Faraday cage continuity question

[vytautas_k]

We have built a faraday cage out of 1,5mm thick galvanized steel sheets. Size is 1.5m*1.25m*1.5m. The steel got powder coated and was assembled on a frame. Each side of the cage is made from separate sheets that are joined using rivets. All of the separate sheets are connected with grounding wires and the whole box has continuity. We are trying to shield against 100Mhz-1000Mhz waves. My question is: Is electrical continuity enough for a faraday cage or was it better for the metal sheets to not be painted and touch each other with bare metal?

 

[Dale]

I would temporarily disconnect the grounding wires and see if you still have continuity through the rivets. It is important to have conduction across the edges and corners. Indeed the powder coating should not have been applied to the seams, but perhaps the rivets offer sufficient conduction.

What is the spacing between rivets?

 

[sophiecentaur]

All of the separate sheets are connected with grounding wires and the whole box has continuity

It all depends on the degree of isolation you are after. How will you decide what screening you are achieving? Local broadcast signals can give a good indication if you have a measuring receiver of some kind. It depends on what signals you are using the cage for; can you choose quiet spots on the spectrum?

A good Faraday cage, as used for critical measurements, will have very good electrical contact between all metal sheets. This should be over the whole length of each seam. A gap will easily allow induced currents to pass around the edges of the gap and merrily pass the EM (at some frequencies) through to the other side, almost as if the sheets were not there. For good contact, the edges of the sheets should be bare metal and perhaps continuous copper braid sandwiched between them along the seams. Certainly, DC continuity is only a rough indication of screening performance.
Any door should also be screwed down all the way round or there should be spring fingers along all sides.

Then there is the issue of power and signal leads in and out. Much better to use battery powered equipment with everything sited inside the cage and no people. Any necessary leads in and out should have suitable blocking filters. Fibre optic links can pass through very small holes and that stuff is very available these days.

This all reads as a bit negative but it it possible to put in a lot of effort and materials making a 'Faraday Cage' which lets you down at just the frequency where you want to operate. You will have looked at available cages, I imagine and decided that they are asking too high a price for them; so you have chosen DIY. High costs will give you reliability and good spec. DIY may give you adequate (or even good) results and that would be 'good engineering' but results can surprise you either way.

But, on the positive side, your required level of screening may not be too demanding. It would be as well to decide on your required spec from the outset. Look at the spec of laboratory grade cages and see whether it's what you need.

If you are really not happy with the performance then drill out the rivets, grind off the power coating along the edges and use copper braid throughout. (But first eliminate paths through in/out cables)

 

[vytautas_k]

It all depends on the degree of isolation you are after. How will you decide what screening you are achieving? Local broadcast signals can give a good indication if you have a measuring receiver of some kind. It depends on what signals you are using the cage for; can you choose quiet spots on the spectrum?

A good Faraday cage, as used for critical measurements, will have very good electrical contact between all metal sheets. This should be over the whole length of each seam. A gap will easily allow induced currents to pass around the edges of the gap and merrily pass the EM (at some frequencies) through to the other side, almost as if the sheets were not there. For good contact, the edges of the sheets should be bare metal and perhaps continuous copper braid sandwiched between them along the seams. Certainly, DC continuity is only a rough indication of screening performance.
Any door should also be screwed down all the way round or there should be spring fingers along all sides.

Then there is the issue of power and signal leads in and out. Much better to use battery powered equipment with everything sited inside the cage and no people. Any necessary leads in and out should have suitable blocking filters. Fibre optic links can pass through very small holes and that stuff is very available these days.

This all reads as a bit negative but it it possible to put in a lot of effort and materials making a 'Faraday Cage' which lets you down at just the frequency where you want to operate. You will have looked at available cages, I imagine and decided that they are asking too high a price for them; so you have chosen DIY. High costs will give you reliability and good spec. DIY may give you adequate (or even good) results and that would be 'good engineering' but results can surprise you either way.

But, on the positive side, your required level of screening may not be too demanding. It would be as well to decide on your required spec from the outset. Look at the spec of laboratory grade cages and see whether it's what you need.

If you are really not happy with the performance then drill out the rivets, grind off the power coating along the edges and use copper braid throughout. (But first eliminate paths through in/out cables)

Thank you for the answers.

I have checked the continuity without the grounding wires, and it seems fine. However, as you mentioned, it only comes through the rivet points.

I am seriously considering disassembling it and sanding down the joining surfaces. Can I use a desoldering wick between the surfaces, or will it not work properly due to the flux?

 

[sophiecentaur]

Look up “braided copper wire” on eBay. Loads available here in u.k.
The desoldering stuff is a bit too thin, imo.

 

[Vanadium 50]

Local broadcast signals can give a good indication

If they aren't coming in through the power lines. Rand into that once. Chased my tail for a couple days.

This all reads as a bit negative

But, on the positive side,

I saw what you did there.

 

[Dale]

Chased my tail for a couple days.

You are lucky you found it so fast! I have had MRI suites where these issues required months of troubleshooting (finally found that the facilities people had changed to a different fluorescent bulb so the RF was being generated in the room, not a leak).

 

[Dale]

I am seriously considering disassembling it and sanding down the joining surfaces

I would recommend that. Getting good conduction at your RF across the seams is critical.

 

[Baluncore]

My question is: Is electrical continuity enough for a faraday cage or was it better for the metal sheets to not be painted and touch each other with bare metal?

100 MHz to 1000 MHz. λ = 3 metres to 300 mm.
Holes, slots and gaps must be smaller than about λ/20 = 15 mm. Your rivets would need to be about 10 mm apart to significantly reduce RF leakage.

I would suggest that you clean off the paint/powder and solder the galvanised seams with a fillet of Pb-Sn solder, to make a continuous electrical contact.

 

[sophiecentaur]

If they aren't coming in through the power lines. Rand into that once. Chased my tail for a couple days.

Not sure what you're saying there. If the cage equipment were not powered from outside and you measured the received interference in the same way (free-standing battery receiver) with the cage present or absent then it would seem to be a fair test. A mains cable can be an excellent antenna.

I saw what you did there.

A balanced view.

Your rivets would need to be about 10 mm apart to significantly reduce RF leakage.

Also the gaps between screws etc around the lid / door. Sneaky little varmints will not give up easily.
Solder is good value and it's lucky he's chosen galvanised. The guys in my lab made a cage with aluminium sheet and it was never good for UHF measurements. These days they could have used the fancy new solder.

 

[Baluncore]

Sneaky little varmints will not give up easily.

I once used the bow thruster room of an ice strengthened vessel as a screened room. The steel hull was over 25 mm thick, and I could hear the morse ident of every VLF NDB within 1500 km.

 

[sophiecentaur]

I once used the bow thruster room of an ice strengthened vessel as a screened room. The steel hull was over 25 mm thick, and I could hear the morse ident of every VLF NDB within 1500 km.

And the shafts of the bow thrusters made excellent feeders into the 'screened room', via the rubber seals. Varmints.

But the sea - going tale was good. Forty years before the mast, man and boy. Aargh Jim Lad.