I am so happy to see someone perform this demonstration with modest test equipment. It's great to see that modest test equipment can be used thoughtfully for meaningful troubleshooting and diagnostic work. Thanks!
A 6 hour video would be a bit much as she mentioned, but a series of say 20-30 minute case studies could be very interesting. Take a single scenario that has been experienced and resolved, similar to the bus conversion story she told, but go into a little more detail on it. Describe the scenario and what was being experienced, show what is happening via a few slides showing the theory and then discuss the solutions tried and which worked or didn’t work as the case may be. This could be board level issues, box level issues, cable issues, system issues, etc.
I really would be interested in a follow up about the low frequency audio example. There seems to be lots of information about the high frequency shielding but not very much about low frequency noise and the associated shielding and filtering. Great video I look forward to any follow ups.
Just as a tip when buying those tinySA units (which are really great by the way, there are several versions). Do double check your seller is listed as an official reseller. There are lots and lots of fake ones being sold for nearly the same price and sometimes even more expensive compared to official resellers. When buying the 'fake' one, you may end up buying it twice. I learned that the hard way. One of the symptoms of having a fake one may be that after upgrading the official firmware the self calibration and self test fails.
She is so amazing at explaining these complex issues. Oh, to have her as a teacher... ( Sigh ... ). I'm just a hobbyist designing things with communication frequencies that rarely exceed 1 MHz, but I learned so much today. Thank you so much for sharing this Mrs. Burnham and Mr. Feranec !.
The part mentioned around 31:03 is one of my favorites in fields where antennas are used. If there isn't sufficient decoupling between an antenna and its coax, it can very easily happen that the outside of the outer conductor starts acting as a part of the antenna. A perfect example of this would be with a simple dipole antenna, where one side is connected directly to the inner conductor of the coaxial connector, and the other side is connected directly to the outer conductor of a coaxial connector.
49:10 what is that picture from EEseal, I can't understand well, those pads are capacitor? How?! Thank you so much. It could be nice if show schmatic of termination of cables with capcitors like in RJ45 and explaining them.
❤ ooh, thank you so much Robert and Karen! I really appreciate this video that explains when to sheild/not to shield cables, and it really helps us in designing and planning layouts of electrical systems. I am dealing with large voltage 100v, large current 400 A, it causes alot of weird noises. This video can be a guide.❤
I have a question in regards to the experiment shown in the first half of the video: What exactly has changed in terms of topology when the big aluminium foil was connected to the shield at both ends? Isn't this just basically connecting a giant lower impedance parallel to the cable shield (at least that's what I'm seeing)? If so, why does it change the results so much? I don't see this changes the termination in any way, right?
Great question, i have the same question , i think that was because the signal generator was in plastic enclosure, that's why she placed the tape close to the signal generator so that both ends of the cable are gnded properly.
@@theIpatix I was wondering much the same, I wish they had discussed that point of the demonstration more. I’m guessing that the function generator is isolated from ground at the power receptacle via an isolated switching power supply, causing the system to float. I’m also guessing that by attaching the shield to the aluminum foil “ground plane” at both ends, there are two parallel paths with different impedances which allows the noise picked up on the cable shield (induced from the 20 MHz and 60 MHz harmonic "signal") to return via the “ground plane” effectively shorting it. Just guessing though.
@@theIpatix the foil is not floating, it is clearly connected to the aluminum box which load is inside. I still think that if the signal generator had a metal chasis the foil was not required. Moreover, as you can see the small spectrum analyzer still shows the radiated noise from the signal generator enclosure. Of course, I am not sure about it. 😅
Thank you for covering this topic. I would indeed like to have a 6-hour video available. I could easily watch it at my own pace over the course of a week or so. Also, I would like to hear recommendations on where to attach (either to signal ground or the chassis) PCB-mounted ESD and TVS protection diodes.
I missed whether the foil on the table is connected to actual ground or "ground point" means "conductor with common base potential shared between the generator and the load"?
I don't understand the ground plane used in this video. Is it just a foil of metal floating on the table or is it connected to circuit GND via the function generator's connector? Is the metal chassis of the function generator connected to circuit GND?
That shielded cable in the demonstration- is that an insulated wire surrounded by a braided conductor shield? What is the return path? Where is the resistor load connected to?
In the example the noise on the cable was being measured by a current clamp. In that case it seems like only common-mode noise could be measured by that clamp, since differential noise would cancel out in the clamp by including both wires inside it. But there's no discussion of common-mode chokes to kill that noise, did I miss something or are common-mode chokes not useful to address this?
Hello Robert, Regarding minute 23 of this video, when she says that to reduce the impedance by reducing the voltage in the internal conductors...if I twist the internal wires, which is what happens in twisted pair, will the voltage also decrease or do I need to reduce the absolute voltage level from the source?
Connecting the shield to the "gnd plane" on the generator side did not improve the shielding performance. It just directed the displacement current which was inflicted from the capacitive coupling between the function generator and the screen current to the foil instead. There would be no difference if the function generator had a proper design. I'd also say that plastic housings and shielded cables are not the problem even for USB, if you design you PCB to do the shielding. And even if the IC are radiating .. one can put a shield on them or ground the heatsink. One interesting question for this setup would be though if the aluminium foil is connected to PE or what would happen if you put a ferrite on the power cable side of the function generator.
I have multiple boards, each one in a metal case, interconnected by shielded cables terminated on the cases (perfect faraday cage system). The question is: How to connect each board ground (0 V reference of circuits) to this faraday cage?
Something that is so hard to get proper answer: Assuming device has metal chassis which has dedicated earth ground connection - How should the device metal chassis/shield/protective ground connect to the device ground? 1) Directly with low impedance 2) capacitor and 1M ohm in parallel 3) capacitor with low ohm value resistor in parallel 4) no connection at all 5) something else?
Luckily, this video contains no revelations. You may continue shielding of your RF cables and connect the shield on both sides, and it will be great. Of course, if you are able to put your whole setup in a large metal box - it would be perfect. Or you may wrap it around by a metal foil in the worst case. ) That's all. Good luck to everyone who is not designing weapons to kill or "protect" people!
I think the same buddy, too much electronics is used for military to "protect" people by bombing civilians; its a shame for the world and EE themselves.
I am so happy to see someone perform this demonstration with modest test equipment. It's great to see that modest test equipment can be used thoughtfully for meaningful troubleshooting and diagnostic work. Thanks!
A 6 hour video would be a bit much as she mentioned, but a series of say 20-30 minute case studies could be very interesting. Take a single scenario that has been experienced and resolved, similar to the bus conversion story she told, but go into a little more detail on it. Describe the scenario and what was being experienced, show what is happening via a few slides showing the theory and then discuss the solutions tried and which worked or didn’t work as the case may be. This could be board level issues, box level issues, cable issues, system issues, etc.
I really would be interested in a follow up about the low frequency audio example. There seems to be lots of information about the high frequency shielding but not very much about low frequency noise and the associated shielding and filtering. Great video I look forward to any follow ups.
Just as a tip when buying those tinySA units (which are really great by the way, there are several versions).
Do double check your seller is listed as an official reseller.
There are lots and lots of fake ones being sold for nearly the same price and sometimes even more expensive compared to official resellers.
When buying the 'fake' one, you may end up buying it twice. I learned that the hard way.
One of the symptoms of having a fake one may be that after upgrading the official firmware the self calibration and self test fails.
She is so amazing at explaining these complex issues. Oh, to have her as a teacher... ( Sigh ... ). I'm just a hobbyist designing things with communication frequencies that rarely exceed 1 MHz, but I learned so much today. Thank you so much for sharing this Mrs. Burnham and Mr. Feranec !.
The part mentioned around 31:03 is one of my favorites in fields where antennas are used. If there isn't sufficient decoupling between an antenna and its coax, it can very easily happen that the outside of the outer conductor starts acting as a part of the antenna.
A perfect example of this would be with a simple dipole antenna, where one side is connected directly to the inner conductor of the coaxial connector, and the other side is connected directly to the outer conductor of a coaxial connector.
Thank you both.
49:10 what is that picture from EEseal, I can't understand well, those pads are capacitor? How?!
Thank you so much.
It could be nice if show schmatic of termination of cables with capcitors like in RJ45 and explaining them.
❤ ooh, thank you so much Robert and Karen!
I really appreciate this video that explains when to sheild/not to shield cables, and it really helps us in designing and planning layouts of electrical systems.
I am dealing with large voltage 100v, large current 400 A, it causes alot of weird noises. This video can be a guide.❤
I have a question in regards to the experiment shown in the first half of the video: What exactly has changed in terms of topology when the big aluminium foil was connected to the shield at both ends? Isn't this just basically connecting a giant lower impedance parallel to the cable shield (at least that's what I'm seeing)? If so, why does it change the results so much? I don't see this changes the termination in any way, right?
Great question, i have the same question , i think that was because the signal generator was in plastic enclosure, that's why she placed the tape close to the signal generator so that both ends of the cable are gnded properly.
@peiman1988 Thing is, what does it actually ground? To me it looks like the aluminium foil is floating apart from the two ends.
@@theIpatix I was wondering much the same, I wish they had discussed that point of the demonstration more. I’m guessing that the function generator is isolated from ground at the power receptacle via an isolated switching power supply, causing the system to float. I’m also guessing that by attaching the shield to the aluminum foil “ground plane” at both ends, there are two parallel paths with different impedances which allows the noise picked up on the cable shield (induced from the 20 MHz and 60 MHz harmonic "signal") to return via the “ground plane” effectively shorting it. Just guessing though.
There is a video on this channel from 4 years ago, which has "Return Current Path" in it's name. I think the answer is there.
@@theIpatix the foil is not floating, it is clearly connected to the aluminum box which load is inside. I still think that if the signal generator had a metal chasis the foil was not required.
Moreover, as you can see the small spectrum analyzer still shows the radiated noise from the signal generator enclosure.
Of course, I am not sure about it. 😅
Thank you for covering this topic. I would indeed like to have a 6-hour video available. I could easily watch it at my own pace over the course of a week or so. Also, I would like to hear recommendations on where to attach (either to signal ground or the chassis) PCB-mounted ESD and TVS protection diodes.
I missed whether the foil on the table is connected to actual ground or "ground point" means "conductor with common base potential shared between the generator and the load"?
I don't understand the ground plane used in this video. Is it just a foil of metal floating on the table or is it connected to circuit GND via the function generator's connector?
Is the metal chassis of the function generator connected to circuit GND?
At 60 MHz, a wavelength is 5 meters long. The length of the cable looks like approximately a quarter wavelength...
That shielded cable in the demonstration- is that an insulated wire surrounded by a braided conductor shield?
What is the return path? Where is the resistor load connected to?
me me i want to see the six hour youtube video!!
Yes, indeed, especially with a timestamp "table of contents."
In the example the noise on the cable was being measured by a current clamp. In that case it seems like only common-mode noise could be measured by that clamp, since differential noise would cancel out in the clamp by including both wires inside it. But there's no discussion of common-mode chokes to kill that noise, did I miss something or are common-mode chokes not useful to address this?
Hello Robert,
Regarding minute 23 of this video, when she says that to reduce the impedance by reducing the voltage in the internal conductors...if I twist the internal wires, which is what happens in twisted pair, will the voltage also decrease or do I need to reduce the absolute voltage level from the source?
I would love to see more Videos with her, lots of interesting stuff.
Grounding of shielded ethernet networks would be nice to know more about.
Connecting the shield to the "gnd plane" on the generator side did not improve the shielding performance. It just directed the displacement current which was inflicted from the capacitive coupling between the function generator and the screen current to the foil instead. There would be no difference if the function generator had a proper design.
I'd also say that plastic housings and shielded cables are not the problem even for USB, if you design you PCB to do the shielding. And even if the IC are radiating .. one can put a shield on them or ground the heatsink.
One interesting question for this setup would be though if the aluminium foil is connected to PE or what would happen if you put a ferrite on the power cable side of the function generator.
I have multiple boards, each one in a metal case, interconnected by shielded cables terminated on the cases (perfect faraday cage system). The question is: How to connect each board ground (0 V reference of circuits) to this faraday cage?
Urrah! New interesting video!
Something that is so hard to get proper answer: Assuming device has metal chassis which has dedicated earth ground connection - How should the device metal chassis/shield/protective ground connect to the device ground? 1) Directly with low impedance 2) capacitor and 1M ohm in parallel 3) capacitor with low ohm value resistor in parallel 4) no connection at all 5) something else?
Luckily, this video contains no revelations.
You may continue shielding of your RF cables and connect the shield on both sides, and it will be great.
Of course, if you are able to put your whole setup in a large metal box - it would be perfect.
Or you may wrap it around by a metal foil in the worst case. )
That's all.
Good luck to everyone who is not designing weapons to kill or "protect" people!
I think the same buddy, too much electronics is used for military to "protect" people by bombing civilians; its a shame for the world and EE themselves.
Surprisingly, the cable shield from the current sensor does not accept FM. 🙃
What is price of frequency generator and oscilloscope sir.?