Length of USB cable w/o twisted data pair to reach Hi Speed 480Mbps

  • Thread starter Ephant
  • Start date
  • #36
Ephant said:
Ok. Let's compute it. Why can't ordinary LEMO connectors survive 480Mbps?? In original USB male plug, is it impedance matched to the twisted wires?

Because most likely that won't result in the right impedance. The impedance when you use a specific connector will depend on the geometry (distance between pins etc) of the connector itself, the dielectric constant of the materials used, how the shield of the cable it attached, in what order the wires are attached etc....There are lots of factors; and even very small differences will result in changes in the impedance; and if the impedance if wrong the signal coming through the cable will be reflected.

There is absolutely no way to calculate this analytically and get even close; simulations of this type are always done using EM simulation software (although a generic FEM solver like COMSOL might also work ).

It is entirely true -as is you were told- that there is nothing "special" about the LEMO connector sold for USB 2.0. But the point is that that particular type happens to have geometry that result in in the right impedance; this obviously does not mean that ALL Lemo connectors will work since Lemo makes connectors of many different sizes and with different pin configurations .

if you want to have a "well document" example of this just compare SMA connectors of different impedance; the 75 ohm and 50 ohms versions look very similar (and I think you they even fit together if you use some force, NOT recommended); however the impedance differs by 50% due to small differences in things like the diameter of the inner pins. If you try to use a 75 ohm connector in a 50ohm environment it will work fine at low frequencies; but at higher frequencies you will get lots of reflections.
 
Last edited:
  • Like
Likes berkeman and Ephant
Engineering news on Phys.org
  • #37
f95toli said:
Because most likely that won't result in the right impedance. The impedance when you use a specific connector will depend on the geometry (distance between pins etc) of the connector itself, the dielectric constant of the materials used, how the shield of the cable it attached, in what order the wires are attached etc....There are lots of factors; and even very small differences will result in changes in the impedance; and if the impedance if wrong the signal coming through the cable will be reflected.

There is absolutely no way to calculate this analytically and get even close; simulations of this type are always done using EM simulation software (although a generic FEM solver like COMSOL might also work ).

It is entirely true -as is you were told- that there is nothing "special" about the LEMO connector sold for USB 2.0. But the point is that that particular type happens to have geometry that result in in the right impedance; this obviously does not mean that ALL Lemo connectors will work since Lemo makes connectors of many different sizes and with different pin configurations .

if you want to have a "well document" example of this just compare SMA connectors of different impedance; the 75 ohm and 50 ohms versions look very similar (and I think you they even fit together if you use some force, NOT recommended); however the impedance differs by 50% due to small differences in things like the diameter of the inner pins. If you try to use a 75 ohm connector in a 50ohm environment it will work fine at low frequencies; but at higher frequencies you will get lots of reflections.

How about the USB 3.1 with 5Gbps vs 480Mps or 10X more. Does the original Lemo connectors matching the 9 pin also happened to have the same or very similar impedances? If so, then it may be easy to match impedances between wires and connectors, is it not, because in the case of the Lemo USB 3.1. It matches coincidentally too.
 
  • #38
Ephant said:
How about the USB 3.1
What do you think?

You've been given dozens of messages on how things work in general. What have you learned, and how do you think it applies here?
 
  • #39
I learnt wavelength of 480Mbps is 0.62 meter versus 0.06 meter for that of 5 Gbps. I assume 5 Gbps is same as 5Gigahertz (?)

if the wavelength is much shorter, the tolerance in the connector must even be more exact (?) I learnt impedance in the connector is related to the geometry (distance between pins etc) of the connector itself, the dielectric constant of the materials used, how the shield of the cable it attached, in what order the wires are attached etc.

Since the impedance of both USB 2.0 and USB 3.1 are both 90 Ohms. How could the connectors of 9 pins versus 4 pins be similar in impedance, and coincidentally? Unless Lemo designed the 9 pin impedance of the connector on purpose. And if they do, then they could also have redesigned the impedance of the 4 pin (?)

Let's wait for f95toli computations of 480Mbps versus 5 Gbps tolerance and see if both connectors can still coincidentally match it without any modifications. Also I watched a lot of youtube about impedance mismatch.



wire einstein.JPG
Without capacitance and inductance to cause impedance and impedance mismatch. Why would it violate Einstein Theory of Relativity which states that information can never travel faster than the speed of lights?
That guy is everywhere, even on wires.
 

Similar threads

Replies
1
Views
9K
Replies
4
Views
7K
Back
Top