Kitz ADSL Broadband Information
adsl spacer  
Support this site
Home Broadband ISPs Tech Routers Wiki Forum
 
     
   Compare ISP   Rate your ISP
 
Please login or register.

Login with username, password and session length
Advanced search  

News:

Author Topic: A Proprietary, Non-Standard, Interface . . .  (Read 523 times)

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 20286
  • Over the Rainbow
    • The ELRepo Project
A Proprietary, Non-Standard, Interface . . .
« on: December 03, 2017, 07:10:35 PM »

. . . thy name is Cisco.

Currently there is some electronic gubbins at The Cattery. ("Quite normal", I sense one thinks.) When powered up it is totally silent. No noisy fans, etc. It is only the front panel LEDs that show it is active.

Also on the front panel is an 8P8C socket marked "Management RS-232". It was decided to connect to that socket, from a computer, with the idea of watching the boot-up process. A flat, pale blue cable, with an 8P8C (male) plug at one end and a DE-9 (female) plug at the other end, was obtained for 2-99 from a charitable concern. The lead was sealed in a clear plastic bag. The label affixed to the bag stated a Cisco part number and the informative string "Made In China". The DE-9 plug carries the Cicso "bridge" logo and the words "Cisco Systems". The lead was plugged into the respective orifices of the gubbins and the computer. A logging screen session was started and the gubbins was powered on. Absolutely nothing was seen.  ???  The gubbins responded to my commands from the computer keyboard . . . the front panel LEDs gave clear evidence of that fact. The experiment was concluded with that failure.

Some weeks later, yesterday to be precise, that pale blue cable was checked with a DVM and this is what was determined --

8P8C Plug        DE-9 Plug
     1                      8
     2                      6
     3                      2
     4                      5
     5                      5
     6                      3
     7                      4
     8                      7

Various on-line resources confirmed that the above pin-outs are correct for one type of Cisco lead.

Eventually I read the documentation for the gubbins in which the recommended adaptor, for use with a normal Ethernet patch lead, is described. It consists of an 8P8C socket back to back with a DE-9 female plug and follows the TIA-561 specification --

8P8C Socket        DE-9 Plug
     4                        5
     5                        2
     6                        3

Suddenly it became clear. By using that Cisco lead, my keyboard commands were reaching the gubbins but the information that the gubbins was trying to show to me was being short-circuited to earth.  :doh:

An appropriate adaptor (for use with a standard Ethernet lead) was sourced and ordered. Delivery is expected this coming Wednesday.

A proprietary, non-standard, interface . . . thy name is Cisco!  :D
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

WWWombat

  • Kitizen
  • ****
  • Posts: 1663
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #1 on: December 03, 2017, 10:39:26 PM »

Welcome to the odd world of Cisco connections...

The wombattery (?!) (not that battery) is currently home to a whole menagerie of Cisco gubbins, but they are pretty consistent ... the console port is used for management of the hardware, and the usual method of connecting is one of those "light blue" RJ45 - DB9 connectors. For computers without a serial port, you can now get the equivalent "light blue" RJ45 - USB (Host connector) cable.

The hardware tends to have a label for the "Console" coloured in matching light-blue.

Online, you often see these console cables also termed "rollover" cables. Looking at the documents, it seems that a proper rollover cable is actually RJ45-RJ45 that isn't wired as a "normal ethernet patch lead." It is wired, instead, 1->8, 2->7, 3->6, 4->5, 5->4, 6->3, 7->2, 8->1. Neither a standard patch cable nor a crossover cable.

It seems to me that, historically, a connection to Cisco kit involved both a "proper" rollover cable, RJ45-RJ45, and an adaptor in the computer that converted the DB9 port to an RJ45 socket. Nowadays, the light blue cable you first acquired is a common short-cut for this pairing of a proper rollover cable and an old-style RB9-RJ45 adaptor.

This thinking seems confirmed in this Youtube video (around 2:45 onward): https://www.youtube.com/watch?v=oHTQ-OSGrDk

The rollover cable alone, without a DB9 adaptor, has a place too: it is used when you connect the consoles of many pieces of hardware into a "terminal server" like the one mentioned here: http://gregsowell.com/?p=4425
(On that page, you'll also see one of the odd octo-cables that can turn an optional port on a Cisco router into 8 console ports)

Once you've learned all this, while sometimes odd and pinouts aren't often described, it does seem to be applied consistently. If you have a 48-port terminal server with 48 identical ports, then you want every piece of hardware to be wired the same ... you don't want 48 different adapters trailing through the rack!

I did find this document that is for a Cisco firewall: https://www.cisco.com/c/en/us/td/docs/security/asa/hw/maintenance/5585guide/5585Xhw/pinouts.pdf

So your device does seem to have an odd pinout. My first thought was that you needed to add another rollover cable (via an RJ45 coupler) to re-arrange the pins to work, but that doesn't look  to be the solution either. Odd...
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 20286
  • Over the Rainbow
    • The ELRepo Project
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #2 on: December 03, 2017, 11:52:44 PM »

Thank you for those links.  :)

I am familiar with the concept of a "Cisco rollover cable" with 8P8C plugs at each end . . . where the plugs are mirror images.

Gubbins actually uses the TIA-561 standard interface (just three signal lines out of the eight defined) and there is a table (at the bottom of the page) which shows the full pin-outs for that standard.

Looking at that table I note that two errors appear which, unfortunately, are widespread and common --
  • 8P8C != RJ45
  • DE-9 != DB-9
A true RJ45 plug & socket pair would have been found on AT&T telephony equipment and is not what we see with networking equipment. The latter are modular 8P8C plugs & sockets.

As for the second error, there is no such entity as DB-9 (either with or without the hyphen). The "big" plugs & sockets associated with with the classical RS-232 interface of, say the late 1970s - early 1980s, are DB-25. The "DB" refers to the size of the shell for the 25 pin interface. For the smaller 9 pin RS-232 interface, the shell size is "DE". (There are, if I am remembering correctly, defined shell sizes ranging from "DA" to "DF".)

Looking on the side of my laptop computer I see a DE-9 male socket (serial port) and a DE-15 female socket (VGA port).
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

WWWombat

  • Kitizen
  • ****
  • Posts: 1663
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #3 on: December 04, 2017, 02:09:42 AM »

Hmmm... Faint bells ringing (almost literally in this case)...

The picture on your link to TIA-561 has an "RI" ping, Ring Indicator. That suggests the intention with that pinout is to connect to a modem, which will provide an RI signal. It doesn't get seen on the Cisco pinout.

That made me think that the pinout differences we see could be explained by whether the hardware is intended to act as the DCE or DTE, and what it expects to be connected to the port. It reminded me of old "null modem" setups for when you connect the "wrong" things directly together, and some assumptions had to be made. Here, the 8P8C connector reduces the signals we can use (further, when pins 4 and 5 both carry ground), so it must be making assumptions too.

I've since come across this page. Not just with some pinouts, but with adapters wired differently (one labelled Cisco/Sun console, the other labelled as a Null cable). And neither matches your pinout... ho hum.

There's also a bit of my mind wondering whether this comes over as an artifact of the two different (TIA-568A and B) ways of wiring ethernet sockets. Its too late to try to figure that out though!

I've come across the DB vs DE thing before. On this front, I'm lazy enough to follow the "well understood" path rather than correct my typing, even though I understand when it is put in the proper form. I haven't come across the 8P8C != RJ45 thing before though.
Logged

WWWombat

  • Kitizen
  • ****
  • Posts: 1663
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #4 on: December 04, 2017, 02:12:27 AM »

Just for fun, I found this: http://www.ossmann.com/5-in-1.html
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 20286
  • Over the Rainbow
    • The ELRepo Project
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #5 on: December 04, 2017, 06:44:51 PM »

Just for fun, I found this: http://www.ossmann.com/5-in-1.html

Ah, quite a fun article.  :)

That made me think that the pinout differences we see could be explained by whether the hardware is intended to act as the DCE or DTE, and what it expects to be connected to the port. It reminded me of old "null modem" setups for when you connect the "wrong" things directly together, and some assumptions had to be made.

You have got right to the relevant point.  :thumbs:  And those "helpful" tables which just label a signal at one pin are generally very unhelpful, often to the point of confusion.  ::)

I come from a time when DB-25 connectors were the standard and an RS-232 interface lead was only used to connect an item of DTE to an item of DCE. Progress occurred and I was regularly making up "special" interface leads for those who wished to connect an item of DTE to another (possibly different) item of DTE. Without much effort I can quote "2, 3, 4, 5, 6, 7, 8, 20 & 22" along the defined signal on each pin.

Here follows my helpful table for a serial console lead to use with the gubbins --

Signal        8P8C        DE-9        Signal
   -               1               6            DSR
   -               2               1            DCD
   -               3               4            DTR
GND            4               5            GND
TxD             5               2            RxD
RxD             6               3            TxD
   -               7               8            CTS
   -               8               7            RTS

The first column, headed "Signal", relates to the second column, headed "8P8C". A hyphen in the first column signifies that there is no connection made to the pin of the socket (within the gubbins). The fourth column, also headed "Signal", relates to the third column, headed "DE-9". For each of the eight rows, the pin number stated in column two is connected to the pin number stated in column three.

Having performed an experiment, I now know the above is correct.

The form of the experiment came to me whilst I was laying in my bed, considering the insides of my eyelids, early this morning. For the experiment I used the following items --
  • A standard Ethernet patch lead (not a cross-over Ethernet lead).
  • A 8P8C socket which was labelled "BT  ISDN 2" and complete with the "prancing piper" logo.
  • Three lengths of 0.5mm solid core, PVC insulated, wire.
  • A DE-9, female, plug with screw terminals for the wires.
Having connected the 8P8C socket to the DE-9 plug as in the table, above, a connection was made from the gubbins' serial console port to the computer's RS-232 port. A logging screen session was started --

screen -L -U /dev/ttyS0 9600

Upon powering up the gubbins its boot time messages, followed by an invitation to login, were displayed --

Code: [Select]
NBpNBfz

Covaro Bootloader: 08/06/09.15:04

Helium 500/400 PP boot v1.6

Loading from active partition.
NP
i 100%

Booting... Please wait.
Boot time: Mon Feb 14 13:51:39

Network Interface Device (NID) 2.1.6-406 (Sep  2 2010)



             @@@       @@@@@@@@       @@@       @@@       @@@
            @@@@@      @@@@@@@@@@     @@@       @@@      @@@@@
           @@@ @@@     @@@     @@@    @@@       @@@     @@@ @@@
          @@@   @@@    @@@      @@@   @@@       @@@    @@@   @@@
         @@@     @@@   @@@       @@@  @@@       @@@   @@@     @@@
        @@@       @@@  @@@       @@@  @@@       @@@  @@@       @@@
        @@@@@@@@@@@@@  @@@       @@@   @@@     @@@   @@@@@@@@@@@@@
        @@@@@@@@@@@@@  @@@      @@@     @@@   @@@    @@@@@@@@@@@@@
        @@@       @@@  @@@     @@@       @@@ @@@     @@@       @@@
        @@@       @@@  @@@@@@@@@@         @@@@@      @@@       @@@
        @@@       @@@  @@@@@@@@            @@@       @@@       @@@

   Copyright (C) 2002-2010 ADVA Optical Networking. All rights reserved.



Login:

b*cat is pleased.  :)
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

WWWombat

  • Kitizen
  • ****
  • Posts: 1663
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #6 on: December 05, 2017, 01:11:35 AM »

Ah, more ADVA equipment  :cool:

On the point of confusion ... Am I right in thinking that these 8P8C, Ethernet connector thingys, have pins numbered in one direction for the socket, but have the lead that plus into the socket numbered in the opposite direction?

I have an old, sneaky, recollection that BT does this with the phone voice connectors, but some of the conversion tables I saw yesterday made me think the same happens in the 8P8C world too.
Logged

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 20286
  • Over the Rainbow
    • The ELRepo Project
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #7 on: December 05, 2017, 02:05:42 AM »

For a modular 8P8C plug, I consider it as if I am holding it in the correct orientation to plug it into a socket -- cable towards me and the pins uppermost, latch below -- and number the pins from the left to right. In my own mind, I always consider the socket to be similarly numbered.

Before I used the (new old stock) "BT  ISDN 2" socket in my experiment I checked its IDCs numbering and how it compared to the pins of a plug. That involved plugging a Chesivale Electronics "lollipop" (an Adaptor Test 50A, also known as an RJ45 Modular Adaptor -  M0110/01A) into the socket and checking the continuity, essentially pin to IDC with a DMM. There was a one to one relationship.

Like you, I am aware of the numbering inversion when the ubiquitous BT/Openreach telephone plugs and sockets are critically examined.
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.

burakkucat

  • Global Moderator
  • Senior Kitizen
  • *
  • Posts: 20286
  • Over the Rainbow
    • The ELRepo Project
Re: A Proprietary, Non-Standard, Interface . . .
« Reply #8 on: December 05, 2017, 02:11:19 AM »

Ah, more ADVA equipment  :cool:

Just between ourselves, the "magic incantation", when in an un-deployed, factory reset state, is "root/ChgMeNOW".
Logged
:cat:  100% Linux and, previously, Unix. Co-founder of the ELRepo Project.

Please consider making a donation to support the running of this site.