EMP, EMI, and EMF Consulting Engineers

"REAL EMP-ENGINEERING FOR 21st Century Electromagnetic Threats"

The Leaders in Electromagnetic Protection Engineering

"The EMP Engineering Expertise you need on YOUR team"

Contact:  EMPEngineering@emag-associates.com

A Good Friend of mine had these comments on a "Trash-Can EMP Shield" he had constructed.    This dialog was so good and his questions so to-the-point I'm including them here.

so steel wool it is! Appreciate the info!  Yep, I had a roll of standard box-store aluminum tape and thought it may have done some good; part of the problem is that I haven't really dug into the propagation paths of signals in that arrangement -- i.e., because of the gap that'd exist because of the adhesive.

Well, the adhesive is a dielectric gap.   You have two methods of propagating energy - conducted emissions across a highly conductive surface - and radiated emissions .  The conducted emissions will go right under the dielectric gap.   When the conducted emissions reach the lip of the trash can (under the lid) they hit a discontinuity and then re-radiate.   So that's why you really have to get the conducted emissions off of the surface of the trash can - that's why the steel wool works.   You have test data to look at now - so really that rules everything.

A few related questions come to mind . . .
- does the can/'cage' need to be grounded to earth?  or should it not be?

It does NOT need to be grounded (and my test data was collected UNgrounded), nor should it be UNLESS.....you have something else connected to it.   For a small stand-alone faraday cage - you do NOT want it grounded.  At with RF a ground can act as a ground at some frequencies, or as an antenna at others.   If you have a shelter, that has external connections - you definitely want to ground it (at multiple locations - I can explain where/why if you are interested) - because you do not want stray currents running across the surface of the shield (potentially magnetically saturating it - reducing shielding to a very low level under saturation)

- does the shielded device (e.g., radio) inside of the can/'cage' need to be insulated from the can/'cage'?  by what or how much distance?

Maybe.   You don't want an antenna of a radio hitting the metallic part of the shield - some shielding effect is likely to be akin to eddy-current shielding at very low frequencies - it could possibly have an adverse impact.  But really it probably doesn't matter - but you likely would want to add something anyway to cushion things from vibration so go ahead and do it with something insulating.  It won't hurt (but won't likely make much of a performance difference either). 

- any particular reason for measuring at 910 MHz?
2 reasons.   #1 902-928MHz is unlicensed band - so I know I'm not interfering with anything important - and any other user has to suck up any interference I cause.   #2 is that an HEMP pulse is about a 1 ns pulse with frequency content up to about 1000MHz - so 910MHz is pretty close to the high end of the EMP - so smaller imperfections would be more visible (because of wavelength proportionality).   You can be assured that your shield will perform BETTER at lower frequencies.   That said, testing at more frequencies would raise confidence in the performance.    Still, chances are my trashcan design will meet HEMP Mil-Spec at lower frequencies.    Most EMP energy is transmitted at lower frequencies (below 300MHz) so this is a very relevant test.  The weakness is keeping the electrical seal in a high moisture environment.  You can improve longevity with a moisture seal of some kind.

- does the can/'cage' material or thickness have any significant relevance, as long as it is RF conductive?  could the same effect be achieved by wrapping something with aluminum foil and rolling over the edges a few times?   Or maybe a metal paint can?

Refresh your memory by googling up "Skin Effect"....basically the principal that the higher the frequency the "thinner" the transmission layer for the energy.   But there is another reason why the galvanized trash can is good.   it's made out of steel.    at low frequencies (say 10kHz to 1MHz) the energy is mainly magnetic in nature - you can't really shield against magnetic fields - but you CAN redirect them with a ferromagnetic material - so the trashcan will direct energy around the shielded space at low frequencies, and act as a faraday cage at higher frequencies.     Faraday cages are only designed for electric fields.    Aluminum foil does nothing for low-frequency magnetic energy.   To act as a good MIL-SPEC shield aluminum has to be at least 1/4" thick (or copper).   On the other hand, at low frequencies a physically small device will absorb little energy even in the magnetic realm.   So an aluminum box should be OK, but a room sized foil shield would probably not meet the EMP shielding at the low end.

- a room-sized Faraday cage typically isn't a seamless or sealed-seam box, is it?  somewhere along the way, I got the impression that a screen/mesh was used, but I guess that depends on the frequency that is being 'blocked'...

Electrically, yes, it's a seamless 6-sided box.   When you have a door, you use something like a Double Knife-Edge door with fingerstock gasketing to maintain the electrical seal when it is closed.   When you have a pipe penetration, like for water or sewer - you use a waveguide penetration, which you could google up - a waveguide will not transmit energy below the cutoff frequency so a 1-inch diameter by 6-inch conductive tube would work as a waveguide penetration - you could stick a fiber optic cable through it (as long as it was all-dielectric) and not hurt the performance.    Electrical connections have to go through a filter.   Power connections have to go through a large inductor too.

As for screen/mesh - the performance is the issue.   As a benchmark a typical 40-mesh copper screen (very fine) will have about 46dB of shielding at 900MHz - that's the physics of it.  It will fall short.   Anything else you do to it will make it worse.   50dB might be good for some applications, but it's 30dB short of the EMP standard, and it won't handle the magnetic realm at low frequency.     You're right though, the size of the mesh is what drives the performance of a mesh shield.

The myth of grounding, insulation are largely performance irrelevant.   With typical electrical grounding techniques it may work to lower performance.

All tests performed with 910MHz 1.5 watt test transmitter on full power.

Approximate Link budget:   33dBm Xmit +5dBi Antenna -2dB cable loss -31dB (FSPL @900MHz @3ft) -10dB attenuator setting
32 + 5 - 2 - 31 - 10 = -6dBm  (reading on graphic is about -8dBm so it's pretty close - but ultimately we don't care exactly because we're just subtracting actual measurements)

Blue line = Lid of the trashcan off -8dBm

Pink Line = Lid secured as tightly as possible -36dBm    (attenuation about 28dB)

Yellow Line = Steel Wool placed on lip of trashcan and Lid secured as tightly as possible -80dBm    (attenuation of 72dB)

This demonstrates at least 72 dB of attenuation.   MIL SPEC is 80dB

This was a stock trashcan (kinda flimsy lid, actually) a stock package of steel wool from Lowe's  Total cost about $25 bucks



SPectrum Analyzer performance Data

Materials List (Lowe's)

Item #:  94614 | Model #: 807-006 

2.  Homax 12-Pack #2/0 Steel Wool Pads 
Item #:  134662 | Model #: 10121100