Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronic Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

[Moved] Best Shielding Material

Status
Not open for further replies.

aguijon

Newbie level 5
Joined
Jun 4, 2014
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
120
Hi all,

I want to manufacture a box which can attenuate a signal @ 5GHz up to 50 dB. I will use this box to create a controlled test environment. We have produced a box from Nickel-Chromium thas has a width around 6mm but it could not give sufficient amount of shielding. Some guys said coating this box with copper would improve shielding performance and some other said design a new box made by fully copper although it is gonna be expensive. Can you give me any idea about this situation, material suggestion, process suggestion etc. please?

Thanks in advance.
 

Dan Mills

Advanced Member level 2
Joined
Nov 10, 2013
Messages
510
Helped
237
Reputation
474
Reaction score
229
Trophy points
43
Activity points
4,558
What is the lowest frequency you need to shield?

Why use Nickel-Chromium? I would have just made the thing out of ally and then had it alcromed for surface conductivity.

For 5GHz copper plating will likely be fine (Skin depth is nearly nothing at 5GHz), but the real issue is that any opening longer then about 1/10th wavelength will allow the thing to leak, my bet is that this (Or leaks around connectors or just plain unshielded or unfiltered connectors) are your real issue.

EMC gaskets, three terminal caps, and coax feedthrus and **Loads** of screws are all appropriate, and be very careful about things like vent grills, these need be be a fine welded mesh with the edges in continuous electrical contact with the rest of the enclosure.

Regards, Dan.
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
47,478
Helped
14,047
Reputation
28,349
Reaction score
12,703
Trophy points
1,393
Location
Bochum, Germany
Activity points
276,115
How do you know that the apparent insufficient shielding is actually a problem of your box and not of the measurement setup? Most likely you'll have much more problems with signal feedthrough than shielding of the bulk case.

I agree with Dan Mills that if the shielding is actually insufficient, it's most likely a problem of electrically sealing the cap to case face and any additional voids. Are you using state-of-the-art RF gaskets?

But first of all, I would like to know how you'll measure the 50 dB shielding effect.
 

aguijon

Newbie level 5
Joined
Jun 4, 2014
Messages
9
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
120
Actually, our aim is placing a Access Point inside this box and make it unreachable for any clients or other devices by air. Because using a set of attenuator, phase shifters etc. we want connect this AP to another client (that will be placed also another shielded box) as conducted to simulate some feature of our devices. Normally our attenuation range has to be up to 100 dB in order to be simulate connection fail when we need. We can use attenuator board that we have designed to 60-70 dB. but after that level although I increase the attenuation level between two devices as conducted, RSSI level seen by client is not changing any more. So, we assumed that after taht certain level signal is somehow prefer to use air (leaking from shield) instead of travelling on variable attenuators and transmission lines.
As both of you said, we were suspicious of cover of box and slots that is used to permit cable transpass, primarily. We intented to reduce their effect using proper gasket and copper tape by covering this discontinuties. Although we observed some amont of improvement, It was not sufficent for us.
Finally, we are using RSSI level as an indicator of shielding effect.

Thanks a lot for your valuable comments.
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
47,478
Helped
14,047
Reputation
28,349
Reaction score
12,703
Trophy points
1,393
Location
Bochum, Germany
Activity points
276,115
The approach sounds all-in-all feasible.

I would allocate part of the signal attenuation (at least 20 dB) inside the box to relax the isolation requirements for the RF cable. Route power and signal connections through RF connectors with coaxial low-pass filters screwed to it. Also place RF absorbing foam inside the box to suppress any standing waves.
 

Dan Mills

Advanced Member level 2
Joined
Nov 10, 2013
Messages
510
Helped
237
Reputation
474
Reaction score
229
Trophy points
43
Activity points
4,558
Also, make sure your RF connectors are done up to the recommended torque, loose SMA connectors leak.

ALL the cables must be filtered right as they enter the box, and you will need to be using good quality coax between the source box and your attenuator, use something designed for lab microwave use.

Picking up a few simple 10dB fixed attenuators (Minicircuits via Ebay) can be a good thing as they can be used to easily find your leaks if it is a cable or connector issue, but I would note that AP RSSI indications are often poor and are sometimes more in the way of BER meters then signal strength ones.

I would note that coax, attenuators, and connectors can all have issues at 5GHz, variable attenuators not designed for use up there are especially problematic.

Regards, Dan.
 

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,111
Helped
1,520
Reputation
3,041
Reaction score
1,119
Trophy points
1,393
Activity points
32,334
During years, I found that the best are the double shielded boxes. The same is valid for coax cables.
Was already proved experimentally that two metal thin foils provides better EM shielding than a single thick metal foil, which is double in thickens than the two thin foils together.
Is somehow related to the energy of the electrons which lose much of the energy passing one foil, and they don't have much energy to penetrate the second one.

For example ETS-Lindgren shielded enclosures, which can reach 120dB isolation, all of them are double shielded.
http://www.ets-lindgren.com/RFTestEnclosures
 

Dan Mills

Advanced Member level 2
Joined
Nov 10, 2013
Messages
510
Helped
237
Reputation
474
Reaction score
229
Trophy points
43
Activity points
4,558
That's just the skin effect, the bulk metal is invisible to the 5GHz current, only the surface layers matter, and a double shielded box has four surface layers rather then two.

It has nothing to do with electron energy as such.

Regards, Dan.
 

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,111
Helped
1,520
Reputation
3,041
Reaction score
1,119
Trophy points
1,393
Activity points
32,334
Skin effect doesn't have anything in common with this situation, or at least when we speak about a shielded enclosure that works at relative low frequencies.
Electromagnetic fields produce forces on the charge carriers (usually electrons) and more metal layers dissipate further any electric currents generated at both sides of the walls.
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
47,478
Helped
14,047
Reputation
28,349
Reaction score
12,703
Trophy points
1,393
Location
Bochum, Germany
Activity points
276,115
I believe, the double shielded coax cable effect is about imperfect shielding by wire meshes rather than fields penatrating a solid metal shield. Similar situation with metal boxes, we can expect 60 dB field attenuation by a 7*skin depth metal layer (or 100 dB by 11.6*). But that's a purely theretical number for an enclosure without a cap (with respective slits) or feedthroughs. In a real world, double shielding is a way to reduce the impact of enclosure discontinuities.
 
Status
Not open for further replies.
Toggle Sidebar

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top