I’m laying out a PCB using a 555 timer. I was able to figure out the “puzzle” and get all but one
connection on just one side of the board. For configuration in “astable mode”, pins 4 and 8 need
to be tied together, as well as pins 2 and 6. My plan was to to route a track from pin 4
to pin 8 on the “front” side. Since I can’t route pin 2 to pin 6 on the “front” also (since it would
cross the track from pin 4 to pin 8), I thought I could just route a single track on the “back” side
from pin 2 to pin 6. I thought the correct way to connect the “back” side track to the pins on the
“front” side was by using a “via”. So far so good?
I ran DRC right before adding the 2 vias (to pin 2 and pin 6) and got no errors or warnings.
Then I added the vias and re-ran DRC which reported Error: Drilled holes too close together…
What am I doing wrong? I thought adding the via would simply result in the instruction for the
inside of the pin holes to be copper coated so they “connect” the tracks from “front” to “back”.
I don’t understand the error since there’s only one hole (not 2 holes next to each other).
If I was trying to make this at home, I would just add two extra pads adjacent to pins 2 and 6, and run
a wire between them. But since I was going to try out one of the Chinese prototype services that seem
to charge the same for 2 sided boards, I thought I would try to add the “crossing” track by adding it
to the “back” and use the vias.
Hi, straubm - thanks for the reply! Are you saying I put the vias on the wrong layer? This is my first
attempt with vias and I figured it was probably “operator error” on my part, but I’m not quite sure
how to get them on the right layer. Since this project will be sent out to fab I want to make sure
my gerber files have all the correct information and I won’t have to do any “rework”, ha, ha!
Hi, thebigg - forgive my ignorance on nomenclature. The “front” side is the side with most of the tracks,
yes? So the components are on the “back” side with the “pins” sticking out on the “front” side that
has the tracks and are soldered? Or maybe I got that backwards. Anyway, if I have a component on
one side, I could solder a wire to pins 2 and 6, but thought I would use the vias so that the board
would look neater, and since the PCB fab places charge about the same for 1 sided or 2 sided.
Am I thinking about this all wrong?
Generally the components are on the front side. Its your choice whether you choose to route the tracks on the front side or the back, or both.
Sorry for the confusing wording: when I said
I meant route a track between the pads on the “other side” of the board.
When 1 layer boards roamed the earth, the back side was the one with all of the tracks. Now with 2 layer boards and SM parts, either side could have more tracks. TH part footprints use through hole pads which act like a via so you can route from a pad from either side. It’s usually when you have to make a track switch sides that you use a via.
Hi, retiredfeline - thanks for the history! Yes, I was thinking like 1 layer board because that’s what I would do if I was making it at home (going to try to use some of that blue photoresist film, etc.). But when I
decided to send it out to a fab house, I was thinking about how I could get a finished product that didn’t
need any “wires” - but could put a few tracks on the “other” side if needed. That’s why I was thinking via.
If I understand you correctly, you are saying that THT footprints use pads which are plated on the inside of the holes? So the “front” and “back” tracks at a particular through hole (like a pin on an IC socket) are already “connected” because the through holes are plated? If that’s the case, then I understand what you are saying.
Just a couple of comments: you do realize that you have placed nearly all your tracks on the same side as your components?
Red is the default colour for tracks on the top layer… same layer as your components. Blue is the default colour for the bottom layer… the underside of the board.
Also, as J1 & J2 are connectors, I’d suggest making their pads larger and the tracks, to which they are attached, wider, so as to increase their mechanical robustness. Few things are more annoying than pads lifting from a board and the tracks cracking after only plugging in the power supply 20 times.
This is a bit of a tangent but the small firm I worked as a trainee decades ago did SM with TH parts on single sided boards. We just bent the legs 90° outwards. Advantage was no board drilling required and rework was easier. Disadvantage was more board space required. This was when the available SM packages were only flatpak.
Note:
This is a Top layer view (Components overlay bottom layer tracks)
Beefed up tracks, especially around terminals (I’ve paid for the copper, so I might as well have it)
Beefed up terminal pads.
Didn’t need top layer tracks or vias.
Hi, jmk - that’s great! Very clever layout! I see that you routed “underneath” some
of the components, like C2. I was over-constraining myself because of a mental block
from the layout of the schematic.
I see that you used “long pads” on the 555, as well as bigger “square pads” on the terminal blocks. I guess the board starts off with copper everywhere, huh? Then the part you don’t want ends up getting etched away - so keep as much as you like?
This is very helpful for a newbie - thanks so much for taking the time!
Glad you had a reprieve from mowing the grass!
Yes, you can keep as much copper as you like. Modern PCBs usually use copper fills, which means that they are completely copper except the clearance area around tracks. Generally, the less copper you need to etch away, the better, as you need less acid and generate less waste. (However that depends on the exact manufacturing process, often you start with very thin copper and then they build it up after etching)
More reprieve from mowing 50mm of rain overnight and another 75mm expected today.
There are pages of articles on the internet on how PCBs are made. I typed “Manufacture of PCBs” into google. It is worth your watching one or two to see the full process.
A few notes on my example:
Thick tracks and bigger pads can be a help if soldering skills are not great.
My board is 26mm X 21mm… usually, the smaller the board, the cheaper to have made.
C2… Although your C1 & C2 are the same value, so probably the same package, as they are TH (through hole) they have leads and leads can be bent, so to save bothering with tracks on the top layer, a wider footprint was used for C2, enabling a track to run between the pads. If using SM (surface mount) you may choose a larger package for C2 to get the same result, or, in either case, make the track narrower.
The two connectors were modified Kicad packages to increase the pad size, so to increase the strength and durability, as these will probably be subject to some use.
I used the DIP8 package with long pads just to show there are different packages in Kicad libraries.
A lot of this design stuff comes down to experience and practice and ALWAYS remember, as I was once told, back in dinosaur days, placement of components is 95% of the job.
Hey, Janek - You’re the bomb! This information is so helpful. I ordered some test boards from PCBWAY and am excited to see how they turn out. They have a lot of videos online that explain their processes. I looked at a couple of youtube videos and other pcb links online as well. I’m a newbie, but with the pointers I picked up here on where to look, I feel like I have a much better grasp on the basics.
