USB Cable impedance

Thread Starter

odm4286

Joined Sep 20, 2009
265
Hello all, let me preface my question by asking whoever responds to please keep things light. I understand the math related to impedance but I'd like focus on getting a more intuitive feel for this concept. Now correct me if I'm wrong, but impedance is resistance plus reactance. Reactance is a resistance that changes with frequency, a capacitor for example.

My question is, what exactly causes the 90ohms of impedance in a USB cable? If you take one apart all you have are 4 conductors and a shield, what contributes to the reactance? Furthermore, why is a "homemade" USB cable that does not have a 90ohms impedance unable to reliably transmit data?
 

nsaspook

Joined Aug 27, 2009
13,085
A intuitive feel for cable impedance involves looking at the cable as a waveguide for the electromagnetic energy that travels in the space between conductors. The characteristic impedance is the ratio for the electric and magnetic field components of that electromagnetic wave due to the geometry and materials of the transmission line . (ratio of the amplitudes of voltage and current of a single wave propagating along the line)
 
Last edited:

Thread Starter

odm4286

Joined Sep 20, 2009
265
A intuitive feel for cable impedance involves looking at the cable as a waveguide for the electromagnetic energy that travels in the space between conductors. The characteristic impedance is the ratio for the electric and magnetic field components of that electromagnetic wave due to the geometry and materials of the transmission line . (ratio of the amplitudes of voltage and current of a single wave propagating along the line)
Interesting, could you explain why data isn't transferred correctly when the impedance value is changed? Does it have to do with maximum power transfer, I.E. impedance matching?
 

ebp

Joined Feb 8, 2018
2,332
You can find all sorts of info on cable impedance on the web. The cable does not look reactive to a fast signal - it looks resistive.

Mathmatically, ignoring resistance of the conductor, which is usually a small player and the conductance of the insulation, which is even smaller
z = (L/C)^0.5 The characteristic impedance of the cable in ohms is equal to the square root of the inductance in henries divided by the capacitance in farads

As you bring two conductors closer together you increase the capacitance between them and decrease the effective inductance (increase the mutual inductance because each conductor "sees" the magnetic field from the other; since the current is flowing in opposite directions in the conductors the directions of the fields are opposing and partially cancel each other). For a given conductor spacing you can alter the capacitance by changing the dielectric constant of the insulation between them. Air has a dielectric constant of 1. Plastics vary according to the specific polymer. Polypropylene, which is a common dielectric for high frequency cables has a dielectric constant of 2.2, meaning for a given thickness between electrodes of a capacitor it will yield 2.2 time as much capacitance as air.

A USB cable is not four simple conductors, it is two conductors for power and ground and a twisted pair for the signal. Twisting the conductors together maintains the spacing so that the characteristic impedance is maintained - a "transmission line" is formed. In fact, stranded wire is pretty awful in terms of trying to maintain consistent spacing. Every time there is a change in impedance in a transmission line occurs, energy is reflected. If you have a whole bunch of reflections the signal integrity is lost as edges "smear" around their intended positions. You can make your own transmission line and have it perform fairly well, but making one of a specific impedance obviously requires the right thickness of dielectric.
 

BobaMosfet

Joined Jul 1, 2009
2,110
As you requested, giving this a 'light' touch. Impedance is really simple.

Impedance is nothing more than the prevention of electron flow. It comes in three forms, singly or altogether: thermal (resistive), capacitive (reactance), and inductive (reactance). Thermal means friction-- like current through a resistor. Reactance means a field-- an electrical field that charges or drains based on changing electron flow. All that you really need to understand is that the less you impede the flow of electrons, the stronger the signal is, and the more you impede the flow of electrons, the weaker the signal is.

You don't need to know anything more than that really, to get an intuitive 'feel' for impedance. Everything else is an engineering detail.
  • Resistors dissipate in heat. What they prevent going through them, they get rid of in friction, in heat.
  • Capacitors react to current flow thusly: A capacitor is a gap- an open circuit- and it simply simulates conductors that have huge diameters. As you load electrons on one face, they push in all directions pushing electrons away from the opposite face, and in fact back at the power-supply from the face they are on. This is why a capacitor only 'holds so much'. Actually, it's taken all the electrons on one plate it can have pushed onto it and it's now pushing back with equal force.
  • Inductors on the other hand, absorb electrons-- they build a field around themselves, and when power is removed from an inductor, that field empties back into the conductor the inductor is made of-- behaving as if it's a power supply. If you charge an inductor up until its field can't take any more, current then flows through the conductor as if a short circuit.
Anywhere you want current to flow is going to have some combination of one or more of these resistances-- whether it's a wire, or anything else. The reason you see people talking about impedance all the time, is because the quantity of electrons allowed to flow determines the strength of a signal. More electrons means stronger signal. Fewer electrons means weaker signal. So a high impedance means high resistance and weak signal. A low impedance means little resistance and a strong signal. If you put that in the context of inputs and outputs, you can begin to see why a strong signal can overwhelm a weak signal, or a weak signal can be overwhelmed by EMI, so you lower the impedance to make it a stronger signal. Or you try to match the output and the input to the same impedance (impedance matching) so the right amount of current flows between and nothing bounces back (causing ringing and malformed waves in analog and digital signaling because that excess current flow has to bleed off by bouncing back and forth on the conductor). Which also means that where there is a change in current flow, there is a change in voltage. And vice versa. If this does not seem to be occurring, it means something is regulating either voltage, or current, or both to maintain them at a level even when one, other, or both try to change. And that regulation is done by changing one of three things-- E, I, or R.

Now that you understand the concept of impedance, the equations and calculations, and how it's talked about will make a lot more sense.
 
Last edited:

nsaspook

Joined Aug 27, 2009
13,085
Impedance is really simple.

Impedance is nothing more than the prevention of electron flow. It comes in three forms, singly or altogether: thermal (resistive), capacitive (reactance), and inductive (reactance). Thermal means friction-- like current through a resistor. Reactance means a field-- an electrical field that charges or drains based on changing electron flow. All that you really need to understand is that the less you impede the flow of electrons, the stronger the signal is, and the more you impede the flow of electrons, the weaker the signal is.

You don't need to know anything more than that really, to get an intuitive 'feel' for impedance. Everything else is an engineering detail.
  • Resistors dissipate in heat. What they prevent going through them, they get rid of in friction, in heat.
  • Capacitors react to current flow thusly: A capacitor is a gap- an open circuit- and it simply simulates conductors that have huge diameters. As you load electrons on one face, they push in all directions pushing electrons away from the opposite face, and in fact back at the power-supply from the face they are on. This is why a capacitor only 'holds so much'. Actually, it's taken all the electrons on one plate it can have pushed onto it and it's now pushing back with equal force.
  • Inductors on the other hand, absorb electrons-- they build a field around themselves, and when power is removed from an inductor, that field empties back into the conductor the inductor is made of-- behaving as if it's a power supply. If you charge an inductor up until its field can't take any more, current then flows through the conductor as if a short circuit.
Anywhere you want current to flow is going to have some combination of one or more of these resistances-- whether it's a wire, or anything else. The reason you see people talking about impedance all the time, is because the quantity of electrons allowed to flow determines the strength of a signal. More electrons means stronger signal. Fewer electrons means weaker signal. So a high impedance means high resistance and weak signal. A low impedance means little resistance and a strong signal. If you put that in the context of inputs and outputs, you can begin to see why a strong signal can overwhelm a weak signal, or a weak signal can be overwhelmed by EMI, so you lower the impedance to make it a stronger signal. Or you try to match the output and the input to the same impedance so the right amount of current flows between and nothing bounces back (causing ringing and malformed waves in analog and digital signaling).

Now that you understand impedance, the equations and calculations, and how it's talked about will make a lot more sense.
:(
 

btebo

Joined Jul 7, 2017
100
As you requested, giving this a 'light' touch. Impedance is really simple.

Impedance is nothing more than the prevention of electron flow. It comes in three forms, singly or altogether: thermal (resistive), capacitive (reactance), and inductive (reactance). Thermal means friction-- like current through a resistor. Reactance means a field-- an electrical field that charges or drains based on changing electron flow. All that you really need to understand is that the less you impede the flow of electrons, the stronger the signal is, and the more you impede the flow of electrons, the weaker the signal is.

You don't need to know anything more than that really, to get an intuitive 'feel' for impedance. Everything else is an engineering detail.
  • Resistors dissipate in heat. What they prevent going through them, they get rid of in friction, in heat.
  • Capacitors react to current flow thusly: A capacitor is a gap- an open circuit- and it simply simulates conductors that have huge diameters. As you load electrons on one face, they push in all directions pushing electrons away from the opposite face, and in fact back at the power-supply from the face they are on. This is why a capacitor only 'holds so much'. Actually, it's taken all the electrons on one plate it can have pushed onto it and it's now pushing back with equal force.
  • Inductors on the other hand, absorb electrons-- they build a field around themselves, and when power is removed from an inductor, that field empties back into the conductor the inductor is made of-- behaving as if it's a power supply. If you charge an inductor up until its field can't take any more, current then flows through the conductor as if a short circuit.
Anywhere you want current to flow is going to have some combination of one or more of these resistances-- whether it's a wire, or anything else. The reason you see people talking about impedance all the time, is because the quantity of electrons allowed to flow determines the strength of a signal. More electrons means stronger signal. Fewer electrons means weaker signal. So a high impedance means high resistance and weak signal. A low impedance means little resistance and a strong signal. If you put that in the context of inputs and outputs, you can begin to see why a strong signal can overwhelm a weak signal, or a weak signal can be overwhelmed by EMI, so you lower the impedance to make it a stronger signal. Or you try to match the output and the input to the same impedance (impedance matching) so the right amount of current flows between and nothing bounces back (causing ringing and malformed waves in analog and digital signaling because that excess current flow has to bleed off by bouncing back and forth on the conductor). Which also means that where there is a change in current flow, there is a change in voltage. And vice versa. If this does not seem to be occurring, it means something is regulating either voltage, or current, or both to maintain them at a level even when one, other, or both try to change. And that regulation is done by changing one of three things-- E, I, or R.

Now that you understand the concept of impedance, the equations and calculations, and how it's talked about will make a lot more sense.
Thanks for this post! Explained more than 4 semesters of school!
 

nsaspook

Joined Aug 27, 2009
13,085
I was just starting with an explanation that's at a slightly lower level than you. :p -- The 'waveguide' concept is farther down the road
I just don't think an electron based snark hunt is the best road to begin with to understand a simple geometric proportions model of distributed capacitance and inductance for transmission lines. In the case of transmission lines no current is flowing across the “characteristic impedance” The electrons are just moving back and forth along the conductors of the transmission line and in a perfection transmission line dissipate no energy and carry no energy.
 

Thread Starter

odm4286

Joined Sep 20, 2009
265
thank you for all of the responses. I have one last question. If I made my own USB "cable" pin for pin and it did not work for data...would this most likely be an impedance issue or a noise issue? If it is impedance related, am I right assuming that the impedance of the homemade cable would be too high, "killing" the signal?

It seems like impedance is more a physics problem than an electronics problem.
 

MrChips

Joined Oct 2, 2009
30,712
Why would you want to make your own USB cable?
How did you make your cable?
What wiring did you use?
What was the length of the cable?
What type of connectors did you use?
Which USB are you testing? USB, USB 1.1, 2.0, 3.0, 3.1?
 

nsaspook

Joined Aug 27, 2009
13,085
thank you for all of the responses. I have one last question. If I made my own USB "cable" pin for pin and it did not work for data...would this most likely be an impedance issue or a noise issue? If it is impedance related, am I right assuming that the impedance of the homemade cable would be too high, "killing" the signal?

It seems like impedance is more a physics problem than an electronics problem.
Electronics is applied physics so it's both. At the 480 Mbps transmission rate of a USB 2.0 it has to be good impedance match (nominally 90 ohms for the signal pair) for cables longer than a few inches. You can try a band-aid fix of adding series damping resistors in the range of 10-20 ohms to reduce ringing with mismatched cables. The real resistance dissipates some of the reflected energy (sees the resistance going and coming back) in the line at the cost of signal strength.

The effects of signal degradation can be easily seen using a 'Eye' diagram.
https://www.edn.com/design/test-and...gram-Basics-Reading-and-applying-eye-diagrams

There's nothing new about eye diagrams, we used them to monitor HF RTTY long distance links years ago. You could usually tell when the signal was fading long before the wave propagation caused signal issues. This allowed us to change frequencies early to maintain 7/24 HF multi-hop communications links over long distances.
 
Last edited:

ebp

Joined Feb 8, 2018
2,332
Your homemade cable will work if you manage to get conductors with the right dimensions (conductor diameter, type and thickness of insulation) and twist them together properly so that the impedance is correct - and consistent along the length, which is not easy with stranded conductors which have a tendency to untwist, increasing the spacing and thus increasing the impedance.

The actual high-speed data transmission for USB is done with "differential" signalling. The receiver is designed to consider the difference in the signal between the two conductors of the transmission line and, as much as possible, reject "common mode" voltage which may be due to noise or difference in the "ground" between the transmitting device and the receiving device. Because the transmission medium is a twisted pair with both conductors very close together, noise that couples to them from external sources tends to couple nearly equally into both, so it is "common mode" - common to both conductors - noise and is largely rejected.

You could make a transmission line of two conductors that are parallel and straight. Think of them lying flat on a surface with a third conductor nearby and parallel. Because that third conductor is closer to one conductor of the pair than it is to the other, it would seem sensible that noise coupling from the third conductor would be greater to the near conductor than the far - and it is. By twisting the conductors of the transmission line, both will spend close to the same time (distance) to the noise generating conductor and hence both will get nearly the same magnitude of noise coupled. The first case was "differential mode noise" because it was not equal in both conductors. The second is "common mode noise" because it is the same in both conductors. It is electrically reasonably easy to reject common mode noise and much more difficult to reject differential or "normal mode" noise because it is nearly impossible to tell the latter from the signal.
 
Top