M
M.I.5¾
[Top posting corrected]
Whilst you are right in the grand scheme of things, the entire gamut of
degradation they discus is in reletively long transmission lines. A 5 metre
USB cable is a relatively short transmission line. Provided the cable is of
adequate specification, the main driver for limitation of length of cable is
reflections. If these could be eliminated (and they could be - if the users
were prepared to pay humungous amounts of cash for the terminations and
connectors), USB cables could be far longer than 5 metres before the other
factors come into play. Indeed the method of termination has to change
between USB1 speeds and USB2, because the simple resistive termination
technique used for USB1 speeds creates far too much reflection for USB2
speeds to reliably operate which has had to employ back termination. Of
course, if cable that is barely able to carry USB1 signals is employed to
carry USB2 signals then your point becomes very valid indeed..
We use serial bus systems for aerospace purposes and as long as the bus
segments are short, there are no problems. But our limit on bus segments is
far shorter than USB at just 5 feet, after which reflections can stop the
system working. Certainly none of the other factors applies to any
significant extent to the expensive highly specified cable at that short
length*. There are techniques that we have to use when a segment exceeds 10
feet to minimise reflections, but the method we use involves transformers
and resistors to absorb the reflections.
*In fact withf the reflections are properly controlled, the cable works
without problems when it runs the length of an entire Airbus A380 (bearing
in mind that the cable is unlikely to go in a straight line), though fibre
optics are taking over.
Whilst you are right in the grand scheme of things, the entire gamut of
degradation they discus is in reletively long transmission lines. A 5 metre
USB cable is a relatively short transmission line. Provided the cable is of
adequate specification, the main driver for limitation of length of cable is
reflections. If these could be eliminated (and they could be - if the users
were prepared to pay humungous amounts of cash for the terminations and
connectors), USB cables could be far longer than 5 metres before the other
factors come into play. Indeed the method of termination has to change
between USB1 speeds and USB2, because the simple resistive termination
technique used for USB1 speeds creates far too much reflection for USB2
speeds to reliably operate which has had to employ back termination. Of
course, if cable that is barely able to carry USB1 signals is employed to
carry USB2 signals then your point becomes very valid indeed..
We use serial bus systems for aerospace purposes and as long as the bus
segments are short, there are no problems. But our limit on bus segments is
far shorter than USB at just 5 feet, after which reflections can stop the
system working. Certainly none of the other factors applies to any
significant extent to the expensive highly specified cable at that short
length*. There are techniques that we have to use when a segment exceeds 10
feet to minimise reflections, but the method we use involves transformers
and resistors to absorb the reflections.
*In fact withf the reflections are properly controlled, the cable works
without problems when it runs the length of an entire Airbus A380 (bearing
in mind that the cable is unlikely to go in a straight line), though fibre
optics are taking over.