DIY RF absorber for the interior of a fully shielded lab

[rhb]

I'm going to be building a 12' x 16' shielded lab soon in my 1530 sq ft shop building.  The shielding will be 26 gauge galvanized steel with fully soldered seams.  Power will come in via  a 2.5 kVA isolation transformer. Air will come in through a carefully screened vent system.

I want to apply an RF absorber to the interior walls.  I expect that even a DIY solution is going to cost around $5k for a mini-split HVAC zone, walls , ESD flooring, etc.  So this is a serious and non-trivial undertaking.  But I've got the lab gear to make it worthwhile.

I have a sheetrock texture gun and more experience using it than I ever wanted.   I think that sheetrock mud mixed with  charcoal and iron powder is likely the only option I can afford for an RF absorber.

But I have a minor problem.  I don't know how much iron and charcoal powder to mix into each coat.  It will take at least 10-20 coats with decreasing proportions of iron and charcoal to produce an effective anti-reflection coating.  I have the resources and ability to calculate the required proportions from first principles, but I *really* don't want to work that hard. 

The proportions will vary logarithmically, but other than that I have no idea what they should be.

Can anyone provide any hints, links, papers or other relevant information?  Thanks.

Have Fun!
Reg

[CatalinaWOW]

Do you know what frequencies you are trying to kill?  SMPS stuff?  Computer bus signal frequencies?  10 MHz frequency references?

With a bit of google you can find the formula for skin depth vs resistivity and frequency which I think would give you the answer you need, though my intuition says you may not be happy with the answer.

It may be more productive to kill the stuff at the source.  Eliminate SMPS.  Shield around sources and the like.  That is tough too, but the wavelength of most of the likely sources makes killing it by absorption tough also.

[coppercone2]

are you sure it will be dimensionally stable and not crack with temperature changes and humidity changes?

and keep in mind that iron dust is flammable, it would be a extreme situation but if the walls are flakey, and it accumulates on top of stuff, it can ignite if dumped on a spark, but I only read about this happening in extremely dirty casting places that were never cleaned (according to US safety chemistry board). Like 80's action movie factory end scene places, where it would accumulate on top of vents, electricals, etc. And crappy sheetrock that does not set properly and flakes off would polish your lungs :-/

[TheMG]

What is your main goal for this? Are you trying to keep EMI out so it doesn't interfere with sensitive measurements/tests you'll be doing inside the lab? Or are you trying to keep your experiments inside the lab from interfering with the outside world?

Usually the main reason for RF absorbing material inside a faraday cage would be to attenuate any RF bouncing around inside, thus creating what is called an anechoic chamber. The main use for such a thing would be if you're characterizing antennas, doing radiated EMC testing on products, that kind of thing. Not really necessary if all you want to do is keep out external EMI.

I'm not sure how effective a conductive coating on the walls alone is going to be, especially if the aim is to absorb a wide range of frequencies.

Have you seen this video?

 

[T3sl4co1l]

Well, upwards of 50MHz, as I would guess from the size of the chamber.

You'll have to test the ratios yourself I think.  Depends on the type of filler (and if it changes during placement -- pure iron powder will end up nothing but an ugly brown streak if you glop it into drywall compound... ), and to some extent on how close it is to the metal wall.

I wonder what a representative small-scale test would be.  Sure, easy enough to put a blob inside a coil and give it a buzz, but how do you relate the geometry factor of a near-field coil with a semi-free wave?  ("Semi" in that, the wave is inside the material to begin with, which presumably has some not-insignificant index of refraction, plus it's in front of a metal panel that more or less doubles up the wave energy near the surface.)

I get the feeling you may be better off with thousands of ferrite plates.  Not at all cheap, or easy to place, but more likely to do a good job.  Last time I was thinking about this, I had been seriously thinking about planar transformer plates (PLTxx in Ferroxcube typeology) arranged in a triangle or square lattice -- assembled edge-on, standing up perpendicular to the surface; this over top of a flat layer directly on the surface.  The idea being, the cavities between plates has a trapping effect, and also a graded index of refraction sort of effect; and the base layer is effective both directly for low frequencies, and at higher frequencies where the lattice concentrates waves into it.

The commercial equivalent is either flat tile alone, or waffle pattern tiles.  The triangle waffle pattern / lattice would mimic this.

You'd still want the stealth foam pyramids for >UHF absorption though.

Tim

[coppercone2]

I assume he wants to 'silence' the area before an anechoic chamber, where the equipment is, not work inside of an anechoic chamber (that would drive him nuts). And they don't even work that well unless you are in the center of them.. and they only work towards the design spec of the compliance document (how many meter chamber).

Something to consider is that most chassis is cheap.. you might be able to get better effect by redesigning the chassis to have EMI screens, EMI gaskets and thicker materials/better bonding, or to put it inside better containers and extend the keypads a bit.

Also you can add ferrite inside of equipment and even redesign stuff to run on linear supplies connected to vents to get rid of switchers.. (like audio processing computers are sold with giant linear supplies.. even though they still have a 4GHz CPU in them......)
[

I can imagine someone building a giant anechoic chamber that has a rack of equipment in the middle leading into another anechoic chamber... it would be like working on the asteroid from the movie 'Armageddon'. Or LV-426....


Getting long high quality VNA cables and 'going outside' might be a better idea, you can feed a fist full of thin 2.5mm hardline cables through a long waveguide into a real anechoic chamber that is quite long and very well shielded (triple shield cable, like braid foil braid), and use RF switch boxes too to reduce cable count.

Anechoic chambers are kind of acoustically unpleasant too. It's a very unnatural work environment that will distract you

[Berni]

What is the goal of the RF absorbers anyway?

Usually they are used is to prevent reflections from setting up standing waves and resonances in the chamber. So it makes the limited size of a RF shielded room look more like an infinitely sized room. This makes EMI compliance testing much more predictable and controlled.

If the goal is to simply have a room with as little as possible RF interference then the absorbers won't really help much. They won't magically reduce the overall RF noise everywhere. Only thing they would help with is if a device inside your room emits RF at the right frequency to resonate between the walls, with smooth flat metal walls this resonance can boost the RF level by a few times. Its the exact same thing as putting curtains into a room to make it less echoey. You can use the same ideas for RF. You probably noticed before how a empty room echoes a lot while putting furniture in it makes it go away. So just putting metal furniture in your room would already cut down most of the RF bounce problem as the complex shapes it introduces scatter the waves all around. Additionally you can prevent easy bouncing of it by making the room dimensions weird, things like making the length,width,height of the room not be a multiple of each other, having walls not be completely parallel to each other helps too, having the wall not be completely flat helps too...etc All this still won't make it an anechoic chamber, but will prevent any significant resonance effects from forming.

Also black antistatic foam can also work pretty well as an RF absorber in the GHz regions, its foamy non uniform structures prevents resonance and the resistive properties turn RF into heat. Tho you might have to do some tests of your own to figure out what type works best since the foam comes in all sorts of different densities and material mixes and i would guess some work better than others.

[fourfathom]

Bags of BBQ charcoal stacked up against the walls?  I think coal has been used as an absorber.  It's cheap, anyway.

[rhb]

I'm a retired seismic research guy.  Reflections were the focus of my entire career.

The goal is to minimize irreducible noise from the stuff in the room, lights, T&M kit, etc.  I *might* build a small anechoic chamber for some antenna modeing work, but that's a *possible, maybe* and would not be in the room.  If I did, it would be immediately above the lab.  The building sidewalls are 10', so I have room if I can go to a high enough frequency for say a 6' x 6' x 6' cube to be viable.

My current bench area has all the power cords up to the IEC connector in a combination of EMT and flex steel conduit.  I have done a lot of sniffing with the SA mode of my Instek scope.  There is a lot of noise coming out the front of the LeCroy from the LCD and the back and sides of the Instek emit a staggering amount of SMPS noise.  I've got an EMI filter on the LED lamps overhead and 1/8" hardware cloth covering  the fixture and any openings on my computers that emit EMI.  I'll eventually address the LeCroy and Instek noise, but right now I need a place for the four 5 ft stacks of TEA in my dining room on mover's dollies.

Latex wall paint may be a better binder than sheetrock mud, but at much higher cost.  However, I might be able to add paint to the mud, iron and charcoal mix.  In studying the issues, I think likely I'll wind up with 1/4" drywall primed with a coat of latex paint, the absorber to a thickness of 3/8" or so and a sheet of 1/4" drywall for the exterior face.  I've shot enough drywall texture I am sure I can put on a 3/8" layer in 10-20 coats.  However without intermediate coats of paint it would be quite susceptible to impact damage and a real nuisance to repair.  A wall face  of 1/4" drywall solves that.

What is required is a logarithmically graded transition from air to steel.  The other requirement is that no two room faces be parallel.  I expect the lab to cost about $5k to build.  ESD flooring is around $1k, HVAC ~$2.5k, Faraday cage about $1k.  I've spent $25k on TEA, so a $5-6k lab  room is not unreasonable.

This is Reg's version of the vintage muscle car he is not at all interested in.

Commercial absorbing materials are too pricey.  The ESD flooring is a good example.  I don't have a quote yet, but I expect it to cost around $5-6/ft for 192 sq ft.  But I can't think of a durable alternative to commercial ESD vinyl flooring.

I have an 8510C coming and already have an 11801 & SD-24, so I can directly measure reflections if I can devise a material property test setup close to 50 ohms.

This forum has some pretty amazing people on it, so I'm hoping to someone who has tried this will appear.

In actual practice it might not provide any benefit, but it can do no harm.  And it is much easier to do before I setup all the test kit, parts bins, etc.  My goal is the quietest play room I can afford, not to build an EMI lab.  I also expect to spend a lot of effort on quieting noisy fans.

Charcoal is easier to get in Arkansas than coal and while not as cheap, is cheap enough.

Have Fun!
Reg

[CatalinaWOW]

Good luck on your quest.  The real problem is that the only answer to your questions is data on the proposed configuration.  Lots of work and time before you know if you are on a promising path.  Even making a single sheet sample would be a fair commitment, but give you a lot of answers.

I am assuming that the noise sources you describe are high kilohertz, low megahertz, maybe even fractional gigahertz stuff.  You might find better performance and possibly easier fab with a semi-discrete configuration.  Wires or strips of copper or aluminum tape on layers of hardboard with discrete resistances between tape elements and layers.

I also still think killing it at the source may be easier.  Tin oxide faceplate for the LCD with appropriate EMI edge gaskets.  Aquadag on inside of cases.  And so on.

A third possibility is going totally the other way.  Depending on your location and the construction of your shop it may be easier go RF transparent on the walls and let the noise escape.  Obviously not viable if you are in an urban area or near an airport, or if your shop is a steel building, but worth thinking about.

[Berni]

Well in my opinion RF absorbers in walls are the least of your worries.

Making a good RF Faraday cage is actually rather difficult. Every single crack has to be sealed. Things like doors need fine pitch copper fingers all the way around them to give it good low impedance connection to the surrounding wall. Any wiring going into it has to go trough multiple stages of filtering to scrub out any RF. Even something as simple as placing a 10cm piece of wire between the door and closing it can raise the amount of RF in the room by an order of magnitude in a good shielding cage. And even then you will be able to see some peaks from strong broadcast bands like FM, TV, Cellular etc...

Then bringing in equipment is a whole new can of worms. Pretty much anything with a LCD spews out RF right out the front, if it has a CCFL backlight then its 10 times worse. Even something as simple as a PC keyboard can spew out RF (I could see a coworkers wired USB keyboard from the next room on a spectrum analyzer). LED lighting is known to be a especially bad RF firehose. Anything with stepper motors blasts out RF. Heck some crappy switching supply even managed to send my spectrum analyzer into overload from across the room due to having a extra LNA on my antenna. Spectrum analyzers themselves spew out a bunch of sweeping tones as they run. We did a lot of sensitive EMI testing(radio receivers near our products are VERY easy to upset) and it turns out pretty much everything spews out a lot of RF noise even if it meets CE regulations(They aren't super stringent). Also 50/60Hz will still get everywhere unless you have shielded mains cable all the way to the equipment.

Even if you had a full on certification EMI lab anechoic cage you would still have some leakage and your equipment spewing out RF all over the place. The absorbers would only push down some humps where the room might resonate if set up in a echoey way. Heck even putting a large house plant in there might solve that problem. Also yes that gradual transition from air to steel does discourage reflections, but the size of this transition has to be comparable to the wavelength. Even 300MHz has a wavelength of 1m and id imagine 1m of absorber material is not very practical while you are still not yet at 100MHz. The better place for absorbers is lots of ferite on all the cables of equipment so that the noise they are producing is turned to heat before it has a chance to radiate out of the cable. For example you put a muffler on a car exaust rather than surround the road with noise absorbing barriers, yet still have the road itself be a noisy place.

But if you do want to coat the walls with something absorbing i would recommend first doing tests using a smaller scale metal box (Designed to be very prone to echo on purpose), putting an antenna inside and then measuring the return loss of it. Place your surface under test against the opposite wall and see how much it improves the return loss.

[rhb]

As stated previously, the seams will be fully soldered.  Yes, the door will be "interesting", but as I will be the only person I can simply clamp a metal sheet across the inside of the doorway with magnets when I need it to be really quiet.  That's not going to be very often.

I am methodically quenching sources as I locate them.

The fundamental room resonances are at 30, 40 and 60 MHz and harmonics of those.  The harmonics will add constructively in places, but I'm not going to worry about that.  This is "best effort".  I'm not trying to meet any particular level, just less noise than without it.

I've swept from 9 kHz to 2.9 GHz with my 8560A and a whip antenna on the input.  I've got significant locally generated noise below 1 MHz which I'll have to locate and remediate at the source, a daytime 1 kW AM station at 1370 and a half dozen or so FM stations.  I'm next to a small general aviation FBO, but that's not likely to be a source of significant trouble.  No radar or ILS to contend with. 

The cage will keep the broadcast stations and sawdust out and allow good temperature control which is all I really need.  The building is insulated, but the lab will have 2" aluminized face foam between the studs so I can economically maintain it to 1-2C or better.   No matter how good a dust collection system I have on the woodworking machines, it will not capture all the sawdust.  That's why I put a partition down the middle to separate the metal and mechanical section from the woodworking and lutherie section when I built it.

If I figure out a good way to  install stuff so I can get unimpeded access to the walls (e.g. put it all on casters) I may just apply a layer of 1/4" sheetrock, tape it, paint and set up the lab so I can measure reflection coefficients and room resonances.  I have a bunch of really nice metal shelving with 3 or 4   rolling sections that allow access to the row behind it.  If I use that it  probably  makes absorber moot.  I certainly can't do anything about reflections off  that or the test gear which will be on the opposite wall.

I got a quote on ESD flooring and that will run ~$1200 including shipping which is not quite as bad as I had feared.  The room will cost about $1000 and the HVAC $2500, so even without RF absorber on the walls it will still not be cheap.  Raising the cost by 10-20% is not a big deal.

This is something I've thought about and researched for a long time (30 years).  I bought Ott's EMC book and some others.  I have a 5000+ volume technical library and  was in the business of doing things that had not been done, or  all I knew was someone else had done it and I needed to figure out how they did it.  I am inexperienced, but not naive. I simply thought it worth asking if anyone had tried making their own RF absorber.

Reg

[DaJMasta]

If you really want absorption, you can get relatively large (2x2 foot) conductive foam panels for not too much, something like the MF22-0003-00 may be what you're interested in ( http://www.masttechnologies.com/rf-absorbers/ ), but I'm sure there are other companies that can do something similar.  One of those on each wall and the ceiling would do a fair amount to deaden the interior for RF reflections, though of course if you were really after that, you'd just want to cover it all (and it does seem like ferrite tiles are the go-to standard for commercial test chambers).

If you've got a directional antenna you can probably do some basic testing on your own mixtures by just coating a panel and measuring the reflection to get close to a good mix, should you go this route, but since the variables that contribute to how good a given commercial mixture works as a binder probably aren't tightly controlled, getting a second bucket of the stuff may produce substantially different performance than the first bucket produced, so it's a bit variable even when you have your ratios dialed in.

If it were me, and I was only focusing on isolation, I'd probably spend the time working on the sealing mechanism (checking and rechecking the joints, having a plan for warpage with use or temperature, reliability of door seals, etc.), maximizing the power input filtering (and keeping it all in its own shielded box), and making sure your connection to earth is as low impedance as possible.  Having some absorbing panels on the inside is nice, but I'd think the spray on sheetrock developing cracks and effecting performance that way would be more trouble than I would want to deal with.

That said, maybe there's a viable option in there too..... just make a conductive-sheetrock-on-metal panel and bond the panel to ground, then you don't have to treat the whole room, don't run the risk of cracks nearly as much, and could just remake the panel if you come up with a better formulation or if it somehow is damaged.

[NiHaoMike]

LED lighting is known to be a especially bad RF firehose.

Pretty trivial to build a linear supply for LEDs, in fact there are even LED modules that just have a long string of LEDs connected to rectified mains with a linear regulator dissipating the remainder.

[rhb]

After I installed a Corcom line filter and covered the 4 ft dual lamp fixture with 1/8" hardware cloth,  the noise level dropped below the other sources.  I may need to revisit it, but for now other sources dominate.

Reg

[Berni]

LED lighting is known to be a especially bad RF firehose.

Pretty trivial to build a linear supply for LEDs, in fact there are even LED modules that just have a long string of LEDs connected to rectified mains with a linear regulator dissipating the remainder.

Yeah its not LEDs themselves but the common practice of running them on crappy $2 switching PSUs from the cheapest chinese vendor with little to no output filtering, combined with connecting the output to a long string of LEDs that acts like an antenna. So most off the shelf LED lighting solutions are an EMI nightmare.

As stated previously, the seams will be fully soldered.  Yes, the door will be "interesting", but as I will be the only person I can simply clamp a metal sheet across the inside of the doorway with magnets when I need it to be really quiet.  That's not going to be very often.

Well you can probably make a normal door into a EMI proof door with some work.

The copper EMI contact fingers can be bought as a continuous reel of it without ridiculous prices. You can use a door made out of metal in the first place or just stick a metal sheet onto a existing wooden door. Then just make a metal frame around it to hold the copper contact finger strips against the metal wall all around it when closed. The hinge side of the door might be tricky as you need to be in front of the hinge for it to work mechanically, but perhaps a harmonica of metal foil could cover the hinging edge of the door.

Oh and shelving units are indeed very good at breaking up echo so id definitely cover any unused walls with them since one can never have too much storage space.

Some photos would also be welcome since it sounds like it is going to be a very impressive room.

[rhb]

It will be some weeks or more before anything happens as I'm recovering from injuring some ligaments in my lower back which takes forever to heal.  And I do *not* want to aggravate it by starting to work before it's fully healed.

With the benches and shelving in the room I'll have about 4.5' x 16' of open floor space.  Everything else is shelving and bench. The moving shelving units are 3' wide.  I've got 3 so I can have 25 linear feet of shelving from floor to ceiling.  It came out of an office building.  It didn't suit the needs of a new tenant, so I had to remove it.  It's really nice stuff so I decided I'd keep it for myself.  I haven't priced the stuff, but I'd guess it's around $3k or more.  It was used for real estate title work folders back when that was all still paper. so they are made to take heavy loads.

I'm going to document the build in detail in another thread.  I've decided that the bench will be several sections on casters so they can be moved away from the wall for access to the cabling.  That will  allow me to retrofit RF absorber if I discover that the room has measurable resonances with the equipment in place. 

I'll revisit the  DIY RF absorber  topic again when I have the 8510C set up and some dielectric and permeability fixtures constructed.  It would be nice to have a small anechoic chamber for testing 2.4 and 5 GHz ISM band antennas.  I could suspend a chamber from the roof purlins so that I could lower it for access and then raise it up out of the way.  I've got space to hold a chamber 15-20 wavelengths or more at 2.4 GHz and still have 8 ft of headroom underneath.  But that's YAP, yet another project.

BTW Copper weather stripping  should provide a good RF gasket for the door.  You don't  see it often now in new work, doubtless because an elastomer is cheaper and easier to apply.  But it's common in older commercial work  where long life is important. Very likely it will make  closing the door require some effort, but it can't get snagged and damaged as readily as multiple fingers.  Applied properly to the door frame, it doesn't make contact with the door until the last few degrees of travel.

Have Fun!
Reg

[jonpaul]

Hello the professional EMC and RF guys build a screen room.

Copper floor, and structure, copper small mesh screen bonded to the structure.

A room for a well equipped lab is 6-12' on a side and cost ~ 5-10k $ in 1970s.

http://www.universalshielding.com/Products-Solutions/Shielded-Chambers

Bon courage

Jon

[CaptDon]

Try ferrel cats as the absorber, they work best above 1000v/m.
You may have to liquid cool them or replace them often. There
seems to be endless supplies, unless you live near Chinese
restaurants! Some cats have resonant peaks, you will have
to sweep each one to be sure.

[coppercone2]

Bags of BBQ charcoal stacked up against the walls?  I think coal has been used as an absorber.  It's cheap, anyway.

are you satan? Do you know what happened when the UAC started to combine demonic shit and RF?

some lecroy crap will catch on fire in the basement and its going to open a hell portal

try to consider fire load when designing this

[fourfathom]

Bags of BBQ charcoal stacked up against the walls?  I think coal has been used as an absorber.  It's cheap, anyway.

are you satan? Do you know what happened when the UAC started to combine demonic shit and RF?

some lecroy crap will catch on fire in the basement and its going to open a hell portal

try to consider fire load when designing this

So put the charcoal in Nomex bags.

[cdev]

AntiStatic foam may be fairly RF absorbent, and may be useful . Also you can load foam with activated charcoal yourself. But be aware, it makes it as hard as a rock and it turns solid very quickly. Much more and faster than foam by itself. The charcoal may act as a catalyst or something. It becomes hard as a rock and its very hard to mix in the charcoal evenly enough. Also, if it gets on your clothes, it could ruin them. So keep that in mind and make a plan, once it gets mixed you only have a few seconds. Wear safety glasses and rubber gloves too.

I wonder if antistatic egg foam is sold. That may be what you need right there. Make sure to read the MSDS.

Please be careful with iron powder. (or any fine powder, for that matter, be careful about the dust possibly created over time, any chemicals which might be contained, etc.)  I would just leave iron out unless you really do know how to do it and the current state of science on it.

 Iron powder is hazardous to your health, I am pretty sure. You should research this on pubmed. I am not saying don't do it, but just do the research needed to figure out how to make sure you remain safe. You may also need to cover foam up for fire safety reasons. In houses foam used in insulation must be behind a fire barrier. Just make sure that you dont create a danger for yourself. Interested in how this works out. Patents might be the best way to find out how its done.

Very interested in your build of this, good luck! Take care of your back!  And your lungs.

If you do use iron powder, I'd make sure the particles are not too small and are of uniform size .  Stay safe!

It will be some weeks or more before anything happens as I'm recovering from injuring some ligaments in my lower back which takes forever to heal.  And I do *not* want to aggravate it by starting to work before it's fully healed.

With the benches and shelving in the room I'll have about 4.5' x 16' of open floor space.  Everything else is shelving and bench. The moving shelving units are 3' wide.  I've got 3 so I can have 25 linear feet of shelving from floor to ceiling.  It came out of an office building.  It didn't suit the needs of a new tenant, so I had to remove it.  It's really nice stuff so I decided I'd keep it for myself.  I haven't priced the stuff, but I'd guess it's around $3k or more.  It was used for real estate title work folders back when that was all still paper. so they are made to take heavy loads.

I'm going to document the build in detail in another thread.  I've decided that the bench will be several sections on casters so they can be moved away from the wall for access to the cabling.  That will  allow me to retrofit RF absorber if I discover that the room has measurable resonances with the equipment in place. 

I'll revisit the  DIY RF absorber  topic again when I have the 8510C set up and some dielectric and permeability fixtures constructed.  It would be nice to have a small anechoic chamber for testing 2.4 and 5 GHz ISM band antennas.  I could suspend a chamber from the roof purlins so that I could lower it for access and then raise it up out of the way.  I've got space to hold a chamber 15-20 wavelengths or more at 2.4 GHz and still have 8 ft of headroom underneath.  But that's YAP, yet another project.

BTW Copper weather stripping  should provide a good RF gasket for the door.  You don't  see it often now in new work, doubtless because an elastomer is cheaper and easier to apply.  But it's common in older commercial work  where long life is important. Very likely it will make  closing the door require some effort, but it can't get snagged and damaged as readily as multiple fingers.  Applied properly to the door frame, it doesn't make contact with the door until the last few degrees of travel.

Have Fun!
Reg

The best place Ive found for small antenna testing is outdoors in an athletic field or parking lot. Some parking lots are deserted on weekends.

[E Kafeman]

> galvanized steel with fully soldered seams.
Galvanized and soldering in same sentence? It is not a solder-able material. Galvanized steel is very demanding to build a Faraday cage with as this oxide never will selfweld during pressure, it is opposite that will happen, it will self-isolate even during mechanical pressure between sheets of galvanized steel, in opposite to many self-welding metals and it is not possible to solder zink-oxide with normal methods. Even hard to weld as zink not can be grinded before welding  and welding creates less healthy zink-fumes and results in poor bonds.

Galvanized steel and copper should not be in close contact to each other due to the big galvanic difference, add minimal moisture and it will create a battery leaving copper sulfate as only remain of copper part.

A replacement to zink is tin as rust protection. It is very solderable and can be both welded  using most types of weld procedures and is soft selfwelding at pressure=> creates no HF leak when ageing. Tinplate is also relative cheap, especially as it can be bought in real thin thickness and used as wall papper.
However is magnetic properties suffering when thickness is low but that is more a problem for frequencies below a few MHz.
A drawback compared to galvanic steel is that tin is not self-healing and rust may occur as result of mechanical damage: https://en.wikipedia.org/wiki/Tinplate

Absorbing 50 MHz signals by several layers of paint is not really useful in an anechoic chamber. If paint had been an effective absorbent had it not been any need for expensive conical RF absorbents occupying valuable
space in the chamber. This 10 mm thick RF absorbing material reduces reflections with 0.1 dB at 50 MHz, if interpolated down from 500 MHz: http://www.masttechnologies.com/wp-content/uploads/2019/02/MR51-0006-00-Tech-Data-Sheet.pdf
It is very few producers of thin absorbing material such as paint or sheets consisting mostly of rubber/carbon/ferrite as absorbing material that shows measured result below 1 GHz as it not is much to show.

A bit more expensive solution is to use thicker sintered ferrite tiles which can be made somewhat effective also for relative thin sheets, 3-10 mm are common used thicknesses. Example at page 7:  https://www.ramayes.com/Data%20Files/TDK%20RF%20Solutions/TDK-IB-017WH-Absorber-2008.pdf

If you is real diy, a friend of mine created his own ferrite bricks for his chamber in a temporary owen (1000-1100C).
These brick was then placed in chamber at all walls, floor and ceiling,  behind the conical RF-absorbers as a way to increase usable frequency range for antenna measurements.

If it only is a minor improvement of RF impedance load that is needed so that an emitted signal will have a reasonable decay time by absorption can that be performed with a bottle of water/salt/sugar. That is often used in echoic reverberation chambers.

Placing noisy instruments inside a chamber while measuring other radiating signals in same chamber and noise and signal will have equally measured signal ratio independent of if chamber is echoic or anechoic.
Avoid to add own noise sources to measured object if possible.

A way to reduce such problem is to have a metal mesh window and placing eventual measurement tools outside of chamber and feed it with signals via both filtered and non filtered connector mounted in the metallic chamber wall.
A PC outside of wall or placed in its own chamber inside chamber can be controlled by mouse and keyboard using IR. A keyboard can be real noisy but an on/off switch can turn it off during measurements.

Avoid LED-light inside chamber. Place light outside of chamber and let light shine thru a metallic mesh. LED inside chamber with a long wire/antenna connected at each side is a passive frequency multiplicator if not shielded/decoupled to metal skin of the chamber.
If it is a problem with electric wires inside chamber do depend on what kind of measurement and needed dynamic range and chamber isolation but better to avoid it if possible.
Soldered copper mesh works well up to several GHz as window in chamber walls and commercial iron mesh from Chomerics have high visibility quality if a LCD screen is placed on opposite side of the mesh: http://www.hitek-ltd.co.uk/media/downloads/47/EMI%20Clare%20(1999).pdf
A mesh can be combined with an ordinary glass-window with glass coated with silveroxide which is rather reflective for RF signals, especially if well connected to chamber wall. Such glass is relative cheap compared to Chomerics EMI-windows and is used in many buildings to reduce sunlight/IR and reduce need for cleaning. It do maybe differ 40 dB in both cost and reflection factor relative Chomerics version but it can still be good enough or at least much better then placing an LCD screen not shielded inside a chamber.

I have anechoic chamber for antenna measurement but if radiation pattern should be measured with high dynamic range is free field outdoor measurement to prefer and measurable frequency range is then not limited due to too small chamber size.
In a chamber, calculate at least 3-6 lambda between reference antenna and DUT for not so directive DUT and additional at least one lambda behind both antenna and DUT, if not a horn antenna is used as fixed reference chamber antenna.
If horn antenna can a such antenna be placed directly at chamber wall which will save some space.

If chamber not is mainly intended for antenna radiation pattern measurement or other measurement methods tha need high level of absorption of reflections, I had not added any absorbers at all, and no paint, but tried to keep inside as clean as possible from not needed noise sources and no AC powerline, filtered or not, would been allowed if it is measurements that intends to be digging at real low noise levels and if line filter is a type with open wires at two sides, are these wires communicating antennas causing almost no attenuation at all at 1 GHz. If a nearby cellphone tower exist, and you will still have almost same signal at cellphone inside chamber as outside and almost all attenuation is punctured by placing antenna in shape of long wires inside chamber that is connected to similar long antennas outside of chamber with almost no attenuation.

As an example, powerline suited for a shielded chamber covering GHz-range with 100 dB or more in attenuation looks something like this: https://czpioneer.en.made-in-china.com/product/gyBmxDClCvkp/China-EMI-Shield-Power-Line-Filter-for-Anechoic-Chamber.html
Notice the separate channel for feeding lines inside chamber and outside in above link, if effective isolation in GHz range with 100 dB or more must filter installation be done a bit careful.
Anyone that have been working with designing such high RF attenuation knows that it is no easy task.
It is important with a good gasket and low loss grounding as unbalanced currents must be allowed to spread at chamber walls and further to a low impedance power ground. Also due to rather big capacitors from live to ground are such leak current rather high, often in range  up to 1A. If chamber not is properly grounded can a such current be lethal.

This kind of filters are expensive but I have built several similar filters with even better attenuation then 100 dB at 1 GHz, costing less then $100 in material and some own job.
Inside filter can it be splitted in several sub-chambers and 40 dB wideband attenuation in each chamber is reasonable possible to reach for from 0.001 to 40 GHz.
100 dB attenuation can be a bit too low if a strong transmitter such as a cellphone tower exist nearby and measurement are done at -100- -150 dBm. It is then a bit irritating to see a big hump in curve when measuring a freq. covering cellphone freq. in an else empty chamber or being able to receive phone calls when inside chamber, due to a ineffective line filter.

It is possible that you already is aware of these kinds of problems and know how to handle them and much of them are possible to solve with low budget even if professional EMI equipment often is very expensive.

[T3sl4co1l]

Galvanized and soldering in same sentence? It is not a solder-able material.

Strange, HVAC was soldering it for decades...

I've soldered zinc plenty of times myself; it's annoying as the zinc dissolves in solder (namely the tin fraction; cool fact, zinc and lead are immiscible), and isn't adequately cleaned by rosin so the longer you keep the joint hot, the scummier it gets as the zinc in the solder oxidizes.  Aluminum does the same thing but worse, but is also harder to tin in the first place (because, unless you have the proper flux for it, well, y'know).

But, get in and get out, hot enough iron, work quickly, absolutely it can be soldered giving a hermetic seal.

Acid flux is very effective, as can also be seen on brass parts -- brass can be difficult to solder with rosin flux, especially if it's been heated, but those basic zinc oxides are washed away effortlessly with an acid somewhat stronger than rosin.

Galvanized steel is very demanding to build a Faraday cage with as this oxide never will selfweld during pressure, it is opposite that will happen, it will self-isolate even during mechanical pressure between sheets of galvanized steel, in opposite to many self-welding metals and it is not possible to solder zink-oxide with normal methods. Even hard to weld as zink not can be grinded before welding  and welding creates less healthy zink-fumes and results in poor bonds.

Have seen galv spot-welded plenty of times before; the joint may well end up brazed, in part, but it's definitely metallurgical in any case.  (Another cool fact: zinc dissolves in iron, forming a system somewhat similar to brass (i.e., zinc dissolved in copper).  I'm not aware of any actual application of this, and actually making a controlled alloy from these two metals, is, well...yeah...  But, apparently the metallurgy is there!)

Spot welding may not be perfectly suited to test-chamber construction, for various reasons, but it's certainly adequate for, say, microwave oven parts!

You definitely can't arc-weld the stuff, though.  Well, not happily anyway, and not without strong ventilation...

Galvanized steel and copper should not be in close contact to each other due to the big galvanic difference, add minimal moisture and it will create a battery leaving copper sulfate as only remain of copper part.

Uh...where'd the sulfate come from?  And where'd the zinc something-ate go..?

The thing about a cell is, you get deposition on the lower potential surfaces, so, the copper and iron get plated preferentially by zinc, at the expense of bulk zinc in the system (which drives the reaction).  And the zinc has to run out first.

And if you're dripping sulfuric acid on your test cell, I guess a little corroded metal is the least of your worries?!

A replacement to zink is tin as rust protection. It is very solderable and can be both welded  using most types of weld procedures and is soft selfwelding at pressure=> creates no HF leak when ageing. Tinplate is also relative cheap, especially as it can be bought in real thin thickness and used as wall papper.

Yup, definitely easier to use.  It has a surface oxide which is thin and weak (and probably conductive anyway), and the metal is pretty soft, so that little wiping force is necessary to ensure contact.

As for metallurgy, hm, don't think there's anything wrong with a few percent tin in iron; probably acts to strengthen it, if anything.  I would wonder if it might cause grain boundary problems, or precipitates with other elements, etc.  Still, probably not something you'd ever have problems with, in ordinary, dry, low stress applications.

Avoid LED-light inside chamber. Place light outside of chamber and let light shine thru a metallic mesh. LED inside chamber with a long wire/antenna connected at each side is a passive frequency multiplicator if not shielded/decoupled to metal skin of the chamber.

Heh, well, depends.  I wouldn't trust just any LED fixture inside a test chamber, but a well filtered DC lamp will do very nicely indeed -- much less heat dissipation than an incandescent one, and the latter will still introduce interference if the mains isn't well filtered to begin with.  A custom build, with LED power supplies separate from the panels/strips/lamps, would be quite effective, just add suitable filtering at the chamber pass-thru.  Not even very custom really, as the blocks are COTS.

And needless to say, anything else entering: ventilation, test equipment power, EUT power and signals; all needs to be filtered or shielded to avoid bringing in outside interference.

A mesh can be combined with an ordinary glass-window with glass coated with silveroxide which is rather reflective for RF signals, especially if well connected to chamber wall. Such glass is relative cheap compared to Chomerics EMI-windows and is used in many buildings to reduce sunlight/IR and reduce need for cleaning. It do maybe differ 40 dB in both cost and reflection factor relative Chomerics version but it can still be good enough or at least much better then placing an LCD screen not shielded inside a chamber.

ITO glass is possibly the more common material, but if Ag2O is available and offers similar transparency-conductivity tradeoff, that's good to know.

Hmm, I wonder if Ag2O is prone to tarnishing?  Can't remember offhand if that's the case, or if it's mostly Ag(metal) that's prone to tarnishing (combining with sulfides to make Ag2S and such).

If chamber not is mainly intended for antenna radiation pattern measurement or other measurement methods tha need high level of absorption of reflections, I had not added any absorbers at all, and no paint, but tried to keep inside as clean as possible from not needed noise sources and no AC powerline ...

Heh, and on a related note, there are other kinds of test chambers -- testing for partial discharge is often done inside a specifically very low loss chamber, so that as much as possible, all energy from the partial discharge can wind up in the receiver, and thus some measure of its charge (say in pC) can be had.  Here, it doesn't matter so much that signals will be bouncing around at lord knows what frequencies; as long as it continues to bounce so that it can find its way to the antenna, that's all that matters.

A room completely plated in copper (with 100% soldered seams) is a wondrous sight...

Cheers!

Tim

[cdev]

I love copper!

There is a bar in SF's Lower Haight (Haight near Fillmore) whose decor used to and maybe still does involve a lot of copper welded together,  perhaps in a seamless fashion, although I kind of doubt it would have been seamless due to the cost. "Noc Noc" It stilll exists and has a web site (but the images on it are too blurry).  It seems to me to likely have gone through multiple decorative revisions since then.

It was a popular place to meet up for drinks, and probably still is.

As I'm sure semiconductor junctions do act as passive frequency doublers if the voltage induced in them exceeds the junction voltage, the issue of transistors, or diodes including LEDs creating nonlinear behavior might be important in a chamber of such kind. (Especially if it was being used to evaluate devices for miniscule RF emissions)

This activity has been used in bug detection for years and its easily observed using any strong local transmitter (even very strong AM or FM or DTV stations nearby willl do) It might be avoidable by using ferrites (but ferrites only will do their magic at certain frequencies ?) around both leads of DC feed to the LED panel and placing the panel both flat on a grounded backing (heatsink) and putting ferrite beads around any leads more than a small fraction of any wavelength anticipated to be used??? (just speculating here, have never done this) One could also put LED panels each inside their own shielded box using some kind of mesh that allowed most light through. (the key is that holes are kept to small fractions of wavelengths) 60 GHz, however, I don't know. As multi GHz RF behavior is over my skill level, unfortunately.

The physical layout of LED lamps likely makes a huge difference in how they behave when hit by an RF signal, and of course its something they dont design them for at all. If care is taken to (re) construct them to make sure most induced RF is common mode and it likely nulls itself out as seen at the junction, maybe?

Very very small diodes might even show visible light if a current goes through the junction, in the conducting direction so that might be helpful in fixing this. This kind of thing I expect may become more necessary for security because 5G provides small devices a ready source of foucusable power and telemetry I suppose too, everywhere, that presumably never wears out.

 Leading to bugging becoming terribly easy for cutting edge spooks and government agencies, no doubt. Because bugs can be embedded in any solid object and might become very difficult to ever find. 60 GHz is an awful lot of haystack to find a beam-forming needle in too!

That (earlier in the thread) company that makes RF absorbent fabric? (One could also use tinfoil, although I wonder how one would handle closures. Maybe with a magnetic closure, like microwave ovens?

I wonder if its suitable for making bags to put cell phones, computers, etc, in. (the devices most likely to be so "enhanced", perhaps even at the factory.)

With RF everywhere they might, at least in theory, interact with their mother ship even when off, with batteries and any attaching cables, power, etc. removed. If now there are enough IP addresses to give everything of note that has a small computer in it (maybe one of the $0.03 variety) its own routable IP address, and little web server or whatever, the sky is the limit. Mesh networks will be everywhere, and devices will pass along messages via the "internet of things" as long as there are devices in range of one another going all the way back to some physically or wireless Internet connected thing. So a place that is isolated from all this may become highly useful - certainly in testing.

[rhb]

Well, at least this is better than bags of charcoal and other nonsense.

The pyramidal absorbers are using the shape to effect a graded transition in impedance.  The earliest ones were horsehair coated with a conductive mixture. Varying the composition of a uniform layer will accomplish the same thing. And be much less fragile.  But will still need a 1/4" wallboard covering for durability.

Commercial TEMPEST grade chambers use plywood panels with galvanized steel sheet on both sides clamped into channels with bolts at very close spacing.  I'm using HVAC sheet because it is easy to get and cheap.  Tinned steel would almost certainly require buying a coil which I could not handle and would be far more than I needed.  I'd love to use copper sheet, but that would raise the price tag by $4-5K to about double the current budget.  The HVAC is about 1/3 the cost, the ESD flooring about 1/3 and the metal sheet and studs about 1/3.  The effectiveness of metal mesh drops off rapidly with frequency,so an actual screen room is not desirable.

With the seams fluxed with acid flux, a thin flat strip of solder placed between the sheets and some strong neodymium magnets to draw the sheets together I don't think I'll have any problems with a 150 watt iron.  I've got drawplates and a rolling machine, so bespoke solder is not a problem.

I'm *not* considering RF paint.  It's ridiculously expensive and wouldn't work well anyway.  *If* I do this I'll be spraying mixtures of sheetrock mud, powdered iron and charcoal with latex paint added to the mud as a binder to make the mixture less brittle.  Each layer will be 1 mm or so starting with the maximum amount of iron and charcoal and grading to a minimum.  It won't be very effective at low frequencies because it will only be 2-3 cm thick.  I was making fireworks at age 12.  I learned to recognize and handle explosive mixtures long ago.

Before I go to all  effort to coat the walls and ceiling I'll be testing the reflection transfer function with an 8510C.

Non-linear junction detectors will find anything except the great seal bug which was purely passive.  You can see Gene Hackman's character use one at the end of "The Conversation".  I think Snowden deserves to live in Russia for the rest of his life for revealing the USB cable bugs.  An FET smaller than a grain of silt and dipole of #40 AWG magnet wire.  Executed well you could never find them by physical inspection and so cheap to make NSA sold them by the bag to CIA.  A non-linear junction detector would find them, so the cat is out of the bag :-(

The LED lamps in my current space have a large Corcom EMI filter to suppress conducted EMI and 1/8" hardware cloth covering it.  It's effective enough it is now below the level of other EMI.  I stopped hunting and suppressing other sources when I bought more test gear and ran out of space, requiring a new lab area.

Have Fun!
Reg

[abinashsubash]

so did you make RF pyramidal absorber?
if yes can I know how
I have an exact requirement for converting normal RF shield room to anechoic chamber

[rhb]

I have not made any pyramidal absorbers and don't have any immediate plans to do so.

The purpose of the pyramidal form is to provide a gradient without having to make graded material.   There are lots of other ways to make an absorbing  panel.

If I were going to build an anechoic chamber, I'd use a mixture of drywall mud, latex paint, iron powder and charcoal with lightweight filler and spray it on the surfaces in graded layers with thin sheets of gypsum board to reinforce it every few inches of depth.  It's cheap, fireproof and durable.  More work though as you have to research the mixture to get the correct properties and verify it was fireproof.   But also vastly more durable a surface. 

Another option would be to blow compressed air into the mixture instead of the beads.

Some of the early commercial stuff was animal hair dipped in a mixture of  carbon and iron powder and then placed in pyramidal forms to harden.

Have Fun!
Reg

[rfguy2020]

I have also considering to make own pyramid absorber. I would like to have compact chamber for antenna measurements at around 900MHz. I think I would need about 300mm thick pyramid absorber. Making own absorber probably does not make much sense, but it is interesting topic. High quality absorbers from well known manufacturer are very expensive. Cheap absorbers from unknown manufacturer can be good, but there is risk that you don't get what you expect.

There is rather detailed document about manufacturing process (page 7) from ETS Lindgren:
http://www.ets-lindgren.com/sites/etsauthor/General_Brochures/Top%2010%20Anechoic%20Absorber%20Considerations.pdf
 
It looks like fairly simple. I did quick test from old mattress. I was using electric paint from Bare Conductive. I mixed some water to the paint. Then I soak the piece of mattress to this solution. To speed up the process I was making curing at elevated temperature. Outcome looked pretty good and feels and looks like commercial absorber. I didn't perform any RF testing, but the foam had similar DC resistance than commercial absorber measured with multimeter.

Cutting pyramids is a bit challenging with DIY tools. My absorber foam was shrinking a bit after curing, so it need to be cut from all sides. I think some kind of special scroll saw would be needed to cut shapes from the foam block.

Another challenge is to find thick mattress foam. I think 150mm foam is easy to find. I'm not sure, what would happen if layers of foam is glued together. I guess some solvent based glue would work in soaking process. Glue layer might have some impact to RF properties.

Final issue would be the characterization of own absorber. Probably the easiest way is to use the NRL arch method, which is basically bistatic measurement. The reflection of absorber compared with the perfect electric conductor. It would take big effort to make prototype sample for the measurement.

It would help if there would be proven formula of the soaking solution and curing process. I'm not sure if some iron oxide or graphite or some other material should be added to formula?   

[coppercone2]

may I suggest some short term investments in a financial market to get enough capital to do this with proper materials, these ideas are janky junk

something is not right when you are using mattresses as a building material as an improvement to loading a basement with bags of charcoal 

being in these rooms is NOT cool enough to warrant building a mattress crack fort, I would expect to find a faraday cage built of wire with little pieces of someones wall paper and alabaster on it in such a structure for some reason

[rfguy2020]

may I suggest some short term investments in a financial market to get enough capital to do this with proper materials, these ideas are janky junk

something is not right when you are using mattresses as a building material as an improvement to loading a basement with bags of charcoal 

being in these rooms is NOT cool enough to warrant building a mattress crack fort

You are right. It does not make sense to make own absorber. Same situation is with many DIY projects. It is better to use commercially available products. Sometimes it is just good to understand how things are done and learn from the process.

[coppercone2]

they are made in a factory with carefully tuned expensive spray guns and skilled operators making stuff when trial and error figures out the enormous amount of bullshit problems involved (a good spray gun costs 5k+ for painting a car even).

[rfguy2020]

they are made in a factory with carefully tuned expensive spray guns and skilled operators making stuff when trial and error figures out the enormous amount of bullshit problems involved (a good spray gun costs 5k+ for painting a car even).

Good point. You can do some rust repair for your car with cheap spray can, if you don't have much requirements for the quality.

[coppercone2]

even if you know what you are doing it ends up looking wrong, some car restorers and collectors are able to spot this kinda stuff from miles away, even if you used a cheap paint gun.. some how they see through the paint correction also

to cut foam anyway you use a hot wire cutter, https://www.amazon.com/Proxxon-37080-Wire-Cutter-THERMOCUT/dp/B0017NS8H6

The factory might have special giant specialty blades though

[rfguy2020]

even if you know what you are doing it ends up looking wrong, some car restorers and collectors are able to spot this kinda stuff from miles away, even if you used a cheap paint gun.. some how they see through the paint correction also

to cut foam anyway you use a hot wire cutter, https://www.amazon.com/Proxxon-37080-Wire-Cutter-THERMOCUT/dp/B0017NS8H6

The factory might have special giant specialty blades though

I don't think you can use hot wire cutter for polyurethane foam. Hot wire works for polystyrene foam. Polyurethane would make nasty fumes and cutting result is not good based on my knowledge. Also surface of the absorber will be more dense due to melted plastics. Mattress foam (polyurethane) should be cut mechanically as far as I know.

[coppercone2]

forgive me, I was focused on imagining someones family finding a satanic worship pit lined with several tons of charcoal and mattresses in the basement

[coppercone2]

looks like Matt Groening came up with this image, prophetic as usual. missed the charcoal though