Certain MiniDin 6 Pinouts Cause Problems on KVM Switches

We have a lot of server room computers hooked up to Belkin 16 port switches.
What we noticed is that maybe 10% of the 6 foot Belkin "all in one" cables
we buy end up causing sporadic problems, such as the keyboard disappears for
a given server.

We have a continuity tester, and I noticed an interesting result on these
cables that maybe some engineer in the group could give me more information
about.

First, according to the published standards I could find, the correct wiring
of the 6 pin mouse and keyboard PS/2 cables on the minidin 6 type connector
are as follows (pins one through six in order):

1=data
2=not connected
3=ground
4=+5V
5=clock
6=not connected

With all of the longer Belkin All In One cables (15 to 25ft), the 6 pins on
the minidin 6 are all connected end to end, and there is no ground
connection on pin 3. We have no problems with those cables.

With the cables that are causing us problems, we measured this result on the
pins using our continuity tester:

1=connected
2=not connected
3=data is not connected, but ground is connected
4=connected
5=connected
6=not connected

The cables that cause us problems seem to follow the standard, whereas the
cables that don't cause problems ignore the standard and just connect all
pins.

Is it possible that the problem here is the ground connection on pin 3?
Maybe the wiring never gets to a true ground, and there is some kind of
ground loop?

If this is a possible cause for the observed symptoms, how could I confirm
it?

--
Will

On Fri, 4 Aug 2006 21:54:41 -0700, "Will" <westes->
wrote:

>We have a lot of server room computers hooked up to Belkin 16 port switches.
>What we noticed is that maybe 10% of the 6 foot Belkin "all in one" cables
>we buy end up causing sporadic problems, such as the keyboard disappears for
>a given server.
>
>We have a continuity tester, and I noticed an interesting result on these
>cables that maybe some engineer in the group could give me more information
>about.
>
>First, according to the published standards I could find, the correct wiring
>of the 6 pin mouse and keyboard PS/2 cables on the minidin 6 type connector
>are as follows (pins one through six in order):
>
>1=data
>2=not connected
>3=ground
>4=+5V
>5=clock
>6=not connected
>
>With all of the longer Belkin All In One cables (15 to 25ft), the 6 pins on
>the minidin 6 are all connected end to end, and there is no ground
>connection on pin 3. We have no problems with those cables.
>
>With the cables that are causing us problems, we measured this result on the
>pins using our continuity tester:
>
>1=connected
>2=not connected
>3=data is not connected, but ground is connected
>4=connected
>5=connected
>6=not connected
>
>The cables that cause us problems seem to follow the standard, whereas the
>cables that don't cause problems ignore the standard and just connect all
>pins.
>
>Is it possible that the problem here is the ground connection on pin 3?
>Maybe the wiring never gets to a true ground, and there is some kind of
>ground loop?
>
>If this is a possible cause for the observed symptoms, how could I confirm
>it?

By "ground" you mean the plug shell? Real IBM keyboard cables use the
shell connection for the ground braid and use the pin 3 for an
isolated DC signal ground. Your problem could be caused because these
get mixed up in some cables.