Shielding of open TEM cell

Do you plan to self-certify for CE? Testing radiated or suceptability or both?

Personally, I'd take it to the mountains and work open-field.

You could get a big cardboard box and line it with aluminum foil, inside and out.

But a small shielded bag or box will have giant cavity resonances... lots of them.

tirsdag den 24. august 2021 kl. 16.58.35 UTC+2 skrev Klaus Kragelund:

stick an Ohmmeter on it, can you actually connector to the metalization? I think the tape would also need to have conductive glue

I'd check ebay et.al. for condutive fabric

Or build a box out of copperclad FR4 with soldered joints.

You could build the TEM cell that way too.

On a sunny day (Tue, 24 Aug 2021 16:58:27 +0200) it happened Klaus Vestergaard Kragelund snipped-for-privacy@hotmail.com wrote in <sg31ej$lsc$ snipped-for-privacy@gioia.aioe.org>:

How about a big cardboard box covered with alu foil?

Hello Klaus, A walk-in RF chamber would be the most convenient, I think, though its a serious investment usually found at commercial test labs or universities I don't suppose you have any friends or fellow IEEE members who would lend you access for the time you need? We have a chamber at our facility, in New York, that's often idle. Before we built that my RF testing colleagues used a large trunk-sized metal boxes. The connection inlet/outlet was itself a smaller box attached and joined to the outside of the big box.

At your local home center store take a look around, e.g. some HVAC ducting (galvanized tin), metal shears, sheet metal screws, ...? The box will need 100% seal i..e. high conductivity all around, and at the edges. The trick at the joined edges is to overlap the folds. To allow you to repeatedly open & close the box, a "piano hinge" could be employed with one sheet to form the door. You can use quick- release snap fasteners to hold the door tight during measurements.

Cheers, RS

Untreated, the boxwill be RF "echo chamber" with all sorts of resonances at different frequencies at different locations. Either, you (a) treat the inside, or (b) 'characterize' the behavior and subtract it out post-analysis. Treating the inside is easier in concept, e.g., energy-absorptive wedges. My collegues (also on a budget) used blocks of various sizes of conductive black foam (that ICs used to be stored/shipped in). Characterize+Post-analyis correction is tricky; it might only work up to 10GHz or so; your DUT setup has to consistent; and your analysis system has to provide a "curve correction" feature, etc.

Goofy idea: make a box out of multilayer PC boards, tiled maybe. Outer layers are ground planes. Inners are a maze of fractal antennas and termination resistors. If not totally absorbing inside, that would at least kill the cavity Q.

Next, build in the transmit/receive connectors.

Would that work? There are planar antennas on PC boards.

No, No! Use the leftovers for baking chicken in foil, a tasty asian dish.

You are basically trying to fabricate a faraday cage. For table-top use, one can fabricate a cage using copper window screen mesh bent over a thicker solid copper wire frame, with mesh soldered to the frame using soft (tin-lead) solder - just as for making copper gutters. Build a flange along the long side with a matching framed-mesh closing plate, with some means to clamp plate to flange.

There will also be an I/O plate studded with feed-through connectors.

And, to break up cavity resonances, make it strange-looking - nothing parallel to anything - a crooked little house

One can also lay sheets of RF absorbing foam or sheet on the floor to further suppress cavity resonances.

Joe Gwinn

I have access to a calibrated 3m semi anechoic chamber, and will be testing both emission and immunity. The reason I write relative measurements is that the TEM results cannot be used directly for evaluating 3m room measurements

Yeah would be best, but no way to do that. I cannot spend the time and move equipment back and forth is a no go

Yes, that is actually what I am doing. Next step is to create some walls in full room height with aluminium foil (for VHF) and metal cage on top (for LF). Also add EMC springs to openings and wire feedthrough filtering. But again, next step, when I get time ;-)

Yes. Some test standards actually use this intentional, with reflectors instead of absorbers

Yes, was my first thought. But spending the time on the box, I can just buy it since that will mean I can bill hours on other stuff

I have access to a chamber, but want my own setup so I can use it when I want and don't need to drive to do testing

Yes, I need overlap, no apertures allowed. I will measure the attenuation of the foil, since I may need double layer to get sufficient attenuation. Measurement is simple done with a antenna and emitter reference, then placing the wall in between

For the large chamber I will get reflections. For the small chamber, that is really just put on top of the TEM. The TEM is in it self a transmission line, an open one, so the small chamber is just to remove ambient noise, so reflections inside wont matter much

I am on a budget, so will be on a lookout for the absorbtion black foam.

Yes, the one I ordered may not be good. Working from home so I took a chance. Next step is a you say to go to the hardware store and probe the different types they have to be sure I can access the metal foil

Will do :-)

Yes, that was my first idea also. Kitchen alu foil may just do the trick ;-)

Very nice, that is a good idea, I was thinking a square box, but you are thinking "outside the box" ;-)

You can improve the flatness of the frequency response in a reflective box by having a large metal-bladed fan inside it. As it spins the resonances change in frequency. A bit of averaging then gets you a much flatter response than before.

John

some microwave ovens do that instead of having a rotating platter to get even heating

Very nice idea. I guess it needs to be very large though