Thanks so much for this, I've been looking around for someone to explain reflections / impedance succinctly, and that's exactly what you've done. Hands on ftw!
great video. Thank you. It would have been good to add a brief explanation to the reflections as a function of wavelength. ie, at what ratio of signal's wavelength to the length of the cable do the reflections begin to be visible/measurable? Why don't we typically see this issue with low frequency signals?
Hats off to you for this excellent video! Great job explaining the concepts and demonstrating some characteristics that may be seen when troubleshooting a signal reflection issue. This video also highlights the need for termination resistors in high-speed digital and RF applications.
thank you so much, i was studying the electrical concept of a Time Domain Reflect-meter, and i saw exactly what you were explaining on this video on my oscillascope. i did not understand, and did not find the answer anywhere until i saw your video. thanks.
Great video! You should have gone a little bit further and mentioned that the width of the reflection pulse can be used to calculate the length of cable, using the speed of electricity in that particular type of coaxial cable, i.e. 82% or 66% the speed of light
Awesome demonstration and explanation. I will have to set this up in my lab now to work through the numbers myself now that I understand what I will be looking at. Thanks a lot.
So that's how you can measure the impedance of a cable with a scope. I guess you can take the potmeter off after leveling the squarewave and measure the resistance and that would be equal to the impedance. That's good to know.
@hlbpr You do actually get reflected pulses even with DC, but the pulses come and go so fast you will never "see" them with a meter. That's why we only worry about high-frequency signals: when the reflection time of the pulses approaches the period of the waveform we're interested in. Then the reflections corrupt the signals we're trying to put on the cable. This is also why you can avoid a lot of signal reflection problems in a computer network by simply turning down the data rate.
@humaxf1 Reflections create standing waves at a constant frequency, so yes this is related to SWR. The better the impedance match at the cable's ends, the less reflections, and the better your SWR.
Gee, i have just rewatched this video great stuff.. CanBuss is on the factory floor with more and more sensors being hooked up to PLC_Can_networks... great when they work as per all things... however, when very short-time errors occur, i need to have a way to catch these problems that may happen once per hour, day, month etc... thanks...:)
Can you explain how an oscilloscope can read RF waveform and voltage accurately with its 1 megohm (typical) impedance? In other words, why aren't there standing waves on oscilloscope leads that throw off the measurements by a huge amount?
It was a really good explanation, thanks a lot. What I'm trying to find out right now is how to use this setup to detect breaks in an XLR-cable, so I can find the exact position of the break, cut the cable before this point and re-solder everything. My problem is that I find a lot of information on coax cables (especially about the velocity factor), but I cannot find any information about the velocity factor for XLR-cables. Which is quite bad if I want to find out the position of the break. Any hints or doesn't it work at all? Thanks a lot!
Questions: 1. How does the signal knows difference between human made joints and factory made joints? i means as you said - "signal reflects from the joint you made to red and black wire". so how does signal knows that , that joint is made by you so signal should reflect from there. and not from in between the cable because that is factory made. does signal have brain? 2. what do you mean by signal propogation? do you mean that signal (current or voltage whatever) runs in the wire with speed of light, so it take some time to reach them to the end. and then when that current come back to the other end, their phase difference is called impedence mismatching? 3. How will you explain impedence without complex numbers and any mathamatics(because i dont want to know exact amount, instead i want to know "process". mathamatics only shows amount, and hides actuall process what going on there.) 4. when they say "for maximum power transfer, impdence should be matched", what does it means to match impedence? do they means phase of current and voltage should be same? because bothe capacitance and inductance make phase shift by 90 degree. so you should add right amount of capacitance and inductance to make voltage and current "in phase" (because phase shift by inductor is opposite of phase shift by capacitor). i just want to know what "causes" this "impedence mismatching" "physically". please dont bring complex numbers and mathamatics here.
I appreciate the attempt to simplify the concept of signal reflection,but i think it can be better explained if we look at your circuit like a LC circuit connected to a square wave source with small internal resistance.The L is of course the loop of wire and the open end of the cable is acting as a capacitor , this model would explain the sudden small step in voltage(voltage across inductor (wire loop)) when square pulse is going positive and also the overshoot as the inductor dissipates its stored energy into the cap. similar analogy can be applied to your other circuits (closed loop and resistive termination) as well. how can you justify that small step is a reflection(voltage divider) and not the voltage across the inductor (transient response)?
The open circuit is like a capacity. That is an unrealistic analogy. The voltage applies an electric pressure on the cable. When it is released (ie the square wave makes the transistion back to zero ) the electric pressure is released and the energy moves back down the cable. When a resistor terminates the cable the energy passes through the terminating resistor. If the resistor is of a low value most energy is dissipated as heat.
Awesome video! What type of cable did you use? I have a 500ft roll of 75ohm coax, a signal generator, and a 25mhz scope and I want to try this experiment! Does the characteristic impedance of the signal generator matter? I think its 50ohm but I'm not sure
very interesting. For a RS485 cable would you always search the perfect resistance with a potentiometer? typically 120 ohm is used but testing the line with scope+function generator + potentiometer would terminate it perfectly.
Thanks again for your videos. I am going to try this but I thought it may be interesting to use the 2nd channel of the scope and look for the voltage drop across the cable and if the scope has the capability, use the math features to add or subtract the channels. Was the ohm value of the variable resistor equal to the specified impedance at the point where the wave looked square again?
I think i understand now what your saying is that a signal goes flying down the cable charging its self capacitance and inductance then at zero cross the inductors and caps release a charge if there is mismatch PLEASE tell me if I'm right. Its a bit like Powerfactor correction in 50 or 60 cycle systems yes? Ok sorry to ask so much but I'm want to learn
So, if you measure the resistance of the potentiometer at the position where you have the clean, square wave form , that would indicate the characteristic impedance of the cable or wire. That should work on an antenna , right? Thanks for the video!!
@NotSoLiberal It's a shame this isn't demonstrated more often -- it's not a very difficult thing to demonstrate. All you need is a square-wave source, a long cable, and a 'scope. Feel free to forward the video link to your instructors, maybe giving them some ideas!
oh yeah, i find during my experiment that the reflection time, is approximately 700ns, doing simple math: 983571056 * 0.0000007/2 * 0.7 speed of light, times reflection time, divide by 2 b/c of cable lengh, and multiply by 0.7 because wire is not ideal (does not travel at the speed of light), its propagation characteritic is 0.7. the answer is 240 feet, that's exactly about the size of my spool, this is a NIFFY trick to calculate total length of a wire.
HI, i would like to ask a question. I have done this for my experiment. The wave I obtained for the reflected wave is positive and not negative as shown in the video although my incident signal still remain in positive signal. Is there any reason?
"Termination" simply means a way to "end" the cable, or better put, how to *properly* "end" the cable so that it won't generate reflected waves that will mess with the integrity of the signal.
Nice video, thanks for taking the time to prepare and post it. Some of the other commenters asked if the signal generators impedance mattered..... A good way to think of it is this: Instigation (as in the signal generators output impedance) Propagation (the cable’s characteristic impedance) Termination (the terminating load impedance) All of the above three components need to match to ensure efficient energy transfer from one to the other without reflections. So if you are using a 50 ohm signal generator, use a cable with a 50 ohm characteristic impedance and ensure that you terminate it into a 50 ohm impedance load.
Hey Tony I'm just getting into differential signaling and communications. I was wondering what the meaning behind the terminating resistors was. It's throwing me for a loop . I would greatly appreciate it. Thanks Tony
C (or capacitance) exists b/c the two wires are spaced close together separated by air, and so it behaves like a capacitor. B/c the capacitance of the wire is removed (when the wires are separated), signal does not have the energy to reflect back. this is just my theory, please explain if you know the answer. thanks
This method does not work when the wires are decoupled from each other (do not run in parrallel). let say, if you remove the + and - strand of the wire, make + wire lays in north to south direction, and - wire lays in west to east direction, you would not see the reflection delay. I do not understand, but what what I remember in my electronic class, a while is modeled by a repeating RC circuit, hence it has an impedance characteristic. When two wires are removed, wire charact. becomes R, no C
That was the most simple, understandable and complete explanation I had ever seen... thank you!!!
Thanks so much for this, I've been looking around for someone to explain reflections / impedance succinctly, and that's exactly what you've done. Hands on ftw!
great video. Thank you. It would have been good to add a brief explanation to the reflections as a function of wavelength. ie, at what ratio of signal's wavelength to the length of the cable do the reflections begin to be visible/measurable? Why don't we typically see this issue with low frequency signals?
awesome .finally i did understood what characterstic impedance really means thank you
Hats off to you for this excellent video! Great job explaining the concepts and demonstrating some characteristics that may be seen when troubleshooting a signal reflection issue. This video also highlights the need for termination resistors in high-speed digital and RF applications.
A brilliant and very intuitive explanation. Thank you!
Thanks a lot . Your practical demonstration about reflecting waves is quite clear and understandable.
thank you so much, i was studying the electrical concept of a Time Domain Reflect-meter, and i saw exactly what you were explaining on this video on my oscillascope. i did not understand, and did not find the answer anywhere until i saw your video. thanks.
Finally a real world explanation of the issue, thank you so much. Now I understand.
I enjoyed watching this video 2 years ago, and I'm enjoying watching it again now!
You saved us almost 5 hours!!!
A big thank you from SNU!
Great video! You should have gone a little bit further and mentioned that the width of the reflection pulse can be used to calculate the length of cable, using the speed of electricity in that particular type of coaxial cable, i.e. 82% or 66% the speed of light
Awesome demonstration and explanation. I will have to set this up in my lab now to work through the numbers myself now that I understand what I will be looking at. Thanks a lot.
You are very welcome. It took me years to finally understand what characteristic impedance means as well!
Best explanation/video I have seen on this subject!! Really enjoyed it !!
first class explanation,,, Excellent
Nicely explained. Great job
Great video, Great explanation. I wish my instructors had shown something like that when we were learning about transmission lines.
So that's how you can measure the impedance of a cable with a scope. I guess you can take the potmeter off after leveling the squarewave and measure the resistance and that would be equal to the impedance. That's good to know.
Good & enjoyable practical tutorial. Thank you.
Also, gratitudes to Oliver Heaviside for, firstly, theorizing these concepts.
Wow. Perfect explanation. Good teacher. Thank you.
Beautiful. Thank you for this.
Thank you... helped me a lot in my presentation on COAX :D
Impedance & capacitance of the cable affects more the signal the longer the network. It acts as a big capacitor in addition to the wire resistance.
@hlbpr You do actually get reflected pulses even with DC, but the pulses come and go so fast you will never "see" them with a meter. That's why we only worry about high-frequency signals: when the reflection time of the pulses approaches the period of the waveform we're interested in. Then the reflections corrupt the signals we're trying to put on the cable.
This is also why you can avoid a lot of signal reflection problems in a computer network by simply turning down the data rate.
Thanks Tony I appreciate it buddy keep up the good work
Thank you for this demonstration.
Excellent practical explanation
@humaxf1 Reflections create standing waves at a constant frequency, so yes this is related to SWR. The better the impedance match at the cable's ends, the less reflections, and the better your SWR.
Very neat explanation.Thx
Great visual explanation, I know know what some of the waveforms I've seen mean.
Well done! Thank you.
Thank You , great job
amazing explanation
Excellent, very appreciate.
Excellent!
Excellent!...and cheap to try
Gee, i have just rewatched this video great stuff.. CanBuss is on the factory floor with more and more sensors being hooked up to PLC_Can_networks... great when they work as per all things... however, when very short-time errors occur, i need to have a way to catch these problems that may happen once per hour, day, month etc... thanks...:)
Can you explain how an oscilloscope can read RF waveform and voltage accurately with its 1 megohm (typical) impedance?
In other words, why aren't there standing waves on oscilloscope leads that throw off the measurements by a huge amount?
That's so great
It was a really good explanation, thanks a lot.
What I'm trying to find out right now is how to use this setup to detect breaks in an XLR-cable, so I can find the exact position of the break, cut the cable before this point and re-solder everything.
My problem is that I find a lot of information on coax cables (especially about the velocity factor), but I cannot find any information about the velocity factor for XLR-cables. Which is quite bad if I want to find out the position of the break.
Any hints or doesn't it work at all?
Thanks a lot!
Questions:
1. How does the signal knows difference between human made joints and factory made joints?
i means as you said - "signal reflects from the joint you made to red and black wire". so how does signal knows that , that joint is made by you so signal should reflect from there. and not from in between the cable because that is factory made. does signal have brain?
2. what do you mean by signal propogation? do you mean that signal (current or voltage whatever) runs in the wire with speed of light, so it take some time to reach them to the end. and then when that current come back to the other end, their phase difference is called impedence mismatching?
3. How will you explain impedence without complex numbers and any mathamatics(because i dont want to know exact amount, instead i want to know "process". mathamatics only shows amount, and hides actuall process what going on there.)
4. when they say "for maximum power transfer, impdence should be matched", what does it means to match impedence? do they means phase of current and voltage should be same? because bothe capacitance and inductance make phase shift by 90 degree. so you should add right amount of capacitance and inductance to make voltage and current "in phase" (because phase shift by inductor is opposite of phase shift by capacitor).
i just want to know what "causes" this "impedence mismatching" "physically". please dont bring complex numbers and mathamatics here.
I appreciate the attempt to simplify the concept of signal reflection,but i think it can be better explained if we look at your circuit like a LC circuit connected to a square wave source with small internal resistance.The L is of course the loop of wire and the open end of the cable is acting as a capacitor , this model would explain the sudden small step in voltage(voltage across inductor (wire loop)) when square pulse is going positive and also the overshoot as the inductor dissipates its stored energy into the cap. similar analogy can be applied to your other circuits (closed loop and resistive termination) as well. how can you justify that small step is a reflection(voltage divider) and not the voltage across the inductor (transient response)?
The open circuit is like a capacity. That is an unrealistic analogy. The voltage applies an electric pressure on the cable. When it is released (ie the square wave makes the transistion back to zero ) the electric pressure is released and the energy moves back down the cable. When a resistor terminates the cable the energy passes through the terminating resistor. If the resistor is of a low value most energy is dissipated as heat.
Awesome video! What type of cable did you use? I have a 500ft roll of 75ohm coax, a signal generator, and a 25mhz scope and I want to try this experiment! Does the characteristic impedance of the signal generator matter? I think its 50ohm but I'm not sure
very interesting. For a RS485 cable would you always search the perfect resistance with a potentiometer? typically 120 ohm is used but testing the line with scope+function generator + potentiometer would terminate it perfectly.
very useful
Thanks again for your videos. I am going to try this but I thought it may be interesting to use the 2nd channel of the scope and look for the voltage drop across the cable and if the scope has the capability, use the math features to add or subtract the channels. Was the ohm value of the variable resistor equal to the specified impedance at the point where the wave looked square again?
Very amazing!
What was the frequency at the end?
What was the resistor value when you got the clean square wave?
Probably 50 Ohms. The typicall characteristic impedance of a cable .
I think i understand now what your saying is that a signal goes flying down the cable charging its self capacitance and inductance then at zero cross the inductors and caps release a charge if there is mismatch PLEASE tell me if I'm right. Its a bit like Powerfactor correction in 50 or 60 cycle systems yes? Ok sorry to ask so much but I'm want to learn
nice
Cool
Hi, experts! I am wondering why the reflection always showed at Vpp/2? Any comments or advices are welcome. Thank you!
So, if you measure the resistance of the potentiometer at the position where you have the clean, square wave form , that would indicate the characteristic impedance of the cable or wire. That should work on an antenna , right? Thanks for the video!!
@NotSoLiberal It's a shame this isn't demonstrated more often -- it's not a very difficult thing to demonstrate. All you need is a square-wave source, a long cable, and a 'scope. Feel free to forward the video link to your instructors, maybe giving them some ideas!
oh yeah, i find during my experiment that the reflection time, is approximately 700ns, doing simple math:
983571056 * 0.0000007/2 * 0.7
speed of light, times reflection time, divide by 2 b/c of cable lengh, and multiply by 0.7 because wire is not ideal (does not travel at the speed of light), its propagation characteritic is 0.7. the answer is 240 feet, that's exactly about the size of my spool, this is a NIFFY trick to calculate total length of a wire.
Is reflection a result of self induction? I just found out about this phenomenon and I'm trying to understand what happens.
HI, i would like to ask a question. I have done this for my experiment. The wave I obtained for the reflected wave is positive and not negative as shown in the video although my incident signal still remain in positive signal. Is there any reason?
How the signal reflect back? Assume we generate pulse signal from device into cable but how extractly the pulse reflect back to device?
"Termination" simply means a way to "end" the cable, or better put, how to *properly* "end" the cable so that it won't generate reflected waves that will mess with the integrity of the signal.
Nice video, thanks for taking the time to prepare and post it.
Some of the other commenters asked if the signal generators impedance mattered.....
A good way to think of it is this:
Instigation (as in the signal generators output impedance)
Propagation (the cable’s characteristic impedance)
Termination (the terminating load impedance)
All of the above three components need to match to ensure efficient energy transfer from one to the other without reflections.
So if you are using a 50 ohm signal generator,
use a cable with a 50 ohm characteristic impedance
and ensure that you terminate it into a 50 ohm impedance load.
Hey Tony I'm just getting into differential signaling and communications. I was wondering what the meaning behind the terminating resistors was. It's throwing me for a loop . I would greatly appreciate it. Thanks Tony
Impedance matching, when you have it you have full transfer of energy to the resistor and none reflected back. If I understand...
Can you calculate the length of the wire from this?
C (or capacitance) exists b/c the two wires are spaced close together separated by air, and so it behaves like a capacitor. B/c the capacitance of the wire is removed (when the wires are separated), signal does not have the energy to reflect back. this is just my theory, please explain if you know the answer. thanks
Looks like you need to fix that overshoot.
This method does not work when the wires are decoupled from each other (do not run in parrallel). let say, if you remove the + and - strand of the wire, make + wire lays in north to south direction, and - wire lays in west to east direction, you would not see the reflection delay. I do not understand, but what what I remember in my electronic class, a while is modeled by a repeating RC circuit, hence it has an impedance characteristic. When two wires are removed, wire charact. becomes R, no C
is this similar to SWR and an antenna?
2.1x10^5 m
Mathiesen :D hehe
that was helpful, but if the connections are showed it would be helpful to do the experiment personally.
I think you forgot to turn on the sound of the microphone.
Maybe not. Maybe it's just capacitance? This has nothing to do with time, that is silly.