Shall We Use a Ferrite Bead in Power Rail or Not? | Explained by Eric Bogatin

[@Jeroen74]

Ferrite beads are also very handy for isolating a chip/subcircuit during debugging or fault finding, or to take it out to measure current. You can also replace it with a zero ohm resistor, or low value resistor and check the influence. So I tend to design them in; after all, removing them from a design in the next iteration is always easier than squeezing them in later ;)

[@TooHear]

Thanks! I really enjoy learning from your videos and people you collaborate. Very interesting topics.

[@RobertFeranec]

@TooHear thank you very much for your support. I appreciate it a lot.

[@marcfruchtman9473]

I just want to thank you Robert for providing some of the best interviews on a such a wide variety of topic for electronics... thank you so much.

[@DerIchBinDa]

Thank you for all the great videos! Learning a lot in how to design my first OHW-project.

[@RobertFeranec]

Thank you very much @DerIchBinDa

[@8546Ken]

When I have used ferrite beads, it was mainly to reduce radiated or conducted EMI in the MHz range. A big advantage is they act as a very lossy inductor so it won't resonate. Ferrite beads don't necessarily have a high DCR. Often it is literally a "bead" of ferrite, which you pass a wire through. Maybe just one pass or 2 or 3 turns. The bead can be large enough to allow the necessary gauge of wire. When you add a ferrite bead to a piece of wire you don't increase the DCR of the wire. We often used ferrite beads on wires that exit the device to suppress EMI. Stray capacitance between the wire and earth may be sufficient, for filtering. Another good use is to pass 2 or more wires through a suitably sized bead or core in order to suppress common mode noise. In fact, you often see giant beads on entire cables, such as those going to a keyboard. Note. even though a ferrite bead s technically an inductor, since is so lossy low Q) they are not specified with an inductance value but as a resistance at a high frequency. For instance, the one I'm looking at is specified as "390 ohms at 100 MHz" even though it's a fraction of an ohm at DC

[@AlbertRei3424]

You are not talking about the same bead, the one you talk about is almost use on every cable as common mode choke, here we talk about PDN design

[@fireblade31967]

Thanks. Very well explained. Clarifies many doubts. Some repetition of already explained stuff makes the lecture longer, but the learning is worth the time. Thanks again.

[@RodrigoFernandesVernini]

It would be interesting to show the difference of LC filter to a simple capacitor. Is there any other video that shows this?

[@wucherer]

Details, Robert. Thank for giving a detailed long explanation. If others cannot follow can simply rewind the video or grab a BANG drink before playing.

[@williameyvaz5565]

Thank you Robert, as always is great to listen to Eric's lectures. Excellent details on PDN filteration design.

[@SuperHaptics]

Wel done Robert! for your channel has grown and become a very valuble source of knowledge for many of us the last couple of years. I remember your doubts at some point but you have found the format to make all your work worth while. Adding the experts that don't necessarily care themselves disseminating their knowledge through their own channels was a game changer and will add value for years to come. Congrats!

[@RobertFeranec]

Thank you

[@Retinatronics]

Hi Robert, I don't know how much I need to thank you for bringing in Eric for these discussion. Please make do make more videos like this.

[@rfrisbee1]

Hmmm. Not a very good explanation or advice with respect to the use of ferrite beads. As at least one other person has commented already, the defining characteristic of a ferrite bead is that over a relatively narrow range of frequencies it converts current into heat. They are otherwise poorly characterised components and when DC current flows through them they quickly saturate. If you do want to put an LC filter in the power supply to a circuit then do the job properly and use an inductor that is characterised for DC filter applications at the current and frequencies that are be present. Eric mentioned the need for significant resistance to damp out oscillations at the LC filter circuit's resonant frequency. This can be applied in a couple of places. The first is in series with the inductor. If the current being drawn by the powered circuit is low enough that the IR drop is not problematic then sometimes it's possible to find an inductor which itself has sufficient resistance so that an additional resistor is not required. If the IR drop would be a problem then the other way to damp the oscillations is to place a series RC network in parallel with the C in the LC network. The C in the RC network typically needs to be much larger than the C in the LC filter. Where the IR drop from 1-ohm or more of series resistance can be tolerated (e.g., Idd < 100mA) it is less problematic to just use an RC filter. There is no ringing to worry about and resistors are far closer to ideal components than inductors. Many ICs have good PSRR rejection up to about 100kHz, so a 1-ohm+10uF filter would provide ~20dB of attenuation at 160kHz. If an IC is being used that is particularly sensitive to noise below 100kHz then an RC filter followed by a (low-noise) LDO regulator would provide very wideband noise filtering. Perhaps even more important is to design the power supply network to have an impedance low enough over a broad range of frequencies so that components powered from it cannot create significant noise that needs extensive filtering before powering other devices.

[@billwest257]

For completeness the test board should have a capacitor filter, maybe have a pair of pins on a header to short out the inductor?

[@pboston6RR]

The discussion would have been easier to follow if the arrow was more visible. Saying “this side” or “that side” doesn’t help if we can’t see the arrow. I’m not an engineering student but I have seen (I think) some of the physical effects that are mentioned, so I watch to see if I can pick up some pointers among the alphabet soup that is abundant in the discussions. Again, even with not understanding the acronyms the discussion is interesting and helpful in my everyday tinkering with Arduino, motors and LEDs. Thanks. Paul Boston (84 yeras old and still interested)

[@voice4voicelessKrzysiek]

I thought that I would be the oldest one listening to this kind of stuff, but I was wrong; you beat me by 10 years 😁🤣👋

[@mattb6001]

doing some delicate power supply design right now for an RF application and much of the power supply design reference indicates ferrite bead usage (to which i oblige) but i still wanted to understand the details behind it. i do not have time to watch this full video right now (should be working...) but i will definitely check this out later today when i get home. Thanks for sharing!

[@guytech7310]

This demo was dishonest because he use a high current switching load, which should have been on a separate voltage supply rail, & there was no load on the probe side. If he use an CLCR filter or set the probe impedance to 50 ohms (instead of 1M) the ringing would have been dampened. Ferrite CLC filters are for filtering out high frequency noise, not low speed, high current switching load (noise). I doubt your RF application has a 10 Khz switching load.

[@jimjjewett]

@@guytech7310 Forgive my ignorance; I'm still trying to understand the basics. Eric Bogatin did talk about a separate voltage supply as a solution, and said the problem was cost. Are you saying an extra supply (or supply per analog pin?) isn't more expensive than a bead and capacitor? Or that the ferrite bead "solution" isn't actually helpful (enough?) to be a cheaper substitute? Or something else that I haven't thought of yet?

[@guytech7310]

@@jimjjewett Digital Rails should be seperate for a high current switching supply. For an application which your using for something like relays, motors, etc just use two rails. One for the Digital logic & another for the switching supply. Really all your need is a separate voltage regulator with sufficient capacitors to handle and transient current loads for the switching loads, and as low current voltage regulator for your digital logical power load. The CLC or CLCR with ferrite is used to dampen high frequency noise create by logic ICs, or perhaps to voltage reference for ADCs when you need ultra low noise for your ADCs. The rest of the ICs just get a decoupling cap close to the IC pins. FWIW: For example I have designed a custom HVAC controller for my home which has 24 relays, a bunch of thermal couplers SPI controllers. This board has two switching regulators (LM2576 3A Buck regulators): 1 for the Digital rail 5V & another 12V for the relays. I also have a couple of 3.3V linear regulators feeding from the 5V rail. One 3.3V is for the microcontroller and the other is for the ethernet MAC controller which draws a lot of current. This Board also has a precision 3.3V reference voltage regulator that is use for the Thermal couplers & the Microcontrollers ADCs.

[@ariparanthaman]

Thanks

[@RobertFeranec]

Thank you very much

[@InssiAjaton]

My practical experience about ferrites has been thar much of the time I want my ferrites to be very low loss, but other times, like the situation of the demo in this video, I want them to be very LOSSY. In other words, a pure LC only shifts the noise frequency spectrum. An LRC filter actually can attenuate. Of course, I often ignore my knowledge of this fact and just start playing with different ferrites, rather than doing a real analysis and picking the correct ferrite(s) after that. Another thing to throw sticks into my spokes is that the ferrites characteristics change by the DC (and AC) current. Ferrite data sheets provide curves about this, but working through all those curves feels so awkward...

[@andrewwhite1793]

A ferrite bead is lossy. These are specified in ohms at a test frequency. If you want a non-lossy inductor, it's called an inductor and specified in uH. With both types the frequency characteristics depend on the core material. Ferrite beads do not work at low frequencies as more turns are needed here.

[@InssiAjaton]

@@andrewwhite1793 Thanks for clarifying, what I tried to say! However, I found some marvelous lossy, yet quite high inductance beads from a Japanese company I had not related to ferrites until I stumbled to them. They show use at pins of TO-147 transistor legs and the results. They work like a charm for a couple of my different applications. You might want to check Toshiba AB4x2x4.5W. Apparently a Japanese number A6J0205 might be a related (or the actual) part number.

[@SeanPorio]

Fantastic and informative presentation!!

[@RicardoPenders]

Thanks for sharing this video Robert, this was a nice refresh to me about this topic. It's always good to refresh your mind if you haven't done anything recently where this is really important.

[@jimjjewett]

Around 38 minutes, am I understanding correctly that (1) The inductor doesn't reduce the peak-to-peak change, but does re-center it around the target voltage. (Which is good for having enough to make logic high, but bad for risking overvoltaage). (2) That the recovery from having wrong voltage is smoothed out, but not really sped up. (3) That adding the capacitor as well greatly reduced the peak-to-peak change.

[@TheBeefiestable]

Thanks, great video. However what i was hoping to have explained by this video: when to use and not use those physical external "ferrite"s that are just a tube around the outside of wires. sometimes a solid piece, sometimes 2 halfs that clamp on (how could that be the same thing?) I see VGA cables with them just thrown on, serial comm cables with them randomly, sometimes multiple loops of the cable, power AND ground wires inside the ferrite, long cables, short cables, why?when?when not? why the word "bead" that clearly is confusing everyone

[@jamen1993]

Here are so many good talks on this topic. It is just amazing!

[@CraftyZA]

Slowly but surely we see switch mode entering high end audio, and we hate it. A new market has emerged for hobbyists to build linear supplies to power these devices. The way I build them is to put the LM7812 right by the exit. Then a 100nf film on the input, and 220nf polorized cap at the output. I then also use a clip on ferrite bead right before the dc plug that plugs into the device. I really want to dig my teeth into active rectification. I use schottky diodes for rectification with some ceramic caps in parallel to each diode.

[@phil85813]

Thanks Robert and Eric. I really like how these videos provide the confidence to go a different way than the manufacturers data sheets. I much prefer Eric's LC-filter values than the ones the vendor is recommending, to get me an improved cutoff frequency in a small package size on the analog supply pins (one of my questions for Eric might be around selecting the optimum value for the inductor, given that resistance value increases with the inductance value). Also, I no longer feel the need to put the same LC-filter on the digital power pins simply because that's what is in the data sheet. So really, this is another great video Robert. A cure for the over-engineered and perhaps counterproductive recommendations in data sheets.

[@arjayUU]

Very helpful. I didn't see much new value in the first 25 mins for just one slide with do much scribbling. The hands-on part in the end did it for me. Such things rrally get the point across. The theory is all over internet and universities anyways already.

[@yoah_an]

Thank you Robert and Eric!! This videos has solved many of my questions about LC filter and Ferrite Bead!

[@BrandonJJackson]

wow!!!! this was great!!!! THANK YOU ROBERT AND ERIC. Im new to EE, actually im a mechanical engineer but this example was just great to understand !!!!

[@NathanSweet]

How about how to choose L and C parameters?

[@mohammadhushki96]

Great content as usual. Thank you Robert

[@Mr_nuke99]

Just what I needed, really great explanation.

[@mixguru9669]

Ferrites are so lossy that q is very low if it forms an LC tank. So resonances are damped in PDNs.

[@barronisme]

Ferrites are inductive to begin with if you see its impedance vs frequency curve. Only for a certain frequency range it is high impedance and resistive. It then becomes capacitive again. The inductive portion creates the trouble!

[@Konecny_M]

Go measure it and you will see you are wrong in absolute most cases for PDNs, the resonant frequency peak happens so low in frequency that the core material loss usually does not have any chance to dampen it substantially.

[@mixguru9669]

Thanks for the informative replies.

[@VictorLarsen-fy9ls]

9:20 Ferrite Bead is not an inductor. Ferrite is a material that increases the inductance of a circuit section or inductor, the amount of magnetic energy that accumulates around it. Instead of ferrite, you can simply make a coil of the conductor in the same place, but then the resistance of the conductor will increase and most likely it will filter high frequencies worse without ferrite. The electrical resistance of ferrite is high and no currents flow through it. Ferrite is not part of the electrical circuit, it only affects the magnetic energy in it. What you use next in the example is not ferrite or ferrite bead, it is an inductance or coil with ferrite in a small case, that is part of an electrical circuit and conducts current. Ferrite Bead is put on a section of the conductor, often found on the leads of diodes and transistors.

[@sunebrian1423]

Totally agree, the basic modelling method is so wrong the in presentation. And the ferrite bead itself is a two way device that never clearly mentioned, which means this component is acting as a noise filter during high frequency and reject the noise coupling back to the source of the power rail. So it is clear that why beads are placed in each rail split and not just at the source.

[@sunebrian1423]

BTW, he mention ringing like that is not because the ferrite bead is not working but the circuit itself is so faulty and completely ridiculous. Where is the gate resistor and the drive circuit? A proper VLSI device on each IO circuit have so low g(tot) aka in fC scale. so the ringing itself coupled by the huge mosfet is completely not true but false to a point that a student could also pointed out. Meanwhile, why not simply replace the power rail of the AVR aka Arduino and give us a strong IO switching result but a breadboard external MOSFET circuit to demonstrate this unreal cases? I doubt his PHD is VLSI or any power analog field.

[@guytech7310]

The issue I see is that a high current draw as created on a IC voltage rail. In a real circuit high current switches are powered from separate power rails to isolate them. I typically use ferrite CLCR filters on IC voltage rails for precision voltage regulators typically used for use ADC ICs (reference voltage) which does remove noise. For really low noise applications I will use a separate power rail for the ADC as well as a precision reference voltage source. Another issue is that ceramic caps can pick up a lot of noise (mechincal vibration) & you need to use polymer caps instead of ceramic caps for very low noise filtering. In the example the ringing can be damped using a resistor on the load side, In this demo, I suspect the probes impedance was set to 1M ohm instead of 50 ohms, which would have damped the ferrite ringing. Try the same test but using a CLCR filter to damp the ringing. Ferrites are used to filter out high frequency noise, in this example the noise frequency was very low caused by a slow switching high current draw transistor switch.

[@mrkv4k]

The "noise frequency" wasn't slow, it was reasonably high (>20 MHz), it's all about the speed of that transition, not how many times per second you'd switch it.

[@guytech7310]

@@mrkv4k Ferrites are for noise above 100mhz. Also he use a CLC filter instead of CLCR filter which would have dampened the noise. There was no load on the CLC filter to dampen the noise. Try a test yourself with and without a resistor. You see what I am mean.

[@mrkv4k]

@@guytech7310 Ferrite is the ferromagnetic ceramic material, which is used for cores of all kinds of RF inductors. Ferrite BEADS are special kind of inductor, used for EMC supression. They can be used for noise supression from low MHz range. Sure, the datasheet will usually specify impedance at 100MHz, but that doesn't mean, they can't be used for lower frequencies. Inductance of an 100ohm@100MHz bead is somewhere around 160nH. Filter with this bead and 1uF capacitor will have cut-off frequency around 400kHz. Also, there was no load on purpose, because he was demonstrating the fact, that you need another component (in this case cap, but it could have been resistor) to finish the filter.

[@guytech7310]

@@mrkv4k Yes, I know, Thats why I commented, but several others didn't understand that it needs a load. He didn't try a CLCR filter, making novices believe Ferrites were useless. Kinda like explaining a devices doesn't work when its missing batteries, not not informing that the batteries are missing.

[@mrkv4k]

@@guytech7310 Did you watch the whole video? He showed that ferrites need capacitor, not that they are useless...

[@dannyboelen1801]

Thanks Robert and Eric.

[@TymexComputing]

Seeing prof Bogatin with a breadboard is really inspiring :) I have heard him recently in some promotional video of LeCroy but this is something better :)

[@87Spectr]

Hello, Robert! Thanks for video - very interesting and useful! But I have some problem with my English and therefore would you be so kind explain some things - 1. About experiment - is there ferrite bead or usual inductor? All topic about FB, but in experiment usual inductor? Is it correct? 2. In scheme with transistor - +9V, but further Eric says about +5V... What's PowerSupply measuring?

[@zaxir2]

I have exactly the same questions

[@mohammedabdeltawab9882]

it is a ferrite bead, usually used as it can keep its inductance at high frequency. normal inductor usually turn into caps in high frequency due to parasitic power supply voltage value will not matter +9 or +5 both are the same, just focus on concept

[@mrkv4k]

@@mohammedabdeltawab9882 Sorry, but that's just nonsense. "Normal" inductors exist in all kinds of packages and all kinds of values and there are plenty of "normal" inductors with resonant frequency well above few GHz. The bead is used because it's much cheaper than an equivalent precise inductor...

[@mohammedabdeltawab9882]

@@mrkv4k yes inductor are available in GHz but with limited maximum inductance value like few nan henry

[@mrkv4k]

@@mohammedabdeltawab9882 And how much do you think these beads have?

[@uploadJ]

Oh brother, a single bead in a line carrying DC? Have you checked the attenuation change versus DC current flowing through the wire? You might be too far to one side or the other of the BH curve to be effective at RF (radio) frequencies.

[@scottpelletier1370]

I wish that new hsde course was more accessible. I'm working with signals in the 35ps RT neighborhood and would take that course in a heartbeat.

[@jackbenyamin8927]

Use a tantalum capacitor in the LC filter and you will get much better filtering

[@DennisEspey]

Really enjoyed, very informative. Keep up the videos!

[@fedimakni1200]

You said we need to use ferrite beads in low current application, can you describe what's the scale of low current? 1 or 2 Amp could be considered as low current for some people.

[@ronpearson1912]

Hello Robert, does this boil down to impedance matching

[@leeslevin7602]

A very informative and interesting discussion thank you both!

[@windward2818]

Robert does some very good engineering, so I think he already understands what the answer is. Eric is one of the more understandable experts in the field of Signal and Power Integrity. At the level he usually teaches or is a guest speaker, he is expecting a certain education level of his participants (students) when giving talks. So, he will quickly move beyond the basics of circuit theory. But, I think Robert is interested in his problem solving process, as I am, and Robert is very thorough and methodical in general. Unfortunately, Eric never really starts at the raw basics. I can't speak for Eric, but I think I know what the expectations are in terms of theory. I will try and explain. On Eric's beginning slide he lists his group, High Speed Digital Engineering Group. Translated means Eric is concerned with electromagnetic field theory in high speed digital, which really just translates to field theory in general. So why does the slide show lumped parameter circuit elements (known as circuit theory to the electrical engineers) and not transmission lines? Eric is making a modeling assumption. He is asking the basic question can we use circuit theory instead of electromagnetic field theory in doing Power Integrity modeling? So, when is it proper to use a circuit diagram to model the system and not have to resort to field equations? There are three basic assumptions that have to be met in order to use circuit theory to model systems: 1) Electrical effects happen instantaneously throughout the system. This assumes electrical signals travel at or near the speed of light, which is indeed true. 2) The net charge on every component is the system is always zero. Capacitors even if charged have a net charge of zero. Capacitors essentially just separate charge. 3) There is no magnetic coupling between the separate components in a system. Transformers couple a changing magnetic field, but the coupling occurs within the transformer. The starting point is in meeting #1 above, with a rule of thumb. For the highest frequency of interest the physical size of the system must be less than 1/10 its wavelength. So, Eric is assuming a bandwidth of up to about 100 MHz in his power integrity study to model decoupling as a basic starting point. The wavelength is c/f = 300,000000/100MHz = 3 Meters. 1/10 of this is 30 cm. So the physical size of the circuit of interest must be 30 cm or less, which is 300 mm or 11.81 inches, and in the video you can see that his example circuit board is around 4 inches wide. So, yes we can use circuit theory to model decoupling, and Eric has drawn a circuit to describe the power distribution network (PDN), or if you like a power impedance diagram. If you do power supply design you already know the basics above and so we are interested in the overall output impedance of the supply as the basic input to the circuit. Then we have to transmit power to the components in the design (the integrated circuit die), which have their own impedance effects. When you add all of this up you get a diagram like Eric has provided.

[@WhiteBream]

TLDR reducing noise with LC filters. Lots of rambling, but nothing about ferrite beads... A ferrite bead is not just a low-current inductor. Also, ferrite beads exist for high currents, or to clamp around cables.

[@p_mouse8676]

Yeah, I was also pretty confused why they didn't mention any of the lossy behavior. Even SMD ferrites up to a couple of A exist. They work super well in gate drivers as well.

[@Konecny_M]

For practical comperhension of target audience here, it is better they understand it this way before they get nasty surprises from ignoring trhe fact that until relatively high frequency, they are indeed forming LC filter with high Q - the misconception of ferrite beads being magic bullet for filtering noise is very widespread. What I dislike is that importance of ESR of the capacitor included after the ferrite bead being skipped - the damping of the filter network is just as important and there are multiple approaches - signle MLCC capacitor not being one of them.

[@Mr.Leeroy]

@@Konecny_M it is easy to give f.beads bad reputation simplifying them to inductors only as everyone is basically doing on YT. But now show one video where proper aplication with engineering example is tought or presented

[@richardj163]

Great content, thank you for sharing!

[@deslomeslager]

My FM radio does no longer receive anything when my MPPT circuit is doing its work. I know it is a cheap thing, and maybe a buck/boost circuit will do this too. I have no clue how I can get rid of this (HF) noise. I tried putting the circuit in a metal box, but it does not help (and also makes it difficult to cool it with a fan unless the complete box is aluminum of course). A bead won't help I understand that. I am just not sure, where does the noise come from. If I understand this lesson well, it is the power lead. The plus and minus wires from my solar panel. Woud it help if I try to use coax cable (as long as it can deliver the 6 Amps at 30 Volt to pass through of course)? If that WOULD work, could I use just a small part of coax (in the room where I have the noise) or do I need to use coax over a big part of its length? I think the current wires are approx 8 meters inside the room where I have no reception, and then they disappear in the ceiling.

[@Putzerlblade]

Great video! What's the difference between a ferrite and an RF inductor? How to determine the right impedance of the ferrite?

[@mrkv4k]

"Ferrite bead" is an inductor typically used for EMC supression. The difference between ferrite *bead* and an RF inductor is in the important parameters. If you look at their respective datasheets, RF inductor will have it's inductance, tolerance of that inductance (typically less than 10%), Q factor, self resonant frequency and rated current. Ferrite bead will have only impedance at 100MHz (in ohms), tolerance (typically 20 - 40%), rated current and DC resistance. In the circuit, RF inductor is used for it's inductance, as a part of LC tank, filter and so on. Ferrite bead is pretty much just an added resistance that blocks unwanted RF.

[@arminleitner1097]

Thank you! As far as I know, ferrite beads are not ideal type inductors. I suppose they are for high frequency only. So I wonder if Eric uses a ferrite bead in his example with this low frequency?

[@voice4voicelessKrzysiek]

In this case, the repetition rate of pulses was low frequency but this fact is irrelevant because we were after the speed of the rising/falling edge of the pulse, which creates a ringing problem in the system. I suggest relistening to this excellent lecture.

[@martinmartinmartin2996]

why is the Ground shown in an IDEAL FORM? switching large LOAD current , di/dt, returns to the power source THROUGH THE GROUND PLANE ! INSTEAD the Load should be isolated : split the ground PC return from the smaller CONTROL ground PC return .

[@Enthusiastic_ECE]

Amazing content as usual Robert and Eric. How to select the L and C values? Can you please give me some reference on that?

[@vsiegel]

Around 30:00 or so I thought "just when I thought the screen can not get more cluttered..." oops! But then, around 36:00 the skype call apology and a skype call come in... Then I knew: This is art!

[@RobertFeranec]

:)

[@robertpeters9438]

You need to use a larger cursor when moving about so rapidly!

[@TheCuteworm]

this is so cool. best class i ever had

[@SaeligCoInc]

Very helpful!

[@rul1175]

Awesome video Robert. I wish the video were 3 hours long🤓

[@霆大]

Hello, I would like to ask you a question. Regarding high-speed differential pairs, can we refer to the power layer? Will it have a bad impact? Thank you.

[@fedimakni1200]

What's the difference between an LC filter with an inductor and LC filter with a ferrite bead?

[@mrkv4k]

Ferrite bead is an inductor. As a component, it's a cheap inductor with high tolerance. You'd use it in situations, where you don't care about the precise cut-off frequency - as long as it's decade below the frequency of your noise. LC filter with a ferrite bead is used when you have a DC signal and you want to separate an unwanted RF component.

[@drkzilla]

Keep doing the long videos they are great 🥰🥰🥰

[@ChrisSmith-tc4df]

Some LDO’s offer a high PSSR up to 1 to 2 MHz

[@sparklee6994]

this is not a engineering LC filter, without knowing the value of L and C...the experiment should go further by trying different value pair of ferrite and capacitor vs resulting waveforms...than people will know how to select different value of ferrite and capacitor pair to get better filtering at specific frequency.

[@viniciusbalincorrea3510]

Great video! Thanks

[@madhavdimble9782]

Thank you. For making videos on such topics. I have one question. Some times differnet grounds are also separated using ferrite beads. Is it good practise? Because some design guides suggest it and some do not suggest it.

[@svenk.5308]

Separating grounds by any means is in general not a good idea. This will cause extra noise on signals transitioning from one reference ground to the other one. If no signals with ground reference are crossing from one ground to the other, this might help to improve EMC.

[@madhavdimble9782]

​​@@svenk.5308Thank you for reply

[@RobertFeranec]

I have had two bad experiences when separating GND: when separated by BEAD , we had ESD problems, when separated by one 0R, we had EMC problems.

[@madhavdimble9782]

​@@RobertFeranecThank you very much Robert.😊

[@wombatillo]

@@RobertFeranec Some datasheets for chips with USB peripherals suggest that the USB shield-ground should be connected to the device ground through a ferrite bead, some that it should be connected through a 1 nF 400V or 600V tolerant capacitor and some with this cap and a large 100k - 1M resistor in parallel. It's confusing and seems like things are communicated extremely badly or worse: some of the authors don't really know what they're doing.

[@y_x2]

You need to use a high speed Differentiel probe to see these problem correctly!

[@ddognine]

I was thinking that too. The actual output looked a little suspect.

[@nicelifekavi]

Nice, thank you.

[@shuashuashua1]

Thanks for next great video

[@avichalid5604]

Thanks for a great video. Can you share the schematics of the LC test board?

[@lasersbee]

Only if needed....

[@pochuanpeng1729]

Thanks.

[@davey2k12]

30min for a practical example 😂😂😂 should of started with that

[@ZhuraYuki]

Robert, you need to learn to explain things in 10 min instead of 40. It is possible with right amount of knowledge. This way more people will learn from your materials.

[@AlbertRei3424]

People need to learn to invest themselves and their time to get better, otherwise they should do python coding instead of engineering

[@ThePetaaaaa]

Shots fired! Honestly, I like both. The short for the TL;DR and the long version to understand the complexity better.

[@ZhuraYuki]

@@AlbertRei3424 Correct, so it would be great to invest 10min, because 40 min does not fit into modern time bugdet.

[@THEELECTRICGUY]

I like the longer version. We as humans need time to process information. A 10 min video is good for engagement but to grasp all the presented information, it needs to be presented steadily and slowly. And thats why a longer video is more suitable. And also, thats how we used to learn in schools and colleges. Imagine a teacher finishing a chapter within 10mins with all information! Will you be able to grasp all of it? :)

[@AlbertRei3424]

@@THEELECTRICGUY let it go,it's a python guy

[@donerskine7935]

Should have edited out that Skype call and done a retake. Otherwise, an interesting video.

[@aeonikus1]

Please mark your videos as not for kids, because those for kids, due to YT terms of use and some laws, cannot be used to identify watching person so it's impossible to SAVE such videos. There is no downside fro you to it, but is's a nuisance for your viewers. Thanks. BTW. Great content, I love what you do to spread knowledge.

[@Mr.Leeroy]

That was a terrible demo, that arduino one. Crafted and misleading. Eric is showing a giant TO220 mosfet that most likely has a lot of gate capacitance that has to be not only slowed down with a gate resistor, but also decoupled with a relatively large capacitance as close to gate switch and terminal as possible. This is the whole purpose of intergrated gate drivers. And what is essentially a result of this absence of mandatory components elswhere in schematic he calls a power rail noise, which it is not. It's a lower frequency relatively slow dip. What he got after adding bead and more importantly capacitance that is a high speed spiking noise that ferrite beads should be dealing with.

[@biswajit681]

Entire video is quite confusing and repeating same things again and again without adding much information...its really hard to understand what he wants to convey.

[@AbdullahKahramanPhD]

Yeah, take that Altium Academy 😅😂😂😂 jk

[@stuntaudio]

First 🎉 !

[@TheOleHermit]

Been developing a PCB that is basically a compilation of others, which have a ferrite resistor. Monkey see, monkey can do. Now I understand what it's for. Thanks to Robert & Eric, this ape descendant has a better understanding of what I'm doing and why.

[@andrewFJ]

Same here, can you oleaae boil down videos to 10-15 minutes and reduce unnecessary talk. Some people trying to learn a lot of material quick, i don't have all day watching several videos for 40min each, but i can find time for 4-5 videos of 5-10min.