Mitigating RF signal leakage using a standard RF relay

I'm involved in a project that requires that we shield a space from ambient EM radiation using a Faraday enclosure in some conditions, and in others, we try to replicate the ambient EM radiation from the outside world between the ranges of 20MHz and 6GHz using two different Rx (outside our Faraday cage) and Tx (inside our Faraday cage) antenna pairs (one operating from 20MHz-2GHz and the other operating at 2-6GHz).  There is a hole with a sleeve in our Faraday cage which connects the Rx and Tx antennas, while preserving EM shieding.  We have two Relcomm RF relays (model # RDS-2S1D6) to fit inbetween the two antenna pairs.  You can review my previous posts in this forum under the heading "How to Unfaraday a Faraday cage".  Now to the problem.

Even when we configure the two RF relays to be normally "off" by connecting output to the "NO"" terminal, and capping the "NC" terminal with a termination load cap (specified to work from DC-6GHz with up to a 1 watt power--being more wattage than we're actually transmitting), we still get a slight bit of RF signal leakage through the RF relays into our Faraday enclosure.  It's actually enough to pick up an FM radio station (albeit, out of many others that are effectively blocked out) as well as some almost non-functional internet connectivity over 2.4GHz wifi in some cases.

So, the question I am posing is how can we ensure that absolutely none of the EM radiation from the outside world (or very close to it) gets through this RF relay when it is in it's unpowered (default) position being "NO" or open/off? Attached is a photo of one of the RF relays we are using, as well as one of the amplifiers that services the signal incoming from the Rx antenna but before it gets to the RF relay.  I appreciate any responses from any of you about how to resolve this signal leakage issue.  Thanks for any incoming help!

Best regards,

Richard Knowles, Ph.D. (in psychology--not RF or electrical engineering    )

Firstly, I would have thought you should use a connector on the cage, not a sleeve.
It needs to be something like a female to female chassis mounting adapter.
Then the relay and amplifier need to be mounted directly onto the connector at the inside wall, no coax at all.
The relay has a finite attenuation, see the data sheet. Coax cable also has leakage unless you use a high quality cable with a non braided screen.
Without being able to look at the setup, I can't say much else, so probably not much help,
cheers,
Richard.

The sleeve is just a shieded tube of metalic fabric that connects to an aperture (hole) that is between the layers of our Faraday enclosure. Please see the attachment for a better understanding of what our dual-layered, pyramid-shaped Faraday enclosure looks like. The dimensions are 11'x11' at the base, and 7' high to the apex. Are you suggesting that we connect the RF output from the Rx antennas (on the outside of the tent enclosure) to the actual Faraday tent itself?

The insertion loss of the relay itself is only .1db, so that is negligible. The coax cables that were included with our lower frequnecy (20MHz-2GHz) antenna as well as our higher frequency antenna (2-6GHz) are supposedly shielded, but I know there must be a way to properly ground out the signals that are intercepted by those cables' shielding, perhaps with a "drain wire" of some sort. Do you have any information about how to do this?

Attached are some (admittedly crude) block diagrams and supplementary material that can possibly shed light on this issue. Please let me know if you have any questions about these diagrams, as they are quite rough.

I once heard that any chunk of metal can serve as an antenna if it has the proper back end to receive it. That may be the effect that we are dealing with here, including RF signals that are piggybacking onto DC power lines that are intersecting with both the RF relays and the amplifiers that they serve/power. In those cases, we end up with slight to moderate RF contamination in the supposedly EM-free Faraday enclosure despite the relay being "open" or "off". It is truly vexing!

Thank you sincerely for your response, Richard! I hope to hear back from you.

Best regards,
Richard Knowles

It looks like only the last attachment that I made went through while none of the previous ones did, so let me try to include the other attachments I was trying to include before:

Get another opinion from an expert, but I don't believe any signal cables should come through the skin without breaking them with a connector, that keeps circulating currents confined to the internal and external circuits. Coax cables are leaky to some extent as well, so you need to put your relay and amplifier directly on the connectors where they come through the skin, avoiding any possibility of re-radiating the external signal from a length of cable.
In your case it probably means a flat conductive plate, with a bunch of connectors mounted on it, electrically bonded to what I assume is a metalised plastic.
That applies to any other wires as well, if power (AC or DC) and control lines come from outside, you need to use a very wide band filter(operating over the full bandwidth of interest) and the cables must come through the skin via feed through capacitors. Otherwise they are likely to re-radiate any signals they pick up from outside the cage. You can't assume that the RF impedance of these lines is very low, particularly at higher frequencies, even with bypass capacitors on them.
Either the ARRL or RSGB manuals (published annually some years back) had an article covering the issue of signal feedthrough using various configurations, from a bare wire to multiple filter stages,, with actual measurements, see if you can find that information, it is very interesting.
It's not insertion loss you have to worry about, it is the off state attenuation, it is not perfect, so you will still get some signal at the output side of the relay when the contacts are open, not much, the data sheets indicate 80-90dB.
You mentioned in a previous post that you had access to some RF expertise, this should be a problem that such people could look into and advise on, best I can do with my very limited knowledge,
cheers,
Richard

Hi Richard

I am investigating cable shielding outside of the cage and the grounding of that shielding as well as mini-Faraday cage shielding for RF components that need to exist outside the big, pyramidal Faraday cage. Might I talk with you more directly about this, perhaps over the phone? This project depends on a solution to this problem, and I would love to get some direct advice about this in the near future, as this is somewhat time sensitive.

Best,

Richard Knowles

P.S. I just responded to this message with a much more detailed response, but I'm not seeing it show up in the lists. This inspired me to respond again with a more brief response to see if it would show up

Hi Richard,
as I said, I am no expert in these matters, and I may be leading you astray (read I cannot be held responsible etc etc). I do recall that when working for Industrial Research Ltd we had to take extraordinary measures to filter the mains supply going into a shielded room to avoid rf leakage into the room.
I have enclosed a diagram of how I believe the Faraday cage should be set up, talk to some of your local RF contacts about what I have suggested.
The supplier of your 'cage" should be able to offer advice as well. If you look up suppliers of faraday cages, you will find that they often have mains filters, as well as filters for signal lines (Ethernet for example) which sit in screened boxes on the outside of their enclosure.
I am on a small farm in the middle of the Canterbury Plains, in New Zealand, surely there is someone closer that you could contact for proper expert advice.
Cheers,
Richard

Hi Richard,

Your diagram seems to indicate that the RF relay and amplifier are in their own mini-Faraday cage within the larger cage, is that right? Also, we only have one tube-like port or sleeve which connects the Faraday enclosure to the outer world, not two as are depicted in your design (although we do have a magnetically sealed door that we can let people in and and out with). We don't intend to run any cabling through that door, though. Also that bulkhead jack/jack seems to imply that there is a solid surface to mount it onto. Remember that our Faraday cage is actually a tent, so it is all flexible fabric. I'm really curious about these filters, though. Our range of interest is from 20MHz up to 6GHz, at frequencies below about 20MHz, the tent becomes progressively more and more transparent to that kind of low-frequency EM energy. I'm therefore surprised that there is a filter which takes out 60Hz hum as this is in the ELF range, and should pass right through most Faraday cages. If there is a filter that can take out all RF energy up to 6GHz, that would be awesome!

The two areas where contamination is occurring are:
1. Through the DC power cables that power the RF relay and the amplifiers and
2. Through the coax cabling that connects the Rx antennas on the outside with the Tx antennas on the inside (and with the relays and amplifiers being in-between. It's probably taking a ride on the shileding of those cables, and we've demonstrated that simply by running a 25' length of coax cable (half inside our Faraday tent and half outside, it serves as a crude antenna and allows for certain FM radio stations to be received on the inside.
3. There is the vague possibility of a third route, being that anything which has exposed metal on it (like the casing of the relays and amplifiers) could be picking up RF and passing it along. I've heard that any chucnk of metal can be used as an antenna as long as you have the right back end to help it receive.

I'm wondering if using shielded DC power cabling http://www.rfindustries.com/pdf/OutdoorPowerCablesSpecsheet-2.pdf and then grounding the drain wire might be helpful. Also, I might be able to ground the outer shielding of my coax cables that are connected to the Rx and Tx antennas. However, grounding the shielding of any wire may not work since what if it is transmitting RF all the way down to the ground? I dunno--that's maybe an issue to figure out empirically.

As far as you being far away, none of that matters these days! I'm in Silicon Valley, and there are some RF folks here, but they all want to charge a large amount of money for consultation, since that's what they're used to, and we have no spare funding for any consultation. I'll check in with Holland Shielding, though. They are the manufacturers of our Faraday cage.

Thanks again, Richard!

The relay and amplifier are already in enclosures, that's all you need. You can buy the necessary connectors to do away with coax cable.
As I understand it, any wire that goes through the skin of a faraday cage destroys the integrity. Hence my suggestion of a conductive plate, the coax connectors and emc filters for the power supply and control lines could all be on the same plate. The diagram was just to get the idea across. I understand that the material is flexible, the plate could sit on it, and it could be clamped around the edges with an additional metal gasket on the other side, with bolts going through the material. It may have to support a bit of weight, so internal support with some sort of simple stand would be needed inside the cage.
EMC filters are available from places like Element 14, but they are not as broad band as you require, they miss out at the top end, hence additional filtering which includes the use of feed through capacitors (again, look them up on Element 14), or custom EMC filters, will be needed. The EMC filter needs to work over the range of frequencies your are interested in.
Cheers,
Richard

Further to my last post, look up Tusonix 4101 and 4209 at element 14, they are the sort of filtering you need on your dc and control lines. If mains goes into the unit, you will need higher voltage devices. Watch out for SMPSs in test gear as well, many in use these days, they can generate EMI in the tens of MHz, although good quality gear should have good EMI suppression,
cheers,
Richard

Ok, and these filters (like the 4101 or 4209) would be mounted on a plate that is sitting on the "skin" of the Faraday enclosure? I assue you just solder your power lines to the terminals on each end, right?

I agree that coaxial cables are somewhat leaky, and I understand that your suggestion is to minimize or even eliminate their use in some cases. I think its safe to say that any coax cabling that is coming through the sleeve of the Faraday tent will have its shielding contaminated with RF from the outside world, but what if we grounded all coax shielding to the fabric of the tent as soon as it entered the tent?

Best,
Richard

Yes to the first part, and not just power, but all control lines as well.
As to the second, the only way to 'ground' a coax is to use a connector, it takes care of the problem of electrically bonding the outer conductor over its full circumference, which is essential.
Since you will already have coax connectors on the cables, all you need is a bulkhead mounting adapter that mates with the existing cables.
Cheers,
Richard

Okay, regarding the second part, we will have one Rx-Tx antenna pairing that uses coax terminating in BNC connectors and the other Rx-Tx pair has SMA connectors. Surely there are filter plates which offer BNC and SMA connectors, right? BTW, I have contacted Holland Shielding (the manufacturers of our Faraday tent) and they say that we would need a filter plate as well. It's good to get confirmation from two different sources. Thanks for being one of our sources, Richard! Also, it's good to have some Richard-to-Richard communications! I haven't met another Richard in quite a while.

Best,
Richard

No, strictly custom, there are just too many permutations and combinations. The same plate will be used for all connections in and out of the cage, which complicated matters even further.
However, the connectors are available off the shelf for

Hi Richard, I have a new update: I contacted Holland Shielding (the makers of our Faraday enclosure) to get a price quote for a grounding plate we could affix to the skin of our Faraday tent. They quoted us a price of $634!! I then looked around on the internet, and found grounding plates at Home Depot for $28.64 (https://www.homedepot.ca/en/home/p.galvanized-ground-plate-cw-connector.1000423074.html). That seems like a much, much more cost-effective solution to me! But the plates are .3 inches thick, and I'm guessing that it would take a machinist with some specialized metal cutting equipment to cut the two holes we'd need (one for an SMA female-female coupler, and one for a BNC female-female coupler). Do you have any idea about how to drill holes like this properly? Plus we'd need the right kind of coupler, and my first thought was something like this (https://www.radioshack.com/products/bnc-female-to-female-coupler), but we'd need to connect two plates on top of each other (one on the outside and one on the inside), and I doubt that a coupler like this could fit through .6 inches of metal. Maybe the BNC and SMA connectors only should be soldered into the holes on the outer plate while the inner plate has larger corresponding holes to allow the plugs to partially stick through while still allowing the cables on the inside to connect to the inside terminals properly?? I dunno, what are your thoughts here?

Looks like the Home Depot grounding plate is designed to be buried, and used for the ground connection from a switchboard. It will probably not be suitable, i'm guessing issues like conductivity, and electropotentials will be important. They will both have an impact on the screening properties, the second will relate to how well the plate bonds, and remains bonded, to the metalised plastic of the cage.
I would need to see a lot more detail about how the plates from the faraday cage supplier are configured to work on your cage before making any further suggestions. It could be that they are over kill for the small number of connections you need, you may be able to do a cheaper copy, but you would need to know what material is used, what the surface finish is, and how do they clamp the plate to the what I suspect is the relatively flimsy skin of your cage,
cheers,
Richard

It's designed to be buried? How did you know? Anyway, I will ask Holland Shielding about what material their grounding plates are made of, wha the finish is, and how they clamp the plate to the metalic tent fabric. Clamping the plate to the flimsy surface of the skin is supposedly accomplished by keeping it very close to the ground, but, perhaps with some support structures beneath it to hold up the weight without it leaning down towards the ground under its own weight. Honestly, I haven't really gotten to the mechanics of how to get the plate stable yet--just working on how to possibly get one installed for the moment. Thanks for your insights about what are good questions to ask them, though. After having paid them over $6k for this Faraday tent, I hope they would be willing to disclose some of these details about the constitution of their grounding plates. That would certainly be nice! I'll let you know what I discover from them.

Best,
Richard

Hi Richard, I have been in dialog with Holland Shielding, and it looks like they use galvanized stainless steel plates for connections with bolts (the inner and outer plates connect with bolts that clamp the two plates together). They said that they use "nickeled" or "chromed" steel for magnetic sealing of the two plates, since those metals are magentic while stainlesss steel is not. Upon further research, I realize that stainless steel is actually magnetic sometimes, though, depending on the alloy used. It looks like electrical connection issues between the fabric and the plates are resolved (according to the guys at Holland Shielding) as long as we get the plates properly clamped together against the metalic fabric of the Faraday tent to create a good electrical connection. Anyway, I asked them if galvanized steel (not stainless teel) could be used for either bolted or magnetic connections, and they said yes. So... I have put together the following diagram of how I could accomplish this grounding plate scenario with those two grounding plates from Home Depot I described before--only with the inner one cut to be a little shorter than the outer one. This way, this outer plate can reach all the way to the ground and then intersect with a wooden board with a groove cut into it to accept that plate, and support its weight. What are your thoughts on this solution, Richard?

Hi,
the concept seems fine. You said in an earlier post that the plates from Home Base were 0.3" thick!
Is that correct?. Sheet metal suppliers often have small off cuts of material, which come in various thicknesses, you may get something more appropriate (thinner) from such a supplier.
Keep the plate as small as possible. Remember you have to put feedthrough filters through the plate as well for your power and control lines, these can be supplied with threaded bushes.

Cheers,
Richard

Well, on the web page for this grounding plate (https://www.homedepot.ca/en/home/p.galvanized-ground-plate-cw-connector.1000423074.html), it says that the "assembled" thickness is .3 inches, while just above this, it is said that the thickness is .25 inches, which is much more manageable. I definitely see what you are saying about the plating needing to be thin in order to accommodate these adapters (of usually very limited length) that need to pass through. Oh, also, in the configuration/version we are using now, only two shielded coax cables will end up passing through the Faraday skin, and so as long as the RF contaminated shielding of those cables is grounded on the plates that will be installed, then all will ostensibly be well. Of course, the shielding integrity of the inside of the Faraday tent will have to be checked in the usual fashion once this procedure is adopted--just to make sure it really does work, and all remaining signal leakage is blocked.

Best,
Richard