This is the best video about ferrite beads on UA-cam I've seen so far. In most videos, the inductive characteristics are not mentioned or dismissed as insignificant.
Excellent video. Ferrite beads are definitely more of an art than a science at this point. The scariest part is that not only does each manufacture have its own secret mixes for making their different ceramic types, but the ceramic is made in batches, and their isn't any real assurance on how closely the batches will match up in terms of the performance characteristics we are interested in. So even if you experiment and find something that works really well, it might not work as well over a period of time in production...
Manufacturers like Transphorm and USci suggest to use a ferrite bead in the gate drive, but your lecture makes its very clear that its not so easy to fight the spikes and in fact may worsen it! Thank you very much for the Lecture Dr. Ben-Yaakov.
This is by far the best explanation of ferrite bead principles that I've been able to find. It seems to be incredibly difficult to get a good theoretical overview of how these things work. I'm beginning to suspect that about 95% of people who put up information on this topic really don't have a good grasp of the topic themselves. Thank you. You've clarified the matter hugely. Do you have any videos that cover higher power ferrite bead applications, for motors and the like? I suspect that is another specialism on its own. [BTW, your two minute overview of noise induced by ground loops would have made the video worth watching on its own. That was a model of clarity]
@@sambenyaakov I was thinking about high power applications, such as the guy demonstrated (practically, at least) in this video: ua-cam.com/video/DJfiOqaeFDg/v-deo.html
@@scollyer.tuition This video if full of errors, too many to count. To the point: he is showing the use of ferrite toroids. These are not the ferrite beads I am talking about i my video. The true ferrite beads include a substantial lossy part. The ferrite toroids used in thr video you referenced are a poor boy substitute common mode chokes. The problem of using FB in high power applications is that they saturate.
Amazing video.Thank you proffesor.You had tell very simple this subject .Because you knowlodge very high. What had say Einstein: "If you can't explain it simply, you don't understand it well enough
Hi Prof. Sam, I'd like to thank you for an insightful video. A few questions which popped while watching - 1. since Rac is basically core losses shouldn't we perceive the added Rac in parallel to the inductance part? I've saw that you disputed the "simple model" by TDK for example. 2. In Clamp bead, I've wondered if the MFR's are taking into account any misalignment in the clamp which might cause an air gap. Does the design takes into consideration that the magnetic path is solid or gapped? - I assume it's the former (just checking). If possible at all - can you help by sharing relevant publications
1. It is not that I am against the lumped equivalent circuits,.They are good for emulating the impedance but, as I have pointed out, I think that the loose the phase information which in most cases is not that important. 2. The clamped bead are assumed to be with practically no gap,. The gap between the bead and wire does not matter (Ampere Law)
Am I right in saying that the reason for the example on the right at 19:55 being optimal is that the fundamental frequency of the falling edge is 5MHz, which matches the cross point on the ferrite bead's ZRX graph? (I estimated 5MHz from the fall time on the left graph appearing to be approximately 100ns, and a sine wave with a 100ns period between its peaks would have a total period of 200ns, i.e. 5MHz)
Thank you for posting such an good informative video. Does this mean that when IC manufactures recommend to put ferrites chips on nets powering the high- speed serdes (to get a clean power), they are actually degrading the serdes performance because the ferrite chip power filter hinders the serdes circuits to draw power rapidly?
Yes, I think there is that one guy (i do not know who he is ) that waits until I publish a video and quickly, within a couple of hours or so, clicks the dislike.Go figure.
Dear Dr.Sam, I really appreciate your efforts in making such an elobarate educational videos on current edge technology to help the upcoming students. However, I am completely confused and surprised to see a ferrite bead in the path of a gate driver on slide 3, 1:36, according to Transform company. It is highly desirable to keep length of gate tracks shorter and wider to maintain as minimum inductance as possible in the gate driver path. Higher inductance in gate path offers higher impedance to high frequency gate signal. Using same reason, how can we endorse a ferrite bead in gate driver path, which is also an inductor basically. (Its frequency response also proves it).
Thanks. You have a good point. I am not sure myself aout FB in gate driver. I tried it and did not se any benefit. But: THe FB might perhaps be useful cases of potential oscillations at very high frequencies 100's of MHz at which the bead is resistive while the added impedance at lower frequencies can be tolerated. i.e. wL/R is smaller than 1 .
The thing I don't understand is, isn't inductive reactance essentially frequency dependent resistance? Or is that precisely what you are saying? That the reactance is welcomed and the DC element of inductance is the parasitic part... in this application
The bead has real power loss which increases with frequency. This can be modeled as a frequency dependent resistor. Reactance does not dissipate real power.
@@sambenyaakov Thank you for the explanation. I was not aware of that distinction. I am studying alone, no schooling. Your response was helpful and much appreciated.
Hi Sam. The video is very nice. Congratulations. I have a question about the use of ferrite beads on both ends of the cable and when it is applicable. Sometimes people use this approach to minimize EMI effects on RF systems in the proximity of the cable (I think so)...
Hi Dr. Sam, Ideally, we would want a ferrite bead that doesn't have an inductance component. Just the resistance at high frequency. Correct? If yes, does that mean we should choose the lowest inductance over frequency ferrite bead every time?
8:40 Earlier comment asked about your phase-accurate models. You said it would be public-domain. Can you provide any update???? I have an immediate need. At least, some pointers on your approach & methods (pspice or whatever). THANKS !!
Hi, I plan to prepare a video on that when time available😊. It short, one can use ELAPLACE behavioral model to create a non linear frequency dependent resistor and then add it to a non linear inductor
Very useful information. In example 4 you say that you use a model, which you developed. Could you, please, explain which model you use for simulation in time domain and how the parameters for this model can be obtained from datasheet graphs Thanks
Hi Lev, Thanks for the comment. You have a keen eye. The simulation models are still in the making. Once published, I will make it public domain. Hold on😊
Excellent video as always :-) I have one question: Are all of these impedance charts for ferrite beads given for one wire going through the bead? What if I put several loops through the toroid?
...I meant what happens to the resistive part. The inductance will roughly (not exactly because it is nonlinear) increase as n^2 but the resistive part I think should behave differently. also where will the resonance (peak) move to when I have several turns?
How do you account for linear regulators that require low ESR capacitors at their inputs? I found a paper from Linear Technology that details using ferrite beads at both the input and output reduces switching noise, but this renders useless the use of any damping resistors to limit high frequency spikes from the ferrite beads.
16:34 to 17:00 ... You state the crossover R/X is the key factor. Why? Because lower freq of crossover is better, for higher freq of interest, so L resonance w/ parasitic C is pushed to higher freq. where R is higher (lower Q)? If so, it seems that would make the lower crossover freq a bad choice for filtering dc/dc converter ripple (lower freq).
Ferrite beads are not meant to replace filter inductors. They are meant to add damping (lower the Q). So if the inductive part dominates it is defeating its purpose.
We need a double thumbs up button.
Great channel 👍👍
😊👍🙏
Namaste Sam Ben Yakov Sir. Great explanation on ferrite beads.
Greetings from India.
👍🙂🙏
This is the best video about ferrite beads on UA-cam I've seen so far.
In most videos, the inductive characteristics are not mentioned or dismissed as insignificant.
Thanks for comment.
Excellent video. Ferrite beads are definitely more of an art than a science at this point. The scariest part is that not only does each manufacture have its own secret mixes for making their different ceramic types, but the ceramic is made in batches, and their isn't any real assurance on how closely the batches will match up in terms of the performance characteristics we are interested in. So even if you experiment and find something that works really well, it might not work as well over a period of time in production...
Well said.If not a member please join www.linkedin.com/groups/13606756/ .
request sent.
Thanks for this video professor. Best regards from Chile 🇨🇱
Thanks and greeting from Israel. Never been to Chile. The closest I got is Brazil.
Manufacturers like Transphorm and USci suggest to use a ferrite bead in the gate drive, but your lecture makes its very clear that its not so easy to fight the spikes and in fact may worsen it! Thank you very much for the Lecture Dr. Ben-Yaakov.
Good points. Thanks.
This is by far the best explanation of ferrite bead principles that I've been able to find. It seems to be incredibly difficult to get a good theoretical overview of how these things work. I'm beginning to suspect that about 95% of people who put up information on this topic really don't have a good grasp of the topic themselves.
Thank you. You've clarified the matter hugely.
Do you have any videos that cover higher power ferrite bead applications, for motors and the like? I suspect that is another specialism on its own. [BTW, your two minute overview of noise induced by ground loops would have made the video worth watching on its own. That was a model of clarity]
Thanks. No no such videos. AS far as I know ferrite beads have not been used in motor control.
@@sambenyaakov I was thinking about high power applications, such as the guy demonstrated (practically, at least) in this video:
ua-cam.com/video/DJfiOqaeFDg/v-deo.html
@@scollyer.tuition This video if full of errors, too many to count. To the point: he is showing the use of ferrite toroids. These are not the ferrite beads I am talking about i my video. The true ferrite beads include a substantial lossy part. The ferrite toroids used in thr video you referenced are a poor boy substitute common mode chokes. The problem of using FB in high power applications is that they saturate.
Amazing video.Thank you proffesor.You had tell very simple this subject .Because you knowlodge very high.
What had say Einstein:
"If you can't explain it simply, you don't understand it well enough
😊🙏
Thank you for this and for your other videos professor.
Thanks 😊. If not a member please join www.linkedin.com/groups/13606756/ .
Thanks. It was very informative for even field designer.
Thanks
Excellent video Professor Sam.
Many thanks!
Hi Prof. Sam, I'd like to thank you for an insightful video. A few questions which popped while watching -
1. since Rac is basically core losses shouldn't we perceive the added Rac in parallel to the inductance part?
I've saw that you disputed the "simple model" by TDK for example.
2. In Clamp bead, I've wondered if the MFR's are taking into account any misalignment in the clamp which might cause an air gap.
Does the design takes into consideration that the magnetic path is solid or gapped? - I assume it's the former (just checking).
If possible at all - can you help by sharing relevant publications
1. It is not that I am against the lumped equivalent circuits,.They are good for emulating the impedance but, as I have pointed out, I think that the loose the phase information which in most cases is not that important.
2. The clamped bead are assumed to be with practically no gap,. The gap between the bead and wire does not matter (Ampere Law)
Am I right in saying that the reason for the example on the right at 19:55 being optimal is that the fundamental frequency of the falling edge is 5MHz, which matches the cross point on the ferrite bead's ZRX graph? (I estimated 5MHz from the fall time on the left graph appearing to be approximately 100ns, and a sine wave with a 100ns period between its peaks would have a total period of 200ns, i.e. 5MHz)
It is optimal in that it is more a resistive than inductive at a lower frequency. Ideally we would like to get rid of XL
Great lecture Dr. Sam
Thanks. How have you been?
Doing good Dr. Sam. Hope you are doing good as well.
Thank you for posting such an good informative video.
Does this mean that when IC manufactures recommend to put ferrites chips on nets powering the high- speed serdes (to get a clean power), they are actually degrading the serdes performance because the ferrite chip power filter hinders the serdes circuits to draw power rapidly?
Not if the bead are selected wisely
Great job, thank you!
Thank you too!
Excellent explanation... I am curious about that one dislike....He would possibly be searching for a necklace of beads....!
Yes, I think there is that one guy (i do not know who he is ) that waits until I publish a video and quickly, within a couple of hours or so, clicks the dislike.Go figure.
Great Explanation.
Thanks
Dear Dr.Sam,
I really appreciate your efforts in making such an elobarate educational videos on current edge technology to help the upcoming students. However, I am completely confused and surprised to see a ferrite bead in the path of a gate driver on slide 3, 1:36, according to Transform company.
It is highly desirable to keep length of gate tracks shorter and wider to maintain as minimum inductance as possible in the gate driver path. Higher inductance in gate path offers higher impedance to high frequency gate signal. Using same reason, how can we endorse a ferrite bead in gate driver path, which is also an inductor basically. (Its frequency response also proves it).
Thanks. You have a good point. I am not sure myself aout FB in gate driver. I tried it and did not se any benefit. But: THe FB might perhaps be useful cases of potential oscillations at very high frequencies 100's of MHz at which the bead is resistive while the added impedance at lower frequencies can be tolerated. i.e. wL/R is smaller than 1 .
The thing I don't understand is, isn't inductive reactance essentially frequency dependent resistance? Or is that precisely what you are saying? That the reactance is welcomed and the DC element of inductance is the parasitic part... in this application
The bead has real power loss which increases with frequency. This can be modeled as a frequency dependent resistor. Reactance does not dissipate real power.
@@sambenyaakov Thank you for the explanation. I was not aware of that distinction. I am studying alone, no schooling. Your response was helpful and much appreciated.
Excellent tutorial
Thanks
Hi Sam. The video is very nice. Congratulations. I have a question about the use of ferrite beads on both ends of the cable and when it is applicable. Sometimes people use this approach to minimize EMI effects on RF systems in the proximity of the cable (I think so)...
Ferrite beads on a cable act like a common mode filter
Hi Dr. Sam,
Ideally, we would want a ferrite bead that doesn't have an inductance component. Just the resistance at high frequency. Correct? If yes, does that mean we should choose the lowest inductance over frequency ferrite bead every time?
Yes, but a low inductance bead will also have a low resistance. Try to get a bead with the highest R/L at the relevant frequency range.
Thank you very much.!
Thanks 😊. If not a member please join www.linkedin.com/groups/13606756/ .
8:40 Earlier comment asked about your phase-accurate models. You said it would be public-domain. Can you provide any update???? I have an immediate need. At least, some pointers on your approach & methods (pspice or whatever). THANKS !!
Hi, I plan to prepare a video on that when time available😊. It short, one can use ELAPLACE behavioral model to create a non linear frequency dependent resistor and then add it to a non linear inductor
Beautiful
Thanks 😊. If not a member please join www.linkedin.com/groups/13606756/ .
Very useful information. In example 4 you say that you use a model, which you developed. Could you, please, explain which model you use for simulation in time domain and how the parameters for this model can be obtained from datasheet graphs
Thanks
Hi Lev, Thanks for the comment.
You have a keen eye. The simulation models are still in the making. Once published, I will make it public domain. Hold on😊
Can you share the ferrite bead you have designed using equation at @26:25 which mimics the R,X and Z of the ferrite bead?
Sorry, I have no record.
Excellent video as always :-) I have one question: Are all of these impedance charts for ferrite beads given for one wire going through the bead? What if I put several loops through the toroid?
...I meant what happens to the resistive part. The inductance will roughly (not exactly because it is nonlinear) increase as n^2 but the resistive part I think should behave differently. also where will the resonance (peak) move to when I have several turns?
Manufacturers show measurement of Z for a number of terms and the impedance behaves as n^2 so both the inductive and resistive parts increase.
Thanks for the video, could you make a video about CM or DM chokes?
Will try.
How do you account for linear regulators that require low ESR capacitors at their inputs? I found a paper from Linear Technology that details using ferrite beads at both the input and output reduces switching noise, but this renders useless the use of any damping resistors to limit high frequency spikes from the ferrite beads.
Ferrite beads are not the ultimate fixer, but could be very useful if used wisely.
16:34 to 17:00 ... You state the crossover R/X is the key factor. Why? Because lower freq of crossover is better, for higher freq of interest, so L resonance w/ parasitic C is pushed to higher freq. where R is higher (lower Q)? If so, it seems that would make the lower crossover freq a bad choice for filtering dc/dc converter ripple (lower freq).
Ferrite beads are not meant to replace filter inductors. They are meant to add damping (lower the Q). So if the inductive part dominates it is defeating its purpose.
@@sambenyaakov I'll take that as either a "yes" or "you emitted an unclear question". Either way, your answer suffices. Thanks. :-))