RF shielding for large projects 2

[davorak]

My question is weather my recent purchase, Rose & Bopla 01234023 85 with rmi-efi gasket, was overkill for my project. What other shielding options make sense?

I have recently started a research project and purchased a box to shield my electronics
I will be making noise measurements ~-215dBm/hz. Frequency range:50Mhz to 3 Ghz, Though the measurement currently focus on f<1Ghz.

I have a similar set up working in a small Hammond aluminum diecast box.

For large rf projects what shielding options do I have? While I may at some point be able to shrink the components in the future the project currently can not be subdivided.

Thank you.

This is cross posted at rf cafe.

 

[Higgler]

when I've done antenna isolation measurements, I've always sealed all rf leakage points (covers and rf connectors and cables) with Chomerics silver paint. I wouldn't even try it without sealing. Metal taping covers is far inferior to silver paint. Metal tape with silver paint at the edges of the tape is a good idea to minimize the amount of silver paint used and it provides a better rf seal than trying to fill a larger gap with silver paint alone.

even though you have an aluminum box, even 0.1 mil gap in a cover will radiate/leak rf. Gaskets help, but they aren't always the best solution for a total seal.

As for internal coupling between components, placing them in a sealed box increases coupling. If internal cross coupling is a concern, add rf absorber or Resistance card inside boxes to reduce internal coupling (if all rf sources aren't sealed perfectly). One cheap source of resistance card is car window tinting film, it's about 20-40 ohms per square.


kch

 

[davorak]

Thank you Higgler, I thought about asking about copper tape vs other sealants.

I made some mistakes in my calculations above so I will correct them here:
You may be right I will tell you how I reached this number.
I am measuring shot noise.
Pn = 2 e I B *R (current noise in a 50 ohm terminated system)
at one micro amp, a bandwidth of ~400 Mhz, 50 termination gives ~-112 dBm and
~-198dbm/hz.

Thermal noise should be
pn = kb T B
Gives ~-88 dBm and ~-174 dBm/hz

What economical enclosures exist for large rf projects.

 

[Comcokid]

Economical enclosures? Well, I have used 1 gallon paint cans (clean new ones) where I wanted to shield a source or device for tests. A new paint can with clean lid, some copper tape or alunimun foil, good feed-thrus can provide very good isolation.

But the Rose/Bopla enclosure you listed is a bit larger. I wonder if they make tin plated 5 gallon cans with removal lids, or 55 gallon drums?

 

[davorak]

Thank you Comcokid, I will keep the paint can Idea around from some smaller side projects.

My current definition of economical is a price lower the the Rose & bopla enclosure(~$300) with approximately the same size. I am planning on opening and closing the box on a semi weekly, which is an important design decision. Using tape and paint each time will be a pain if it is absolutely necessary. I should be able to try the new box out some time next week.

 

[Higgler]

Tape and paint once a week isn't too bad. The paint on Chomerics has the consistency of toothpaste, so it comes off as easy as scraping toothpaste off the bathroom sink, and also goes back on easily.

Gasketed covers work, but after many off/on tests, if you really need super high isolation, you may be suspect if the gaskets have compressed too much.

Watch out for Aluminum Oxide buildup too, limits gasketing success.

kch

 

[biff44]

Paint can, huh? That is an EXCELLENT idea. I will have to keep that one on file!

I once used a big coffee can to build a very high Q resonator for use in a ultra low phase noise measurement apparatus. The hard part was getting the coffee out without opening the can(drilled two holes for coax connectors, and put a vacuum cleaner hose on one hole and sucked out all the coffee). I could even tune it by slightly bowing the top of the can with a micrometer.

http://www.MaguffinMicrowave.com

Maguffin Microwave wireless design consulting

 

[Comcokid]

I learned the paint can idea from a very experienced RF engineer who designed devices for the automatic meter reading market (wireless power, gas, and water meters) He had about million dollars in equipment available including a large GTEM cell, but still had paint cans at hand.

Worked well for me when I needed to test a number of 802.15.4 type transceivers, and only wanted them to mesh together in a fixed way to test the networking method and software stability. Put together an array of 1qt cans linked by coax and attenuators.

 

[davorak]

Thank ofr the tip on Chomerics the sliver paint that I have now is definitely not that thick that should make things easier.

 

[Higgler]

We add MEK to the silver paint to thin it out. You can have very thick or very thin silver paint. It won't be as thick as toothpaste but I've built up to 50 mils worth before with multiple applications.
It's for sealing the tape edges, which means you don't need much thickness.
Q tips or cheap paint brushes for added neatness.

Chomerics has a spray paint for $140 per can if you want it to look prettier. Spray from the paint can dries hard and thin, just like normal spray paint.

kch