DIY low cost RF absorber for a 4' x' 4' x 4' anechoic chamber ???

[coppercone2]

I think it would be really worthwhile to look into ferrite powder prices to do a real comparison if you are going that far, maybe send some emails
g2g

[rhb]

I found iron filings for about $3.40/lb in the US delivered on eBay.  Ferrite powder around $150/lb from the one place I found that had prices.  But I need to research the materials.  It might well be that a usable ferrite is being sold as sand blasting media or as sandpaper.

[tautech]

Old thread that may offer some tips and guidance:
https://www.eevblog.com/forum/testgear/emc-chamber-build-log/

[ogden]

Old thread that may offer some tips and guidance:
https://www.eevblog.com/forum/testgear/emc-chamber-build-log/

Very interesting report. Thank you!

You mention ferrite used in this post:

When its all complete, 1.5 tonnes of ferrite tiles will cover every inch of the inside.

After a quick google search I think I have these ones:
SFA600 or SFA600A
http://www.samwha.co.kr/SW_catalogue/catImage/37/Ferrite_Absorber.pdf
Each individual tile is 100mm sqaure. They are mounted on a board in a 6x6 grid.
I'm told each tile costs $5 US direct from the manufacturer,  don't know how much we paid for each.
I worked out (roughly) that we would need a little over 10,000 tiles to fill the inside of the chamber so $50k just on tiles would not leave me with anything for the equipment.

[coppercone2]

bulk powder of shady grade from china might be alot cheaper then sintered tiles with specifications.

You can just make the cones bigger I think with little consequence so long you stay in certain frequency bands. Ferrite is mainly designed for inductors I think, so using it as a general 'i dont want reflections here' might allow for loose specifications.

With a ferrite tile they probobly sinter it and measure it and there are rejects due to process errors, inspection costs of initial powder and finished product, chipped stock that does not sinter properly, bad distributions of ferrite, improper geometries, mold maintence costs, sintering cost, furnace upkeep costs, factory calibration costs, stock maintenance cost (they are kind of difficult to store as they are heavy and fragile) skilled ferrite technician costs, shipping cost and then multiply cost by 4-10 for company profit depending on how niche it is. If they sell a tile for 5$ that means probably ~1$ of raw material went into it (or maybe much less).

since we are spraying it in a garage with ceiling material some liberties might be allowed to reduce costs, I kind of imagine it showing up in a giant 'chemical bag' like you get seed, rocks or industrial chemicals in, the one that you basically lift into the air and cut it on the bottom made of nylon.

With some kind of home made antenna tester I think what you are looking for is a way to detect which signals are bogies so you can ignore them in your head rather then try to get some kind of universal 'every joe on the street will know what this graph means' thing going on, especially when you are using it as a calibration/reference standard against a hill you found outdoors.

I don't want to stifle discussion of high end solutions but I don't want to see this degrade to cave technology because someone found some price of some manufacturers product, we have yet anyone with experience in the ferrite tile or anechonic pyramid/cone manufacturing field to tell us how sane or insane we are, but I have yet to find some kind of encouraging ferrite price war between two companies undercutting so it might be a stagnant niche.


To put prospective on it, he is getting 3000lb of tiles for 50k, meaning that the material finished inspected certified all that jazz is 16$ a pound, vs the 160$ a pound of raw unworked material someone found.

That means 3000 units get you a bulk discount of 90% of finished product. Thats kind of strange imo. I figure that 150$ a pound is worth something like <<20$ a pound, 90% for 10,000 units is a bit of a steep curve for bulk buying IMO, 30-50% price would be a good deal @10,000, and thats finished product. I would like to know more about how those factories run so I can understand the cost reduction with volume that occurs. It makes sense to me because you get 16$ a pound finished for 10,000 units at say 50% discount, so a single unit might cost like 32-40 dollars, and thats a damn finished product, with 4x markup on materials say, so thats 8-10$ a single pound, but they wanna charge 150. Even if you multiply it by a really bad 400% small guy bullying/inconvenience cost, you still get like what, 40$ a pound?

150 a pound is cringe, even with a 400% bullying small customers cost. I bet you can get chinese stuff at 5$ (hell or less) a pound for 50 pounds.

The powder screening process is probobly throwing it into a ball mill and paying someone with a particle mask 12.50-15 an hour to beat a seive with a rug beater to get 'engineering particle grades' and some tech 17-22$ an hour to look at a sample of a 100kg bag of it under a microscope (they probobly get the raw materials from china too, off alibaba)

I think people are just willing to pay because the people that are typically interested in ferrite tiles would be rather well educated RF engineers that are all neat, understand mathematics, nice clean temperature controlled office etc. I don't see them wanting to fuck around with giant bags of dust and 'kiln conditions' that bring the movie 'deer hunter' to mind. It seems that even as far as metalurgical interest goes, making ferrite is the bottom of the barrel as far as work conditions.

Before someone says I am saying RF engineers are posh, I just mean its a dirty ass nasty job. My friend works in a factory that has a kind of department related to heat treating, kiln work, etc.. and it pays more to work in that location. Most people don't wanna do it, blue collar guys like mill wrights, stock workers, tool operators. Apparently it feels like walking into hell. But it means a hell bent individual can save alot of money.

[rhb]

Ferrites are probably more in line with geology. They are sort of between ceramics and metal.  Before I strayed from the strait and narrow and went into geophysics, I was trained as an igneous petrologist and became quite familiar with phase diagrams of solid solid solutions of things like garnets and spinels.  It is highly likely that there is an inexpensive alternative to expensive "optimal" materials.

The chief difficulty is likely to be that the purveyors of cheap materials will know little about the composition of what they sell.

[coppercone2]

I recently found out that you can do x-ray metal analysis at home very cheaply but it won't give you insight into the grain/crystal (what is it?) structure, though maybe you can sinter a small block of it and then polish it and do a acid etch and look at it on a metalurgical microscope to get some kind of idea. The videos on youtube show that you can get a elemental composition from a metal sample using a small home made benchtop device.



https://www.ebay.com/itm/AmpTek-XR-100T-CdTe-Gamma-X-Ray-Detector-w-PX2T-CZT-Power-Supply-Amp-MAY-BE-NEW/132718107889?hash=item1ee69d9cf1:g:Ca4AAOSwBrlbMQld

I thought it would run at least 50,000$ for one.

http://www.gammaspectacular.com/phpBB3/index.php

Beyond that I think you would need X-ray crystalography right?

[rhb]

Barton garnet media is pyrope (Mg3Al2Si3O12) and almandine (Fe3Al2Si3O12).  These form a continuous solid solution series with spessartine (Mn3Al2Si3O12).  So it is highly likely that there is garnet sandblast media with a Mn,Mg,Fe composition very close to that of  the expensive material at prices dominated by shipping costs.

A Von Lau camera will handle the X-ray diffraction albeit a bit tedious to interpret.  That just requires an X-ray source.  One could also use a petrographic microscope and immersion liquids.

At the end of the day though, what matters is the lossiness of the material and that is best measured electrically in the same manner used for core materials.

[cdev]

I used activated charcoal from several spent BRITA water filters and off the shelf "Great Stuff" foam. I tried to make it densest in the back where it touched the metal. It was used by being poured in layers into a round metal cookie box. Its very hard to work with because if it gets into any of your clothing or on your skin its soon hard as a rock when it sets and its very hard to get off, and it hardens quickly.

Even if you use a LOT of carbon its not that conductive. Iron filings would probably help there. (not ferrite, iron)

The idea was to make an RF absorber to go underneath a spiral antenna. I think I used four thin layers (15 mm each?) of the black(really black+pink) foam. Once it sets its not possible to change it. It was flat, not shaped. I still may have this thing somewhere. I'll try to find it, I haven't seen it in a while. My wife quite possibly may have thrown it out. (Understandably)

I've also used anti-static foam in a small plastic bag experimentally to see if LNAs were oscillating. (If it is pressing it down on top of the active device  stops it, changing the current drawn.) thats a useful thing to know.

[David Hess]

I could see using a ball mill to grind appropriate materials to be added to a binder and testing that for absorption but I have no idea how to extrude a consistent foam.

[cdev]

The foam is expanding as it leaves the nozzle and then you have to rapidly mix it and pour it into something else (or mix it in the container which is what I tried to do. This is extremely difficult to do fast enough to have it mix throughout the foam. As the goal of what I wanted to do involved needing some precision (I wanted to see if I could use this in a (receive only!) GPS antenna) I considered the whole effort not really at all successful. The signal strength was higher with the cookie box (because it was acting as a reflector- probably!) but I think the antenna was more accurate without it. Just in free space. The spiral antenna was a fairly good GPS antenna without any additional metal around it. (except for the feedline going straight down and to the side after a bit)

I think the earth in my back yard with its natural increase in moisture going down made a better RF absorber.  Which makes me suggest. (I think the best testing environment thats likely available to you and the rest of us is some open space - like a field or a parking lot.)

And I found that the granular carbon I had was not that conductive. If you squeezed it the resistance would go down markedly (this is the principle that telephone mouthpieces used to use) but I think some kind of carbon fiber would likely work best, not granulated carbon like I had. What I wanted to try was many layers of different batches of conductive antistatic foam pressed together. I tried to do this with some carbon foam I had but I didn't have enough of it.

[rhb]

I think adding foam beads to a binder is more practical.  In principle you can inject air bubbles into a binder.  That's done with concrete.  But cleanup with a non-water soluble binder would be a real headache and if you don't achieve the right viscosity, the bubbles just migrate to the surface.

 I also think loss  testing small samples is probably the most efficient approach.  IIRC there was a thread in a VNA group which referenced an HPAK fixture for testing toroidal cores and a link to an article about solenoidal cores.

[coppercone2]

Yea I have serious concerns that trying to make voids without the use of some kind of filler media is gonna be a pain in the ass process. There is probobly a reason those hydrogen bomb plasma foams cost billions of dollars to develop.

Hmm, if you grind up foam, there is no reason you can't sinter a plastic mixture right?

Another binder idea is maybe using sand or anything else really soaked in sodium silicate, which is hardened by CO2 gas diffusion to make a glass like binder thats not time sensitive or nothing. You can store this stuff premixed pretty much indefinatly, put it into a mold and fill it with gas and it will begin to solidify, you just poke it with a fine needle to inject gas into it if its really deep. It makes for good metal casting molds. Also high temperature stability, inflammable and fairly hard but it will crumble if you start banging on it hard. If you set it right you need to basically use a chisel to break it.

[E Kafeman]

Think about the wall absorbents as antennas.
If the wall should absorb and not reflect must it be impedance match.
Using ferrite/iron/coal powder will create a reflecting wall if wall impedance and wall structure not is matched to the space wave.
To achieve impedance match and not create a reflector is a certain deep relative lambda of absorber thickness needed as for most antennas.
That is why cones-shapes often are used as a kind of successive impedance matching, but these cones are still frequency and depending on angle of RF signal to it should be able to absorb well. If checking inside a chamber can it be seen that often different lengths of cones are used in middle of chamber compared to its ends. Size and angels  and amount of absorbent in the cone decides its optimal effective frequency range and usable angles.
Ordinary coal is the most common material. In some cones is coal density different at different heights.

If only near-field magnetic field are needed to be absorbed can ferrite tiles with high resistance for actual frequency be used, no need to give them any impedance matching shape.

Open field measurement is so much simpler. Open field distance should have an length of 5-10 lambda. Reason to use open field is to avoid all none planar waves. That includes reflections in ground between Rx and Tx antenna, That is why a certain measurement height is needed.
VNA with gating function can also help a lot. Also using directive measurement antenna reduces unwanted reflections to be mixed with direct wave.
Wide-band horn antenna is often used for that purpose.
A very simple way to avoid ground reflections is to measure vertically in open field upward against open sky. This kind of measurements can be done even in a rather noisy urban environment as long as own signal can be identified, transmitting a relative short distance and using a directive antenna. In worst case, reduce measurement bandwidth to as low as possible, 1-10-100 Hz and let any frequency sweep be slow enough.

[coppercone2]

How do you work out the geometry with the cone shapes and angles based on distance from a radiator? The great thing is if we make big molds we can cut whatever cones we want out of them with simple tools like sawzall.

How can we categorize a cone that we make?

Also if you mix up a foam plastic/other mixture, maybe you can modify a 3d printer to grow them with layers using a infrared heat source to sinter it.

If its too hot you will smoke cones, or maybe you can smoke cones to your advantage because char is kinda conductive, add some saw dust in there and hit it with a laser, or pump smoke from an acetylene flame into it to paint it as its doing layers, so the cone will be smoked

maybe you can just set a tire on fire near the 3d printer and blow the fumes at it with a fan, who knows

[E Kafeman]

Cones need to have a controlled and even density of coal and a thickness enough to be absorber at particular frequency.
See thickness as a kind of successive attenuation. If impedance is 377 Ohm, fine, then just let cone be thick enough to absorb as much of reflection that remaining reflections energy not interfere with main planar wave. If thickness is much less then Lambda or impedance is way of from 377 Ohm will it work less good.
Suppress unwanted reflections with more then 20-30 dB in whole chamber is more or less impossible but if chamber design is good will there be a zone with low amount of reflections, known as quite zone, where a dynamic range of 40 dB or more is possible.

Geometry of cone is depending on actual needs. Calculate its shape and impedance can be done in most RF design software such as CST..

[coppercone2]

How exactly do you define the impedance of the cone? 377 is the impedance of free space.

If the thickness is much less then lambda - is this why I sometime see the cones with the top cut off to make flat faces on top in EMI chamber pictures?

I have trouble relating what you wrote to a cone covered surface.

Can HFSS do what you want, if I modeled a foam wall in solidworks and imported to HFSS?

[coppercone2]

Can you tell us how to measure a foam sample to get the data relevant towards a HFSS or other program simulation?  Can you wedge a block of particular dimensions into a VNA somehow? I just built my first heatsink in solidworks and it was remarkably easy, I found a way to import my solidworks models into HFSS, and youtube can probobly tell me how to run a 3d simulation, but I assume that HFSS will have some kind of table or graph import or something related to the material. How do you get this data from the laboratory. What equipment/setup do you need?

I assume it would be something like setting up a foam sample of particular dimensions connected to some kind of test equipment, either by direct connection or some kind of antenna system (like facing horns).

[E Kafeman]

In CST/HFSS do you define each structure and shape before resulting RF fields can be analyzed.
It is you that provide material data such as foam conductivity and dielectric constants, same for coal and coal density.
Basic material data can either be found in books or internet or you have to measure it yourself.
How to do measure, search VNA+dielectric measurements. It is well described at R&S and Keysight websites
Some providers of absorbing material do have very informative data sheets that can be of help when calculating your own absorbents.

[cdev]

Does anybody know what the properties of carbon fiber are with RF?

[E Kafeman]

Carbone fiber is a wide range of different materials. Fiber size and type of filler can vary in a rather wide range.
Measure a sample of your DUT is recommended. A lot of different types of carbone fiber are measured and results are published but results are nothing more then typical indicators. Some results here.

[babysitter]

My idea of testing material suitability:

Get basic directional antennas for the interesting range and go to the "hilltop" test site.
Make samples that will "block" the estimated main lobe. I would expect a good sample to be the one that doesn't act as a effective reflector for a "backfire yagi" and can get to the antenna proximity while changing S11 less than others.
When building your chamber, put your tiles/cardboard pyramids/whatever  most dense in the expected main lobe, fill up when you can afford it. Consider a air gap behind the absorbers in your chamber, so reflected radiation can enter its demise from the backside too. Movable absorbers can be pushed around to see if they act better somewhere else.

[Wolfgang]

Hi,

when I was young at university we hacked a chamber from egg cartons covered with a laquer made from shellack and powdered iron. After drying, we applied a layer of graphite spray normally user for the repair of CRT tube shieldings.

To our utter surprise, it worked quite well up to the 13cm band.

[rhb]

Some excellent suggestions. Thank you.

My interest is investigating  multiband HF  dipole designs by scale modeling in addition to using CEM tools. Not sure it will work, but I thought it would be an interesting experiment if I can keep the cost low enough.

I am a huge fan of the Amateur Scientist columns  by C. L. Stong.

[heyecanyoq]

Hi. Did you accomplish building your anechoic chamber finally?