Folkert Rienstra wrote:
> Michael Baeuerle wrote in news:u4crk5-kvm.ln1@micha.freeshell.org
> > "R. LaCasse" wrote:
> > > On 20 Jun 2008 11:21:50 -0400, jeffj@panix.com (Jeff Jonas) wrote:
> > > >
> > > > Using a 5 connector cable with built-in termination is
> > > > neat and easy: just roll up the unused end and connectors!
> > >
> > > I thought the (modular) scsi terminator would have to be within
> > > 3"-->4" past the last drive and that would do it....
> >
> > No, it must always be placed at the end of the cable.
> >
> > Otherwise the signal is not absorbed correctly.
>
> Terminators do not absorb signal reflection.
They should absorb the signal itself (leaving nothing left for
reflection).
> They prevent signal reflection by mimicking that the cable is endless.
Yes, and this don't work in the middle of a cable: Because the
terminator electrically behaves like an endless cable you get the
equivalent of an Y-cable configuration that is forbidden for the same
reason (impedance mismatch and reflection).
Michael Baeuerle wrote in news:gs1ml5-s3d.ln1@micha.freeshell.org
> Folkert Rienstra wrote:
> > Michael Baeuerle wrote in news:u4crk5-kvm.ln1@micha.freeshell.org
> > > "R. LaCasse" wrote:
> > > > On 20 Jun 2008 11:21:50 -0400, jeffj@panix.com (Jeff Jonas) wrote:
> > > > >
> > > > > Using a 5 connector cable with built-in termination is
> > > > > neat and easy: just roll up the unused end and connectors!
> > > >
> > > > I thought the (modular) scsi terminator would have to be within
> > > > 3"-->4" past the last drive and that would do it....
> > >
> > > No, it must always be placed at the end of the cable.
> > >
> > > Otherwise the signal is not absorbed correctly.
> >
> > Terminators do not absorb signal reflection.
>
> They should absorb the signal itself (leaving nothing left for
> reflection).
Nope. In that case they should be zero ohms, which they are not.
>
> > They prevent signal reflection by mimicking that the cable is endless.
>
> Yes, and this don't work in the middle of a cable:
> Because the terminator electrically behaves like an endless cable
> you get the equivalent of an Y-cable configuration
Which in this case that leg is not a problem, as to reflection,
the problem being still the open end on the other leg.
Folkert Rienstra wrote:
> Michael Baeuerle wrote in news:gs1ml5-s3d.ln1@micha.freeshell.org
> > Folkert Rienstra wrote:
> > >
> > > [...]
> > > Terminators do not absorb signal reflection.
> >
> > They should absorb the signal itself (leaving nothing left for
> > reflection).
>
> Nope. In that case they should be zero ohms, which they are not.
No. The transmitter injects energy into the cable when a "signal" is
created. Therefore the law of conservation of energy forces that a
signal can only be "terminated" by absorbing all of its energy.
Zero ohms means zero voltage, zero voltage multiplied with any current
results in zero power and zero power multiplied with any time results in
zero energy. In other words, a zero ohm terminator can't absorb the
signals energy, not even a part of it. Because the signals energy is
still inside the cable, it must be completely reflected.
The difference is the polarity: An open end will result in 100% positive
reflection (that doubles the voltage at the open end), zero ohms will
result in 100% negative reflection (because incoming plus reflected
signal must be zero volts at the zero ohms). Only a match with the cable
impedance will result in no reflection and all of the signals energy can
move inside the terminator.
> > > They prevent signal reflection by mimicking that the cable is endless.
> >
> > Yes, and this don't work in the middle of a cable:
> > Because the terminator electrically behaves like an endless cable
>
> > you get the equivalent of an Y-cable configuration
>
> Which in this case that leg is not a problem, as to reflection,
> the problem being still the open end on the other leg.
>
> > that is forbidden for the same reason
>
> > (impedance mismatch
>
> impedance disturbance within the cable stretch.
>
> > and reflection).
>
> as a result.
The presence of the terminator in the middle in parallel with the rest
of the cable let the incoming signal see half of the matching impedance.
This reflects one third of the signal [1] in the middle at the
terminator, the polarity is inverted because the impedance changes to
lower. One third is absorbed by the terminator and one third runs into
the rest of the cable. Whether the end is open or not has no influence
at this time.
Later, at the open end this third of the signal is competely reflected
without changing the polarity because the impedance changes to higher.
The reflected signal from the end hits the same impedance mismatch at
the terminator and is split again in three parts and so on.
On the transmitters side of the terminator this results in multiple
reflections with different delays.
|>> I thought
|>
|>That appears to be a problem. Are you any good at thinking?
|>
About as good as you are at sarcasm, maybe I'm retarded..an IQ of
80??? **** happens....
|>> the (modular) scsi terminator
|>
|>What about the integrated one?
|>
There isn't one on this drive.... no TP to bus either...some Quantum
drive....
|>> would have to be within 3"-->4"
|>
|>Anything particular about that 3"-->4" ?
|>
Just a quoted note from SCSI help suggestions...... that's why the
External Capacitor is usually close to the last connector, making it awkward
to install sometimes...
|>> past the last drive and that would do it....
|>
|>What exactly made you think that, do what?
|>
SCSI comprehension references....
|>> Now I'm getting the thought
|>
|>You should have that looked after. Something is not quite right.
|>
Try this and that heh?
|>> that the (modular) scsi terminator's
|>
|>> capacitors
|>
|>Capacitors?
That's what terminators use to mark the endless cable or reflect the
impendence....
"R. LaCasse" wrote:
> On Wed, 23 Jul 2008 00:42:17 +0200, "Folkert Rienstra" <see_reply-to@myweb.nl> wrote:
> >
> > [...]
> > Capacitors?
>
> That's what terminators use to mark the endless cable or reflect the
> impendence....
Resistors not capacitors, because the cable impedance is approximately
resistive in the relevant frequency range. The SPI document recommend a
capacitor for the TERMPWR line inside the terminator (but this is only
to bypass its supply, not for termination).
Re: Internal cable purchase issue 
Linear Mode
