Fascinating.
What is the make/model of the MB? Name of the program?
I'd sure like to know how they're doing this with a 'normal' chip set.

Of course one can do simple TDR using a fast rise
time square wave generator and a fast scope.
And there are a number of $200 'cable testers' that can
do this.
And Circuit Cellar had a winning project some years back
which did TDR, crosstalk tests and the like with results
displayed in text on a 2x20 LCD.

Robert

"M. Adam Davis" wrote:
>
> My latest motherboard has built in 1000baseT (gigabit ethernet over
> twisted pair) and comes with a utility that tests cables.  It'll provide
> minimal TDR capabilities - in addition to the usual shorts/opens, it'll
> tell you where in the cable the short/open is, how long the cable is,
> various parameters about the cable, improperly connected pairs.  Some
> tests need a similar network card on the other end, but most run with
> just one end connected to this board.
>
> I'm hoping to find a PCI or PCMCIA card with this capability - it'll
> save me tons of time!  What I'd /really/ like is a handheld driving one
> of these cards.  It'd be neat to look at the chipset info and see if
> it's interfaceable with a low level controller...
>
> -Adam
>
> Howard Winter wrote:
>
> >Russell,
> >
> >On Mon, 29 Mar 2004 16:28:52 +1200, Russell McMahon wrote:
> >>Best of all was including a short length of wrong impedance cable in a coaxial ethernet installation.Guess
> >>
> >>
> >how I know :-(
> >
> >Ah, memories!  I remember getting a call from a secretary at our office, who said "Get here and fix this
> >(expletive deleted) thing before I throw it out of the window!"
> >
> >Never being one to underestimate a threat like this, I headed back (from a client's) and found that
> >Wordperfect would take 5 minutes to load, then another 5 to open a file, etc.  It was a network wide problem
> >so I set to with the time-domain-reflectometer (praising the accounts department that I'd been allowed to buy
> >one!) and it showed some odd readings - usually you open the termination at the one end end of the 10-base-2
> >(50-ohm coax) and connect the TDR at the other end and it gives the length of cable to the open end of the
> >network segment, then you work along and as you exclude a length of cable the remaining length shown reduces
> >accordingly.  This didn't!  It intermitently showed a much shorter length, and it wasn't consistent from the
> >two ends,  so I did a "binary chop" to locate where the problem was.  I found that someone had inserted a
> >piece of coax they'd found in a cupboard, to add their machine to the network, and it was (I think) IBM
> >terminal cable: 75-ohm instead of 50.  I replaced it with some of the right stuff, and harmomy was restored.
> >I'd like to say that the secretary was eternally grateful and demonstrated it... but sadly not  :-(
> >
> >I'm rather glad that 10-base-T (twisted pair) caught on and replaced coax - it's much easier and more reliable
> >to use all round.
> >
> >I wonder who got custody of the TDR?

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu