pads, holes, trace sizes, with a rule of thumb for current, spacings for different voltage conditions, and appears to be based upon IPC standards. Find it here:
2. Colonial DFM Guidelines has more information on board mechanics. It is useful if you are going to have your boards manufactured by a board house. Find it here:
3. EATON Corp, Cutler-Hammer Div Printed Circuit Design Guidelines. A very comprehensive design guide that includes information about about device packages, dimensions and board layout. Find it here:
The documents seem to follow IPC guidelines, and are easier to get than the actual IPC standards which are very expensive. There is a high degree of agreement between them, and with much of the information you have been given here. Many more documents are available under a Google search for PCB Layout Guidelines. For specialized types of boards, such as RF and high frequency work, there are documents specific to those areas.
Good hunting
(The other) Harvey
On 3/7/2019 9:04 AM, Harvey White wrote:
On Thu, 7 Mar 2019 09:28:13 -0600, you wrote:
Harvey, thanks again for the detailed post. I also found a trace width
calculator online that helps size for amps. Another question seeing I
have never done this before. I have various tools like a shear, tin
snips, and scroll saw to cut my boards. But what is the best way? I bought the *large* harbor freight shear, 300 some pounds, because I
have a bit of a metal shop.? I'd suggest a smaller one for sheer size
and portability.? Most people deal with pcb material that's 12 inches
wide or less.
And
what is the minimum boarder I need for best or design rules? I have 12
little circuits that fit on a 2x6 copper clad board but they looked
crowed so I changed it to 10. The gap between the long side center is
3/16". If I cut or sheer accurately is that enough meat left on the sides?
In a sense, you don't care how much distance is left between the board
and board edge, it's nominally no less than two trace spacings. The
limit is how close you can get to the trace when cutting.? For a
shear, I'd say that you'd want about 1/8 inch border minimum around
the board, which says 1/4 inch or so between boards.? Note that you
can always do a rough cut with almost anything and then trim with the
shear.? That's final sheared dimensions, shearing when using panelized
patterns, etc.
Scroll saws will dull almost immediately.
Tile saws with diamond blades may not, and who cares if the board is
wet?? Never tried it myself, though.? Be very cautious with the
fiberglass dust if cutting with a saw.
The main problem with the shear is setting it up properly (it needs
different settings for thicker PC board material), and shearing very
narrow boards along the length.? Board material will tend to want to
pop up and wedge between the blade and table.? This imposes some
practical limits.? (as well as needs hold down clamps of some sort for
those rare boards).
Copper clad, when sheared, may have a tendency to delaminate depending
on the shear settings, so we get a second border limit as well as a
need for an unused border around the whole board pattern (regardless
of what's on it).? I typically wanted about 3/8 to 1/2 an inch around
the board pattern that was unused.? That allowed space to tape the
transfer paper down (one edge only), allowed any lifting effects to be
on a part that didn't have traces, gave me space for holes to hang the
board in the etchant tank, and enough space that trimming the board
with the shear worked well.? Trimming the board to the arbitrary two
trace spacing was fine if the shear would do it.
You'll note that the material sheared off fares worse than the
material left behind, but that will vary with setups.
Harvey
Thanks
On 3/6/2019 3:49 PM, Harvey White wrote:
On Wed, 6 Mar 2019 12:33:13 -0600, you wrote:
Thanks for the info Harvey. I have a new question. What is the closest
pads should be to each other? And if you are making a bunch of small
boards and transferring them to copper for etching what is the minimal
space between designs? In other words, what should the remaining border
width be? Seems I may have placed my circuits too close to the center
and have to shrink the width a little. There's two limits, design and process.
Process limits simply say that you can't physically make the board
like that.? Either too close to the edge, or holes too small, or
pads/traces too small.
Those limits vary with respect to the process, but typically 0.030 or
0.060 are decent limits (roughly 1/16 or 1/32) for pad to pad, and
spacing betwen tracks, but that's process....
Now, voltage wise, that determines the minimum track spacing as well.
I'm not sure of the limits here, you'd have to look that up. Ditto
with track current carrying capacity for width (varies also with
respect to the copper foil thickness).
You also want to allow for heat dissipation for things like resistors.
Typical power resistors are often at the end of longer leads and
spaced off the boards.? Seen scorched boards because of resistors. Now
on the other hand, those resistor long leads will fail any sort of
vibration test.? (military stuff used clamps for bulky parts like
that).
So for home analog stuff, not necessarily digital, you want a small
enough trace that you can put one between the pads of a DIP (more
requires finer stuff than can be happily done at home).
WIthout an autorouter (which is of limited use for home built boards),
I tend to use the airwires (unrouted traces) to determine which parts
need to be near which parts for shortest wire runs, then try to
manually route the shortest wires.? It's a learning process...
Once you go to surface mount parts, the design rules start to change,
but I'm not sure where you are on this design.
Harvey
|
