Circuit Board Design Tips

For The Popcorn

Well-Known Member
Most Helpful Member
The picture shows the vending machine control board I'm just finishing up. You may notice an abundance of 0.1" pitch KK type headers i have a couple tips to help get 0.1" pitch headers nicely installed.



The first tip is to use "lock" pcb footprints developed by Spark Fun. Alternate holes are positioned slightly above or below the baseline. This offset ensures the connector is perpendicular to the board. Details and links to the Spark Fun are at this link.

Jon's Imaginarium Lock Footprints

The second tip is to space header connectors on a grid according to the pitch of the headers, leaving an empty position between connectors. This leave room for the mating connectors. In the context of this discussion, it allows you to bridge all of the connectors with a female header strip, aligning the connectors. Add a couple of spring clamps and the connectors are held in position while you get them soldered in place.


 
Pretty useful, thanks for sharing this
 
That might solve one problem, but it creates others. Forcing a pin to be pushed against the wall of a plated-through hole during soldering is not a god idea.

First, it creates a void in the solder where the only thing assuring contact between the pin and copper is friction. This is fine for press-fit connectors designed for this, but for "normal" parts it is not a "gas-tight" connection.

Second, it makes the connector very difficult to remove when using a vacuum desoldering station.

A better way to assure that the pins are nicely aligned across the row is to use your second tip: Using the strip socket as a carrier, load all of the connectors into the socket, then install the entire row on the board as a single component, solder everything, and remove the strip socket.

ak
 
The only "problem" I have observed using lock footprints is finding out some months after soldering a board together was when I was unpluging a connector and the header came off the circuit board – I had missed soldering it, and it was working well despite that omission. This "problem" can be handy for making temporary connections like an ICSP header. Insert a header during dev, then remove it when you're finished.

I have to question your "gas tight" point as well. These headers have square pins, just like wire-wrap headers and sockets. The very slight interference fit created by a lock footprint causes the header pins to slightly score into the hole plating.... if you believe claims about wire-wrapping, the sharp corners of the pins bite into the wire, creating a gas tight connection. Hmmm. A lot of complex (i.e., hundreds if not thousands of connections) of gear with wire-wrap connections is still going strong, so maybe they knew what they were talking about.

The Sparkfun article documents the efforts made to determine and test the optimal spacing. It was pretty well researched, and Sparkfun has been using it for years.

I'm not here to debate the idea. I have found it useful without any problems. Please feel free to use it or not as you see fit.

Of course, one has to wonder how many headers you unsolder that this would be a major concern......
 
JON,
Generally, I agree with your points. Today, most stuff is SMD -- at least what boards I have done are. Solder w/ flux will get into very small spaces. Note that voids are not a problem when sweating copper pipes, if done correctly.

More relevant is that solderable breadboards, which I use a lot of, don''t have staggered holes. In that case, if it's a header, I put a socket or jumper on the pins, then use a paper tape (aka masking) to hold it to the board. Sockets don't need anything. I far prefer 3M blue tapes as they don't leave goo like the tan types do. Then I solder one pin. Turn the board upright and check whether the thing is vertical and against the board. If not, it's easy to re-apply the iron while holding the thing with fingers on your other hand. Once that is done, I solder another pin, check, then solder the rest. The jumper/socket on headers helps keep the pins nicely aligned and protects your fingers from getting burned while doing the first two pins. For right-angle headers, adding a socket also assures it will fit after soldering.

Attached is an image of a recent board. (The extension is to accomodate a 9-pin FFC cable from a keyboard.)

Edit: typo
 

Attachments

  • Final Top All 110924.jpg
    366.2 KB · Views: 23
Last edited:
Jon;
I particularly like your idea of using staggered pin connectors as a “programming” connector which you can easily remove after the fact.
 
Jon;
I particularly like your idea of using staggered pin connectors as a “programming” connector which you can easily remove after the fact.
Well, maybe so, but the most useful thing to me for my programming connector is some key, so I can't plug it in backwards. You never know what your mental state will be when you finally have a solution to try. Thus, for ICSP (Microchip), which requires 5 pins, I use a 6-pin connector and blank the 6th pin, so it can't be plugged in backwards. The upper left header in the board I show has that.
 
I use a 6 pin header for Microchip's ICSP connection, sometimes including an extra port pin on pin 6 – you never know when you'll want to connect something else

I have had a few "Oh crap" moments when I've gotten that connector backwards... and remarkably, no damage has occurred to the micro or other parts of the board. I don't know if Microchip's engineers were ingenious or just lucky, but their layout has saved me a couple times.
 
I simply solder one pin of the connector (sloppily), then I pick up the board with my non-dominant hand and gently pinch the through-hole part while I touch the single solder joint with the iron. The part falls into perfect position and I can solder the rest of the joints without problem.
 

Likewise - simple and easy.
 
burned finger or messy result?
Comical
I didn't know I needed to post kindergarten level common sense, but here goes.
Pro-Tip #1: Don't pinch the pin you are reheating to set the component in its final place.

Pro-Tip #2: it takes effort do download custom footprints or adjust zig-zaggy footprints that put mechanical stress on the pins or component. Nothing causes micro racing in plastics like stress - especially after solvent or solvent vapor hits the stressed area of a plastic part. Be careful with anything beyond a hobby project using the stressed-footprint method. It also takes effort to clamp each component - the lazy method is in post #11

Pro-Tip #3: I've never seen a cleaner/neater method than as described in post 11 so, I have no confidence in most anything the guy watching the popcorn is saying if he things it's a sloppy method.
 
Last edited:
Cookies are required to use this site. You must accept them to continue using the site. Learn more…