Model M Cables, 4 leads vs. 6 leads

[Pretendo]

...


That's about the extent of it.  You can see that the above SDL connector has only 4 leads, while the lower one has 6 leads.  Green and yellow are added at the edges.  The cable with 6 leads is a longer cable, about 10 feet.  It caused certain PS2 controller boards to completely lock up on me (including mouse functionality), until I yanked out the two pins underneath the wires to sever the connection.  I've seen several posts about Model M boards behaving like this on Geekhack forums.

Are these black sheep wires documented?  Perhaps they're meant for terminal boards??

[theferenc]

The only 6 lead SDL cable I have is the one the belongs to the M5, which has 2 PS/2 connections at the end. In that case, I'm assuming that the ground and +5v lines are shared, and those 2 extra lines are clock and data for the trackball.

Could be wrong though.

Edit: Oh, and I checked 11 different cables, for what it's worth.

[Pretendo]

The only 6 lead SDL cable I have is the one the belongs to the M5, which has 2 PS/2 connections at the end. In that case, I'm assuming that the ground and +5v lines are shared, and those 2 extra lines are clock and data for the trackball.

Could be wrong though.

Edit: Oh, and I checked 11 different cables, for what it's worth.

This wire has only one PS2 connection at the end. The only oddity about it (besides the extra leads) is its 10' length.  My father, who works at an IBM plant, was allowed to take this cable home for me.  They claimed it was broken.

  Because it came from an actual IBM plant, I'm thinking that it may have belonged to some odd form of the M. It could also just be an earlier cable design.

The wire worked fine for me with a USB/PS2 adapter, but locked up when I plugged it into a real controller board.  Some of the LEDs would stay on, but it wouldn't work after the BIOS.  It was only after I acquired another wire and specifically looked for differences that I realized what was different.  I disconnected the two odd leads and it's worked fine for me ever since.

I'm just suggesting that if people have a keyboard that's acting like this, they check out their SDL connectors.  Since it seems like these leads never had a use, it would pay to get rid of the extra leads to restore functionality to the wire.

[Pretendo]

Absolutely correct...that's what the extra 2 get used for. Makes me wonder what would be found if the OP cut apart the PS/2 connector on the cable in question...I wonder if those wires are just N/C or actually attach to something. Could always poke at it with a meter.

See my post above yours.  The extra wires caused problems.  In my pic you can actually see where I ripped out the pins underneath them to sever the connection.

[Pretendo]

Yeah, I see that...interesting. If they happened to be connected to the data and clock pins in addition to the normal lines, in theory, such a cable could let an M5-1 or M5-2 properly operate on a system with a combined mouse+keyboard PS/2 port.

A non-trackball Model M might do weirder things though. I don't have a controller handy to poke at, but I would think the extra pins either go NC or GND - but if they are indeed duplicate clock and data, that could cause seriously erratic behaviour to have two clock and data lines...(but then, how are two duplicate wires a problem? It is possible the wires being active on one end were creating an interference problem)

Did they ever make such a dual ps2 port on the hardware end?  I had a wild guess that it might be a terminal cable, or possibly a POS (point of sale) wire, with the extra leads allowing for a credit card swiper to transfer data.

But that wouldn't explain why it would cause a freeze up with a regular M board, unless, like you said, the M is outputting some extra data on these lines.  I'm guessing the ps2/USB adapter worked because it's not implemented on the hardware interrupt level; it's simply a USB dongle which probably ignores the data from those lines anyway.  Actual PS2 controller boards may not know how to deal with these lines since they were never really used, and some (like my old Sony) may lock up upon receiving any data through them.  A clock signal would certainly cause problems!

At any rate, I don't think they were ever officially used for any commercial application.  If somebody knows otherwise, correct me!

[False_Dmitry_II]

I think certain laptops had a port like that. They are all obsolete though.

[Pretendo]

Yeah, whatever the function was is probably obsolete.  I'm just worried that somebody will come across one of these and fry their keyboard, or possibly even their PS2 controller!  They seem to be rare, which is a good thing in that most won't have to worry about it, but also a bad thing because nobody's looking for the problem...

[rootwyrm]

Looks like it's not a PC Model M there; looks like that cable's from a 3151 "general purpose" terminal. Technically, PS/2 specifies 6 conductors - two are reserved. The 3151 doubles down on some keys - they have multiple possible modes and functions associated to specific standard scancodes based on key combinations. (Confusing, no?) Thing is, the 3151's an RJ45 connector (basically SDL-RJ45, not so easy to tell those two apart.)
I'd be a lot more interested to have the keyboard that cable came with apart, to see what controller's in there. Weird.

[Pretendo]

Looks like it's not a PC Model M there; looks like that cable's from a 3151 "general purpose" terminal. Technically, PS/2 specifies 6 conductors - two are reserved. The 3151 doubles down on some keys - they have multiple possible modes and functions associated to specific standard scancodes based on key combinations. (Confusing, no?) Thing is, the 3151's an RJ45 connector (basically SDL-RJ45, not so easy to tell those two apart.)
I'd be a lot more interested to have the keyboard that cable came with apart, to see what controller's in there. Weird.

The wire is most certainly an SDL connector, and not an RJ45.  Fits perfectly into the keyboard and has tabs at the sides instead of the top.  The tabs look a little bent out because, well, they ARE a little bent... I didn't receive this thing in the best of condition .

There's another small difference I noticed.  The 6 lead cable has two markings that the 4 lead cable does not; UL component certification (in the form of a bacwards 'R' attached to a 'U') and CSA certification (a 'C' enclosing the SA.)  These are located opposite the pin side, underneath where it says 'AMP.'  Picture of this attached below:


6 lead left, 4 lead right.

EDIT: Added picture.

[Pretendo]

BTW all mine are 4pin.

That seems to be the consensus; since it came from an actual IBM plant I'm leaning more and more towards it being a REALLY early version of the wire or some weird in-house project.  It's not of any real use without modding it down though, unless you like frozen controller boards.