Hi,

I am doing a research on a sensitive analog circuit project. I havent got much experience, but I know that using an EMI shield is a considerable move (it would not compensate for the poor layout, but presumably improve the circuit.) I noticed that some EMI shields are solid and some of them have holes inside. Is there any theory behind the hole diameter and wavelength ratio, or is it just reducing the material consumption in large production quantities? Is there any difference in the functionality. Does EMI shield thickness play any role in the shielding quality? Thank you in advance.

 

For the most part, yes, yes, and yes.

If the holes in the shield are small compared to the wavelength, then the shield is almost completely intact. Consider the mesh screen in the window of a microwave oven. The wavelength if about 125 mm and the holes are a few millimeters. There's a rule of thumb, but I don't remember what it is. Something like 10:1 or 20:1.

The thickness matters up to a certain degree, related to the skin-depth for EM waves at that frequency in that material. Beyond that thickness, it has almost no additional effect.

 

It's probably easier to look at it from the other side; if you're building a project, what specific frequencies do you need to keep in or out? Your sensitive circuits should also have filtering to filter out frequencies you're not interested in.

 

Hi,

I am doing a research on a sensitive analog circuit project. I havent got much experience, but I know that using an EMI shield is a considerable move (it would not compensate for the poor layout, but presumably improve the circuit.) I noticed that some EMI shields are solid and some of them have holes inside. Is there any theory behind the hole diameter and wavelength ratio, or is it just reducing the material consumption in large production quantities? Is there any difference in the functionality. Does EMI shield thickness play any role in the shielding quality? Thank you in advance.

https://www.farnell.com/datasheets/1965028.pdf

 

A word of caution about what has bitten others: Aluminum and copper both make good shields against electric fields, but neither does much to avoid magnetic coupling. Coils in aluminum cans can couple to other coils in other aluminum cans.

 

RE: Bill.--> As yes as no. Indeed the anti-magnetic metals cannot play the significant role alone, them can be used extra efficiently when combined with ferrous materials. Sorry I cannot appreciate the precise construction before the patent is issued to the name of inventor to whom I made the expertise, but there exists a highly effective construction using the bunch of perforated steel-plates, aluminum foliage and paper sheets to create the shield making a Faraday ecrane simultaneously with magnetic ecrane using no any mju-metall. One side of this wonder-deed I put the 4 Tesla super-magnet creating about 40 kg force to the piece of steel on the weighting scales. Putting that hand sized sandwich shield between both objects, the 40 kg reverted to 10 milligrams. However my suspicion was heated by magnetometer data what shown only decrease of field magnitude by an order, not four. Probably my sensor was too large, just there is obvious that shield not "eats up" the magnetic lines but just declines it aside.
Anyway, the perforations somehow makes the ecrane more effective in comparison with identical construction with intact steel plate.

 

RE: Vilis/Willy the Green. Salut from today`s snowy Riga. Hope You are bit Southerner there in Lithuania, so weather is more like to summer.

For the case of Faraday Cage, there is no problem on small holes what customarily are created for better air circulation and cooling aim.

But the problem always is with narrow but long gaps, what are leaking always and seriously. Thus, the all doors, sticking, screw-together sewing etc must be made imagining there inside the very bright light source and outside is absolute dark - You should not be able to see any LONG gap-hole with light. How to do so - simply make the double corner, because the reflection from grounded metal is near zero. Just stick a metal wall not in L shaped but Z shaped or E shaped etc gaps. When EM field have to make a sharp bend, it is unable to leak out.

Genaral law, when field meets the hole, that hole is fully transparent if any longest of sizes of this hole is at least lambda divide to four or larger. When hole size [or contrapetive terminology surface roughness] (I mean, a longest of it possible to measure diameters when form is not perfectly round) is at least lambda divide to 10, then in optics the EM wave scatters so well that lenses become transparent and mirrors refractive. For antennas at far lower EM range then are count to be insignificant, and cables begin to not need any impedance equalizing. When lambda divide to 40, then at astronomy is written, telecope needs no any smallest altering of surface quality, because it is ideal. Thus, the same criteria may be drawn for ecrane at lower frequency. When lambda divide to 40, then NO ANY field are leaking out.

However, I never have met the formula appreciating HOW MUCH is this "nothing".... sorry, none cant answer of such question.
BUT, simultaneusly, in the net are floating one inspirating picture with ecranated cables effectiveness comparizon. Dozen of cases - one braiding, one brading grounded from one side, then from other side, then from both sides, then double braiding grounded in all possible good and bad ways, and at last the triple grounding where the middle one is fed by inner wire signal in anti-phase. The worst ecranation variant was 3 dB, well made about 40 dB, the best ever possible 100-120 dB. And NO MORE!!!

For thickness. As far the Focault depth is enclosed in the ecrane material, the going through component is weakened. Just take the skin-effect formula and calculate, for example in case of 0.5 mm thick platelet, how strong is field at Your frequency in death of 0.5 mm. That will exactly be the part leaking out the ecrane even the any holes are absent. Look for formulas 4 and 5 in the reference https://en.wikipedia.org/wiki/Skin_effect
However there works one combinated effect. Your ecrane gonna be grounded of course It means it is not only using the Faraday cage effect but also the GND effect. Thus may draw the resistive divider circuit where lower resistor is inductive X(L) for all the grounding wire geometry and upper is inductive plus active resistance (Pithagor formula Z=sqrt(R^2+X^2)). Thus the attenuation again is calculable, however the calculation accuracy is under the large and shameless grin.

Last but no least. The Faraday Cage effect is UNIDIRECTIONAL (!!!!!!!!) effect. What is said by Faraday Principle - all the charges from inside is flowing outside and never vice versa. Thus, cage what are not connected to GND is effect-less against the radiation leaking from inside to outside. Whilst it is highly effective when radiation is coming from outside to inside. Thus, for this first case, the dominating EMF decrease effect is created by GND and nothing more.

PS. I forgot to give any directing motions about how large the ecrane box must be. In the older literature for example brochure of seventies колебательный контур, sorry I forgot the author name, probably Emeljanov(**), was said that ecranating coils of contcentrated selection filters to eliminate the unpleasant interference betwen ecrane and coil, the distance between outer coil points and ecrane steel must be optimum 3 radiuses of coil but never smaller as 1 radius. Am sure the same thumb-rule may obey even for boxes. May read instead of coil radius the bigger of radius of the PCB patches or current providing sensitive leads. So, the PCB size may be significanly bigger, but if one sensitive wire is 3 cm long, then box maust stay off at least a 3 scm or best case 10 cm. Therefore all sensitive leads must be as short as possible. Otherhow inductive coupling factor is large. What is it, shows formula from contact-less recharging theory. https://en.wikipedia.org/wiki/Wireless_power_transfer or coupling factor k=M/sqrt(L1*L2) and then U out=(di/dt in)*k. Here is good primer in wireless energy transfer math: https://www.sciencedirect.com/science/article/pii/S0045790618326326

And, one word of discourage - The Zenneck waves. Just avoid to involuntary to create it in the box. https://en.wikipedia.org/wiki/Zenneck_wave
(**) Just checked, author was another - Skripņikov Ju.F., here the link https://libcats.org/book/496557?ysclid=lg3i3628e5617414425

 

Last but no least. The Faraday Cage effect is UNIDIRECTIONAL (!!!!!!!!) effect. What is said by Faraday Principle - all the charges from inside is flowing outside and never vice versa. Thus, cage what are not connected to GND is effect-less against the radiation leaking from inside to outside. Whilst it is highly effective when radiation is coming from outside to inside. Thus, for this first case, the dominating EMF decrease effect is created by GND and nothing more.

So all of the security shielding we had on airplanes to prevent EM emissions from internal secure communications leaking out never worked.

 

RE: <<So all of the security shielding we had on airplanes to prevent EM emissions from internal secure communications leaking out never worked>>
But WHO really cares about EMI disturbances somewhere in city produced by airplane over that city??
Its so small and so far
Just contrary is about electronics inside the plane. It is well secured by just the elementary cage. No any need to ground it by the wire hanging down to the soil
And why the mobile phones in the airplane are so poisonous - just because them are IN THE SAME cage made of duraluminium "box" named fuselage.

 

RE: <<So all of the security shielding we had on airplanes to prevent EM emissions from internal secure communications leaking out never worked>>
But WHO really cares about EMI disturbances somewhere in city produced by airplane over that city??
Its so small and so far
Just contrary is about electronics inside the plane. It is well secured by just the elementary cage. No any need to ground it by the wire hanging down to the soil
And why the mobile phones in the airplane are so poisonous - just because them are IN THE SAME cage made of duraluminium "box" named fuselage.

What does that non-response about EMI disturbances even mean?

Properly designed (for the frequency and types of field emissions) Faraday cages are effective inside -> outside, outside -> inside and ungrounded.

 

<<A Faraday cage operates because an external electrical field causes the electric charges within the cage's conducting material to be distributed so that they cancel the field's effect in the cage's interior.>> https://en.wikipedia.org/wiki/Faraday_cage

Of course You may to try contaradict to largest peer reviewed encyclopedia worldwide, but I shout the hundreds af scientific books and publications on this topic says the same, sorry.

And I may bet the two years ago when I made a monitoring system to our taste-special beer factory at countryside reserve diesel electro-generator station for it monitoring system transmitting every minute the number of parameters "telegrams" toward the central office 2 km afar, I met the problem that station warranty is broken if I make a least hole into steel housing (container like the railway wagon). The GND was regulated by electrical grid rules so 4 Ohms indeed there was. How to make the LORA transmitter to work if I have forbidden to pull antenna outside?! Already I fell near a stupor but I find a strength to say, guys, at least let we should try!

Then, out-coming signal was diminished only some 10-fold, cant say I wasnt surprized. However in-coming signal indeed was just about cancelled. We installed the repeater just on the nearest tree aside the wagon, and data radiolink begun to work faultless, however the out to in telegrams goes with full power while in to out telegrams goes with minimal power. May it be convincing proof for the theory is always right??

Why this unidirectionality is not much bolded in low class primer books - because NORMALLY (except few special cases) the cage is grounded. Airplane is one of those "special cases", however I am not sure, the total capacitance of so large metal piece is indeed large, thus the cage (probably) may act like a local GND with capacitance against the Planet. Have never had a need to understand until the kidneys deep such a riddle.!!!

!!! Highly welcome if have any proofs (please references on the table) stating the other !!!

Overall:
Cages: ungrounded = in to out - small efect; out to in - large effect
Cages: grounded = in to out - notable effect; out to in - maximal effect

Note: If a "proper engineered" cages would give the ultimate++ effect, the 120 dB would not be a limitation, the all femtoamper circuits never would fail, and kiloampere supply chains would never radiate EMI. Its my everday life to beat out all of that sh**

Ah, ya, one more story of my praxis. Some years ago the certain factory came to me, what produces 85% of all world products in narrow medicine goods market nische. They erected the special laser cutting CNC equipment what drives the laser scalpel very accurate and makes the job on the quartz glass fibre. And, hundreds of times every day that CNC becomes mad, try to cut the ceiling of machine, or machine itself, or clash into end-stopper become bent etc etc. My task was to find what damn is wrong there. At first I thought a software, then some cables braiding grounding. At last the oscilloscope shows a fantastic cause-less peaks on the cables. I started to measure the EMI in room. It was teen of kV to meter at 50 Hz and slight less on the 100 kHz. Worth to tell the computer and drivers/controllers was in grounded aluminium box staying in grounded steel box, but just across the window in third floor heigth the local 20 kV electric grid line was going just some 5 or 10 meters afar the nearest wire. Oppps! We tried with hardly altered grounding wire. It helped just slight. The madness sessions become 3x less frequent. Then I gave an advice to just rearrange the unit to next (inner) wall of the room. There it already works for 10 years and everything works perfect. Note, the direction of disturbance was oriented to maximum diminishing direction. Just the field was strong. 100 kHz (or probably 110) is used for technologic telecomunication between far beating HES operators, thus the voltages isnt very weak.

Third example. In Sweden one strongfield laser lab asked me to solve the synchro signal providing 30 meters off the superlaser to scanning system. Signal was high speed TTL (about 1 nsec) and laser room had EMI in frequency range between 100 MHz and 1 GHz the several kilovolt to meter. Because of so high 50 kV field the room must stay locked when laser is on. So, firstly I tried, of course coaxial, then three layer coaxial with antiphase on the middle-braidings braiding. Results was woeful. So, all my knowledge about braidings and Faraday cages was too tiny to "engineering it properly". Why, - the field was coming into cable not outward, but no effect is absolute, sad to tell. Even in right direction there are some limitations what are not possible to overstep. Then I found the last measure (and may advice it to everyone in similar situation) - HFBR 1528-1523 or in newer successor HFBR 2528-2523. The circuit was elementary and worked faultless. May try to find the EGAS publications on my surname, there are the technical details on this.

 

<<A Faraday cage operates because an external electrical field causes the electric charges within the cage's conducting material to be distributed so that they cancel the field's effect in the cage's interior.>> https://en.wikipedia.org/wiki/Faraday_cage

Of course You may to try contradicts to largest peer reviewed encyclopedia worldwide, but I shout the hundreds af scientific books and publications on this topic says the same, sorry.

And I may bet the two years ago when I made a monitoring system to our taste-special beer factory at countryside reserve diesel electro-generator station monitoring system transmitting every minute the number of parameters "telegrams" toward the central office 2 km afar, I met the problem that station warranty is broken if I make a least hole into steel housing (container like the railway wagon). The GND was regulated by electrical grid rules so 4 Ohms indeed there was. How to make the LORA transmitter to work?! Already I fell near a stuporbut I find a strength to say, guys, at least let we should try!

Then, out-coming signal was diminished only some 10-fold. However in-coming signal indeed was just about cancelled. We installed the repeater just on the nearest tree aside the wagon, and data radiolink begun to work faultless, however the out to in telegrams goes with full power while in to out telegrams goes with minimal power. May it be convincing proof for the theory is always right??

Sorry but you IMO have selective reading and understanding as demonstrated by the video I linked above in mixing electrostatic conditions with dynamically changing EM wave energy conditions. What I'm saying doesn't contradict wiki or any scientific books and publications on this topic. At RF, because of skin-effect, there are two effectually electrically separated inner and outer surfaces on a proper shield (faraday cage), just like there are RF inner and outer surfaces of a coax shield causing one side to be a transmission shield and the outer side to be a radiator with unbalanced RF currents. Effectively you have two boundary conditions for RF on the shield for the inner RF source and the outer RF source emissions.

Airplane is one of those "special cases"

Sure, the EM fields and atoms know they are flying. I think your beer-factory monitoring system example is mixing causes with solutions. It really has little in common with faraday cage applications and requirements.

 

RE: <<there are two effectually electrically separated inner and outer surfaces on a proper shield, just like there are RF inner and outer surfaces of a coax shield causing one side to be a....>>
Hmmm, fine thought. I must swallow it and bit to chaw it certain time.

However the first taste - Yepp, indeed. But what is an impedance between both virtual shells then??

Then HOW the inner surface will duplicate the outer wall potentials (variating in time) if definition of effect and boundary conditions when solving the integral for Cage (*) is "inner wall have definite zero voltage on it, all the charges run to outside wall". So, when source is inside, its clear, outside wall catch every single electron and accurately repeats the inner field. When the source is outside, the inner surface want to capture the outer field but cage principle cannot be cancelled, the all charges goes outside so the inner wall still keeps the ZERO.
Yes, there always may be set the dynamic resistance or even capacitance and inductance between those virtual shells thus the signal is slight-slightly weaker in one copying the other. Thats a true reason why ecranating effect (when non-grounded) is never absolute. Probably.

(*) being the University radiophysicist by my first educattion before all another educations, I had in my biography the period when I managed to solve that differential equations of field and pull out the integrals, yet never in my 64 years them turned out to be needed so them are well forgotten, but I still remember the boundary conditions they demanded, the inner surface V=0 if closed shell.

 

RE: <<there are two effectually electrically separated inner and outer surfaces on a proper shield, just like there are RF inner and outer surfaces of a coax shield causing one side to be a....>>
Hmmm, fine thought. I must swallow it and bit to chaw it certain time.

However the first taste - Yepp, indeed. But what is an impedance between both virtual shells then??

Then HOW the inner surface will duplicate the outer wall potentials (variating in time) if definition of effect and boundary conditions when solving the integral for Cage (*) is "inner wall have definite zero voltage on it, all the charges run to outside wall". So, when source is inside, its clear, outside wall catch every single electron and accurately repeats the inner field. When the source is outside, the inner surface want to capture the outer field but cage principle cannot be cancelled, the all charges goes outside so the inner wall still keeps the ZERO.
Yes, there always may be set the dynamic resistance or even capacitance and inductance between those virtual shells thus the signal is slight-slightly weaker in one copying the other. Thats a true reason why ecranating effect (when non-grounded) is never absolute. Probably.

(*) being the University radiophysicist by my first educattion before all another educations, I had in my biography the period when I managed to solve that differential equations of field and pull out the integrals, yet never in my 64 years them turned out to be needed so them are well forgotten, but I still remember the boundary conditions they demanded, the inner surface V=0 if closed shell.

Good, seems you needed to jog a few memory cells to get back on track. At RF the inner and outer surface don't duplicate. The energy is reflected or absorbed as heat due to EM wave propagation in a good conductor.
https://farside.ph.utexas.edu/teaching/315/Waveshtml/node65.html

(See Exercise 7.) Equations (867) and (868) indicate that the amplitude of an electromagnetic wave propagating through a conductor decays exponentially on a characteristic lengthscale,

, that is known as the skin-depth. Consequently, an electromagnetic wave cannot penetrate more than a few skin-depths into a conducting medium.
...
According to the previous analysis, a good conductor reflects a normally incident electromagnetic wave with a phase shift of almost

radians (i.e.,

). The coefficient of reflection is just less than unity, indicating that, while most of the incident energy is reflected by the conductor, a small fraction of it is absorbed.

Grounding affects the electrostatic potential of the shield conductor IRT 'ground'. There is no effect on shielding at the effective EM shielding frequency range.

There's never absolute isolation or attenuation between 'shells' but with layers the reduction in detectable EM energy is very effective.

 

About saltwater immersed antennas - have a bit heated a fingertips making a cow stomach immersed monitoring transmitter. Yet signal came out hardly weak, but enough for "telegram" receiving when in near distance. So yes, about saltwater its written near the truth.
About the statement of Focault effect is only factor determining the shielding efficiency, its obvious the statement is one-sided and badly tied with reality if it is really meant be solely existent effect. Meanwhile all said on the Focault effect is true, of course.
About the graph, no dispute the curve is indeed like this, however I would bit correct the maximum possible limit but no much. But I dont realize what means the underscript "the DEMANDED ... effectiveness". We are speaking on factually possible not a virtually wished one, isnt so?
Anyway, thanks for the quote, its well written, it recall me a youth and that a equations we immersed was far more complex. At least Focault effect in our Phys-Math Methods Primer was explained in some 20 pages not in one. I would like to find that book indeed to have a stronger arguments. However, strongly later I found the 90% of text was "closely adapted" from well known Western textbook having another author name on the cover But anyway time is washing everything away, sooner or later, and only what are left are personal experiences or "fisherman tales" - what I told You those had happened with me. And why I spend so much a time for in-deep with narrow topic, because I feel the Lithuanian guy is my neighbour country inhabitant having rather similar language to be called a brotherly nation. And I know their problematics (about) in laser industry Im been familiar with.
RE:<<There is no effect on shielding at the effective EM shielding frequency range.>> WHEN the gnd path is large enough to have milihenries, hhilst between duoshells exist femtohenries, You are absolutely right. Except the cases when its not like that, GND is ultra short, or case is having a huge capacitance against the physical Globe as in the case of that wagon-like container with electrostation, but also shield is badly made (magnetic steel) thus I believe it is effective on 50 Hz, and, as I told at the 868 MHz was rather uneffective - what mighty worked in my benefit. So both ends must be compared always, instead on one alone. IMHO.

OK, Im going to sleep. its near the ten o clock evening. Buy buy!

 

I didn't say it's the only factor, only that earth grounding is not a factor for (far-field) RF radiation shielding. Near-field (reactive), the added ground conductor might have a small effect (usually within measurement error) due to induction effects. Think logically about how energy moves here, we don't have a static or quasi-static utility frequency circuit theory current loop as the RF energy hits the conductor. IMO most of fixation with Faraday type RF shield grounding comes from thinking with simplistic circuit theory concepts that unfortunately see current loops as an energy carrier instead of field/charge carrier interactions causing localized reflections and absorption. I would ground it for safety like we do for most things electrical.

What's much more important is actually keeping a good RF (no long slot antennas) seal on the thing.

I don't fear EMP attacks but the basic info is good.

Yes, that's the DEMANDED effectiveness as the minimal specification. The actual sweep and tested effectiveness at X location and application are usually classified.

 

nsaSpook - now we are staying on the much nearer key. Absolute agree that Faraday cage is damn multi-effect complex creating still hiding some quirks. For example, that multilayer pie what is capable to reflect the magnetic lines within concurring quqlity as mju-metal but thousandfold cheaper (I wrote about it in another chapter). Until now I dont deep enough understand how it happens if neodimium magnet effect to steel piece on the weight scale it affects with factor of 1000 and higher but measuring field with precision professional magnetometer it shows only some 4x diminishing of magnetic field. Theoretically impossible but observed with my own eyes and fingers. If someone answers the riddle, the Nobel Prize is practically at the hands, probably.

 

nsaSpook - now we are staying on the much nearer key. Absolute agree that Faraday cage is damn multi-effect complex creating still hiding some quirks. For example, that multilayer pie what is capable to reflect the magnetic lines within concurring quqlity as mju-metal but thousandfold cheaper (I wrote about it in another chapter). Until now I dont deep enough understand how it happens if neodimium magnet effect to steel piece on the weight scale it affects with factor of 1000 and higher but measuring field with precision professional magnetometer it shows only some 4x diminishing of magnetic field. Theoretically impossible but observed with my own eyes and fingers. If someone answers the riddle, the Nobel Prize is practically at the hands, probably.

Magnetic fields can be tricky.
You can shunt it with shielding or null the field if it's fairly static

In sensitive systems where full magnetic field shielding is not practical we use magnetic field cancelling systems like this:
https://www.vibeng.com/blog/how-emi-cancellation-systems-work

https://www.techmfg.com/products/electric-and-magnetic-field-cancellation