0.156" Pitch Header

[NivagSwerdna]

I'm trying to create a footprint for a PCB that will connect to a 12x2 female 0.156" edge connector...

... got me thinking...

Why 0.156"?  Where did that come from?

[Benta]

Good question

It was very common for single board computers in the 1970s/80s, but whoever defined it? No idea.

[Ian.M]

5/32" rounded to the nearest mil.  On a 50 position connector the difference between 5/32" and 0.156" pitch at the last position is less than 8% of the pitch.

Why 5/32" ? Well that was probably down to constraints like track spacing and required track width for whoever had the original application, then everyone else jumped on the band wagon once a connector was available.

Edit for clarification: 'mil' is very much a USAism for what the rest of the world used to call a 'thou', namely a thousandth of an inch.

[coppice]

Good question

It was very common for single board computers in the 1970s/80s, but whoever defined it? No idea.

In the early days of PCB edge connectors they all seemed to be 0.156" pitch.

[Pawelr98]

I happen to have a soviet VFD display board that uses 2mm/4mm pitch.
Made in 1976.

0.156" is roughly 4mm.

USSR commonly used 2.5mm pitch instead of 2.54mm because they didn't want to use inches.
So with the 0.156" they decided to round it off to 4mm.

Must have been some common standard back then, if the USSR used similar dimensions.

[tkamiya]

In Japan, standard spacing for analog board was 4mm.  In earlier days, there were ICs made with 2.5mm but it didn't last long.  Eventually, market demand dictated we'd start using 2.54mm US standard.

[Berni]

Sounds just as infuriating as the zoo of SMD chip pin pitches these days:
- 1.5 mm
- 1.27 mm SOIC okay so half the DIP makes sense
- 1.0 mm
- 0.8 mm
- 0.75 mm
- 0.7 mm
- 0.65 mm
- 0.6 mm
- 0.5 mm
- 0.4 mm

I probably forgot some but everyone seams to just pick some random pin pitch for the small SMD chips or connectors.
Then you still have some connectors that use 2.5mm and 5mm (Like JST for example)

[bsdphk]

I had the very same problem recently, creating an extender-board for a Documation M200 punched card reader.

The relevant mil.std is 21097 which has all the dimensions for the connector you slot the PCB into, but surprisingly nothing even remotely like a PCB-footprint.  In particular you will want to look at the "M21079" documents.

The only place I found guidance on the PCB footprint was in old versions of EDAC's 307/357 series datasheet.

A really good "app" to supplement KiCad would be a edge-connector footprint generator, either as stand alone (python?) code, or as a webapp.

[NivagSwerdna]

The relevant mil.std is 21097 which has all the dimensions for the connector you slot the PCB into, but surprisingly nothing even remotely like a PCB-footprint.  In particular you will want to look at the "M21079" documents.

Funny you mention that because the MIL Spec was one of my google-ings...
It contains a table which defines 0.050, 0.075, .100, .125, .156, .200 being contact spacings in inches (with approx metric equivalents) and that the error should be +/- 0.005 non-accumulating.
So I think that definitely defines it as exactly 0.156" centered spacing... I'm now looking at IEEE488 plugs... what a minefield.

[tkamiya]

It's really annoying when I guess wrong and things ALMOST fit.....

[bsdphk]

that sounds like a CBM project to me :-)

For the IEEE end, buy a used cable on ebay, and chop one end of.