Electronics Tutorial - Cables and noise immunity

#90 In this video I look at cables - how their structure influences their noise immunity and why. To get a better understanding of the phenomenon I look at both magnetic field and electric field immunity separately and how sensitive multiple types of wire are.
Related videos:
NearField Magnetic probes: • EMC debugging - Near f...
NearField Electric probes: • EMC debugging - Near f...
Power supply noise: • Electronics tutorial -...
CM-DM signal separation: • Electronics tutorial -...
Special Thanks to all my supporters on Patreon! Especially @afiskon and Xegomen!
If you liked this video be sure to check out my other videos and you can also subscribe to be up to date with all the new ones!
If you want to support the creation of more and better videos please consider checking out: / feszelectronics

[@danpayerle]

Best video yet I’ve seen on common and differential noise. Excellent information and presentation. Thanks for sharing!

[@asaphemagno3044]

Huge Thanks Fesz. A lot of key concepts. I believe you solved many doubts for many people, including my owns, with this video.

[@TheCalvinSkinner]

I love the large variety of topics in your videos. They are relatively short and concise as well.

[@gwc1410]

Very good. In fact excellent. You are among the best electronics instructors. I am surprised you don't have more subscribers.

[@Pumoneon]

Brilliant demonstrations about basic EMI concepts. Congratulations! Keep up the good work!.

[@FesZElectronics]

Thanks a lot!

[@shaggygoooxide]

"electrically conductive hand" - I always suspected you were a cyborg 😁. Thanks for all your amazing content! Yours is one of my favorite channels!

[@FesZElectronics]

Thank you for the kind words! Cheers!

[@supersuperintendant]

Brilliant instruction.

[@bertronixtv]

Superb. Thanks for sharing free knowledge.

[@MuhammadHanif-bx4pb]

Fesz, your knowledge and your method of teaching is incredible! I'm no electrical engineer or have degree in it and yet I understand it all. I am quite surprised that you didn't have a lot of subs yet, heck even I just stumble on your channel a couple weeks ago! and I almost watch all of your vid! keep it up!

[@FesZElectronics]

Thank you for the kind words! Cheers!

[@rohitchaturvedi2271]

13:30 is this phenomenon called ground loop? Please make a video on ground loop in relation to hum in audio amplifiers.

[@vladimirzibtsev2889]

Thank you so much, your videos are simply amazing. I have a question though - how do I determine which fields I need to protect myself from, magnetic or electric? After all, both are always present in the real environment.

[@konturgestalter]

excellent topics

[@plinsburg]

Thank you

[@marconis.giacomini1543]

Is it possible to build a coil version of the EMC Magnetic Tester for 60Hz, with high magnetic fields between the coils? Have plans to test low frequency rejection magnetic fields on shielded/twisted cables.

[@tai08cdt2]

Hi Frez, many thanks your sharing ! 9:01 Could you tell me about the braided shield of your circuit with 1000R that braided shield wire no connect both ends with ground, isn't it ? I find on internet that it must connect one ends with ground. May i missunderstand something !?

[@FesZElectronics]

Well, the cable is some generic coax - and both ends are connected - I mean one end is connected to the resistor, the other goes into the oscilloscope; there is no part left floating if that is what you mean.

[@tai08cdt2]

@@FesZElectronics So , you are used a coax with one inner core and one outer conduct(shield wire) ! I confused that you are used a twisted pair cable with two inner core and one outer conduct(shield wire). Thank you your reply .

[@juanchirino7135]

sometimes I saw ferite cores used in the main AC cable line of power supply. The proupose of this is reduce noice coming back from the power supply to the AC power network cause of the added inductance to the line? Regards from Argentina

[@egoauxiliatussum...5926]

Those would be for high frequency noise rejection yes. For example a ferrite core might add a few hundred ohms of impedance at 100MHz, but nothing noticeable at 50/60Hz. The mains AC noise at those high frequencies would then be attenuated by the voltage divider created by the ferrite core and the input impedance of your system. It can be very effective in practice, and of course the actual values can vary greatly. Great video once again! Looking forward to part 3 of the effects series

[@FesZElectronics]

I would just add something on this topic - if the ferrite cores are added on both wires at the same time, they are used as common mode chokes - the added inductance is on signals traveling on both lines at the same time rather than on differential signals (signal coming in one line and going out the other)

[@david_6063]

This is the first of your non LTSpice tutorials that I've had the time to watch. I need to say this upfront, and then going forward, I won't repeat myself on this point. I've been in the profession for 30+ years and this is some of the best discussion and teaching of electronics that I've ever seen. As I far as I'm concerned the work that you do as a teacher is nothing less than a tremendous service to humanity. Now that that's out of the way, let's get into the EE. 1. In the initial set-up (at about 2:00) we see that the naked wires are quite good at picking up EMI. I understand that this tutorial is on how different shielding techniques enable immunity to EMI, and not on component noise, but I am interested to know how much of that noise is thermal noise in the 1K ohm resistor? If it is say 1%, then who cares; let's just focus on the EMI, but if it is larger, say 10% or more, it would be good to know. An order of magnitude comparison would be very interesting. Maybe you could shield the wires, and look at only the thermal noise? 2. Electromagnetic radiation is a wave of orthogonal electric and magnetic fields; the point being that they always occur together. However, as I am shown by your tutorial, they add noise via different mechanisms. It is my understanding from the video, that using devices you have previously built, you are able to look at the noise effects of the magnetic fields and electric fields SEPARATELY. This is great! So we see that the noise due to the magnetic field component is given by Faraday's law; in other words, a voltage is induced (we say that an emf is induced), and this generates a current. The emf will be present whether the circuit (the loop) is open or closed. In the closed case, which is the one we care about, since we have no circuit without a closed loop, the induced emf will cause a current to flow. By twisting the pairs this effect is cancelled (the detailed explanation of this is very complicated), but this much of an explanation is enormously helpful. MY QUESTION here has to do with the next step in building the noise immunity from the magnetic field, which is the coax cable and the shielding around the signal line. I listened several times on this point, and my best understanding here is "the shielding around the signal line acts to absorb or short circuit the magnetic field". Is there anything further that you can say on this point? For example, if I was explaining the shielding effect of a Faraday cage I would say that the external electric field moves the free carriers (electrons) in the metal of the cage in such a way that they create their own electric field within the cage, which when combined with the external field, leaves a net E field of zero within the cage. In other words, can you add a bit more of a physical explanation of how the coax shield attenuates the effect of the magnetic field? Probably like most people, I have a better feel for the effects of the electric field than the magnetic field, and I don't have any physical understanding of this point. 3. Now let's get to the electric field component. Admittedly my thoughts here are less clear. Perhaps you can help. In circuits, if we have a high pass RC circuit and we apply a voltage step with a very fast rise time to the input side of the capacitor, we say that the voltage is coupled from the input side of the cap to the output side of the cap. This is explained in terms of power. There can be no instantaneous voltage change across the cap because that would require infinite power (cap current = C(dv/dt), so dv/dt cannot be infinite). Once the input voltage is coupled to the output side of the cap, that output voltage drives current through the resistor and the cap charges up with the RC time constant, with the output voltage (high pass filter) going to zero. I hate this explanation because it is not physical. I want to know, physically, how does the voltage get from the input plate of the cap to the output plate of the cap? Do you know? In your drawing where you begin to look at the electric field getting into the circuit, you say that "the electric field is capacitively coupled into the circuit, into either the signal line or the ground line." I need some help understanding this. Yes, the EMI contains an E field. But in circuits (not transmission lines), we look at voltage not E field. As I just described for the high pass filter, we say that a voltage is coupled across the cap plates. Is it different in this case? Is the E field literally coupled from one plate to another? I think not, because then why would we need a capacitance, the E field can hit the signal wires just because it is flowing through space. If you want to talk about capacitive coupling, then maybe we need to think about that E field in terms of a voltage that it is generating, and then we need to explain how that voltage gets from one plate to the other. 4. I could ask more about the final part of this tutorial that deals with ground loops, but this comment is already too long, so I'll just repeat what others have said, which is, it would be very much appreciated if you could come up with something to show how to do low noise design at the system level focusing exclusively on grounding; maybe by having sensitive signals on a ground plane which would be isolated from a noisy power return, and how all of these ground signals are best connected. My field is deep sub-micron IC design, and I am very interested and grateful for the opportunity that your videos give me to get better at board level and system level design.

[@FesZElectronics]

Hello David! Thank you for the kind words! I will try to answer point by point 1. This one is relatively easy - based on the value (1k) and the oscilloscope BW (70MHz), the thermal noise is

[@david_6063]

@@FesZElectronics Hello Fes Z, Thank-you so much for your reply and for answering my questions! I really appreciate it. I think it would be worthwhile to look more at the idea of “capacitive coupling”. It occurs everywhere in electronics and it shows up on your white board in the explanation of how the electric field component of EMI gets into the signal and ground lines. For example, you show a high pass RC consisting of a coupling cap and the resistance on the ground line between that cap and ground. To get going, let us look at a simple high-pass circuit. Consider a DC supply VIN which is connected to a switch which is followed in series by a cap and resistor to ground. The input side of the cap is connected to the switch and its output side is connected to the resistor. At t = 0 the switch is open and both cap plates are at ground. At t = (0 + delta), where delta is as small as you like, the switch is closed. Ideally, the switch resistance is close to zero and the resistance of the conducting path between the switch and the input side of the cap is also close to zero. As a result at t = (0 + delta), the voltage of the input side of the cap pops up to equal the voltage VIN of the supply. What happens to output side of the cap? It follows the voltage on the input side of the cap; it also pops up to VIN. WHY? How does that happen? At t = (0 + delta) there has not been any time for any charge flow; both cap plates are still uncharged. The answer to this must have something to do with wave theory. Although my wave theory is weak, I believe that the when we connect the DC source to the cap, this change creates an electromagnetic wave. For typical circuits (not microwave or transmission line), we don’t normally consider such waves because the wavelengths are gigantic in comparison to the circuit dimensions. However, this wave is possibly the explanation for how the output side of the cap follows the input side prior to the movement of charge (prior to the flow of displacement current). But if the wave can pass through the air gap between the plates, what is to prevent it from going in all other directions? There must be something about the fact of the parallel plates that allows the wave to move from the input plate, across the gap between the plates, and then onto the output plate, and nowhere else. At time t = 0 stationary charges fill all points of the closed circuit loop, just like water fills a hose. At t = (0 + delta), when we get VIN volts on the output side of the cap, charge must begin to flow through the resistor because we have VIN volts across the resistor. Just as the flow rate of water through a hose is the same at every cross section of the hose, so too here does the charge move through the circuit. As a positive charge flows down through the resistor, a positive charge moves off of the cap’s output plate and a positive charge moves onto the cap’s input plate. When that first positive charge moves off of the cap’s output plate that plate holds a net negative charge. At the same time the arrival of an equal amount of positive charge on the input plate gives the input plate a net positive charge. As current continues to flow, the amount of negative charge on the output plate increases but is always balanced by an equal amount of positive charge on the input plate. For every positive charge on the input plate there is a corresponding negative charge on the output plate, and for every such pair of charges we have an electric field line starting on the positive charge and ending on the negative charge such that a uniform electric field exists between the plates. Similarly if we now switch the input side of the cap to ground, positive charge flows off of the input plate and appears on the output plate such that the reduction of net positive charge on the input plate is matched by a reduction of net negative charge on the output plate. The cap discharges. So capacitive coupling is really some sort of wave phenomena which is followed by an associated displacement current. The wave part of this, the actual voltage coupling from one plate to the other is not really talked about in regular circuit design, and I don’t understand it. We are just taught that the voltage difference between cap plates cannot change instantaneously. But at a deeper level I believe this capacitive coupling is an electromagnetic wave thing. Maybe if I was a microwave engineer I would have a better understanding of this. In your whiteboard figure the input signal IS actually a wave (EMI) and its capacitive coupling into the signal and the ground lines IS the point. If you have any physical understanding of how capacitive coupling itself works; that is, of how the output plate sort of acts in concert with the input plate, or is some kind of wave guide such that (prior to the start of displacement current), the output plate voltage follows the input plate voltage, I would like to hear it. My thinking has been significantly enhanced and sharpened by just this one video. And I see that you have scores of additional videos on your playlist. I am truly awestruck by this goldmine of knowledge. A whole university level EE department could be built around your videos. Thank you so much for this incredible body of work! My next video will be the one on transmission lines. Beyond that, I look forward to slowly working my way through your incredible video library.

[@FesZElectronics]

Hello David! Well, now that you mentioned the transmission line video - one of the topics I look at there is the transmission delay - Even though you apply a certain voltage on one end of a conductor, you won't see it on the other end instantly (it takes the time it takes for the charges to move trough the conductor - the charges behave like wave whose speed is limited by light speed and the dielectric constant of the environment). I guess its a similar delay within a capacitor - based on the physical distance between the plates. Since you mentioned you work in IC design, I'm not sure if you ran into this topic, but when you have high speed interfaces one of the things to keep in mind is that trace delays need to be equal (to get all signals reaching the other side at the same time). This is done in practice by controlling trace lengths. The reason I bring this up is that because of limited PCB space, the way to achieve this forced trace lengthening is by making a serpentine layout - and depending on how close the loops are, the delay will be shorter (even if the trace length is the same) - this is because of capacitive coupling between sections on the same trace - signal ends up going partially trough the trace, and partially trough the gap between the sections. (more on the topic - www.sigcon.com/Pubs/edn/serpentine.htm ) Now you don't need microwaves to see this effect, just long elements.

[@david_6063]

@@FesZElectronics The use of serpentine routing to control trace delays sounds interesting and gives more insight into the topic of capacitive coupling. I will check out the link! Thank-you!

[@Arijit_VU3ICT]

Ow..

[@juanchirino7135]

First comment!!!!

[@Impedancenetwork]

You seriously need to get better test equipment.

[@FesZElectronics]

Any good suggestions on not extremely expensive equipment? :D