header pin outs

[Simon]

I notice that both male and female pin headers have the same pin numbering when viewed from the top. This means that when you rotate one 180 degrees to plug them together if it is a 2 row header pins 1 and 2 mate. Pins 3 and 4 mate and so on. I take it this is normal? Presumably there has never been any standardisation.

[DerekG]

I take it this is normal?

I presume you are talking about IDC connectors as used in older desktop computers?

Originally this was a big problem as you could connect HDD & CD burners etc "backwards".

Eventually "most" manufacturers learnt the hard way (product returned due to be burnt out etc) & they moved over to a key-way moulding (so the connector could only be pressed in the correct way around).

[Simon]

I'm talking any old header. Yes there is no difficulty in plugging them in upside down but I am referring to the lack of standardisation in that pins will never mate with a pin of the same number.

[Wendy_Preston]

I can only assume that this has effectively occurred because these no-frills pin headers (like our M20-9981046 for instance) have nowhere on them for polarisation. Also they are designed to be cut down, so no way of marking a number one contact on the housing.

What you gain in flexibility you lose in design-for-assembly, I guess. These pin headers - in this same design - have been around far far longer than I have been at Harwin (23 yrs and counting) - and we've never promoted a pin-numbering system for these connectors because of this.

[Simon]

Well it is just that even with no polarisation after single rows you have to be careful with your design. If all female headers had the numbering reversed to the male headers it would word if they are mated the right way around. With no polarisation features each PCB designer is free to do as the please and I will in future be differentiating male and female so that when mated the numbers coincide and it makes designing the two halves easier. What the manufacturer chooses to call each pin is of no concern to me as they are not polarised.

[donotdespisethesnake]

There is no standard, since there is no standard way to use these connectors.

It's down to topology, and that depends on your design. There are several ways to do board to board connectors, some of these involve "flipping" the connector on one end, some don't. If you use a ribbon cable, the wiring is probably 1-1, if you use a board stacker, then you probably need to flip - unless you have a bottom entry connector.

Relying on the manufacturer to number the pins "correctly" is not a good option. But don't worry, even professionals get it wrong . We just had a little interface board come back and ALL the connectors are flipped! What a mess of wiring that created

[T3sl4co1l]

Shrouded headers usually have a consistent pinning convention; check the datasheet.  (Sometimes the polarity mark can be hard to see, or is buried in a tiny note.)

On the PCB footprint, the convention is usually with the key down, pins in horizontal row, pin 1 bottom left.

Breakaway headers are completely arbitrary and it's up to you to check that the pins are being used consistently.

Tim

[Simon]

Well i am working on my own system now with opposite pin rows for male/female so that the numbers will match.

[jpanhalt]

For breakaway headers, I add an index pin.  They come as a strip.   Cut one off and insert it in a female contact. Then remove or cut the mating pin.   Of course, don't use the center pin, if there is one.

[Doctorandus_P]

I find it hard to imagine that there are connectors in which the Male pins are not supposed to match with the corresponding Female pin numbers.

What software and connector type do you have in mind?

There can be a difference between the representation of how a symbol for a connector looks on a schematic and the location of the pins on the Footprint of the PCB.
Are you sure you are not mistakingly looking at mirrored representation of the connectors?

I once had to cut the wires of an 14 pin IDC cable with a knife and swap the even and odd wires because they layout of an LCD connector was designed to be inserted in the front of the LCD, but there is of course a frontpanel in front of the LCD, so the IDC connector was soldered into the back of the LCD.
Shit like that happens all the time. It seems to be an intrinsic part of life as we now it.

There are some rare hermafrotitic connectors, where any can be mated with any. Those were used in old IBM computers (Was it the 360?) and I think also with GPIB, but I'm not sure.

Lately USB-C is gaining popularity. Not hermafroditic, but plugs can be rotated 180 degrees.
The power pins are symmetrical over the design, also because multiple power pins are good for reliability and current handling capability. Current computers have enough smarts to figure out how the signal wires are connected.

But for ordinary connectors, if M & F pin numbers don't match then probably (& in this order):
1). You made a mistake.
2). There is a mistake in your libary symbols.
3). Something weird and/or unsual is going on, and it is intentional (for a good reason?).

[bson]

If all female headers had the numbering reversed to the male headers it would word if they are mated the right way around.

Well, FWIW they do match up in KiCad v5.  I imagine in any other layout tool as well if you judiciously use the right footprint for the gender.

 

[free_electron]

and then you stack board on top of each other ( male bottom mounted , female top mounted) and it still screws up.

i use the same pin order as if it would be a box header. pin 1 is clearly indicated on the silkscreen
it is up to the board designer to flip them correctly.