Are you trying to shield for WLAN security, prevent interference for other devices? Are you looking to shield an entire room, window?
Why not try woven copper screen material, (much like screen door material)? I used a shield cage (8X8 room), made from this material for tuning/testing RF components, and it provided adequate shielding from outside interference. This material could be applied to windows, in wall interiors, etc. for shielding.
OK, well let me clarify then..
I need to sheild a square plywood box (8 cubic feet) from a WLAN. So, it's a small sized area in comparison to a full sized room.
It seems you are saying that a highly conductive metal will work best.
I need the sheilding for experimental research, so when I use the civilian term "completely" it's because any leaks of signal make results invalid.
I'm not sure exactly sure what the signal strength of an average consumer wireless router is in terms of dB, but that is my aim.
I apologize for my lack of knowlege.
You may need to think this through a bit more. Let’s suppose you put copper foil all around the outside of the box. That will make a pretty good screen provided you overlap the foil at the joints to get a nice low impedance joint. Now what? You can’t see into the box and you can’t get power into the box. Ok , so you put a wire into the box, just a simple wire not connected to anything either inside or outside. You have now taken 100dB shielding down to perhaps 10dB to 20dB. Oopps! The internal to external communications will be your biggest hurdle. However before you can even do any sort of ball park calculations you need to know some signal strengths. If you don’t know and can’t measure the field strength outside, and you don’t know or can’t calculate what field strength you require inside, then the whole project is either going to work by luck or require mystical intervention.
Put you thinking head on and do some sums. Then you can work out what sort of feedthroughs you will need on the power and control lines (if any) and how big a window you can fit if you want to see what’s going on.
You may be better to just hire a consultant on this one.
One USB cable will most likely be the only thing going in and out, serving as data transfer and power.
Are you trying to tell me that this one cable will reduce the shielding capacity 80-90%? Why? That's crazy.
Anyway, I'm just asking about materials/procedure. Since, you really have no clue what I'm doing, it's hard for you to predict the outcome. haha
It seems like the consensus for the best low budget shielding for something like WLAN is copper sheeting of some sort.
Also, Thanks for attempting to talk to and put-up with someone without some sort of degree in Electrical Engineering.
I know it's hard for you, logbook. haha
Exactly. Except in a very few circumstances, don't use percentages when discussing relative RF levels. It will negatively affect your ability to understand the basic concepts. RF power can be a 'qizillion' watts, or it can be one 'qizillionth' of a watt. Numbers like '180 dB' (or more) are often used in RF work; this is too large a range for percentages. Thus the log-scale dB. Long live the dB.
What sort of measurements are you trying to accomplish in this shielded box? Could you simply make the measurements late at night when the other system(s) could simply be shut-down for a few hours? Would 20dB isolation be enough to separate (100:1) your measurement from the external interference? How accurate do you need to measure?
For one USB cable, you could have the cable enter the shielded box via a long & narrow copper tube that is firmly grounded (clamped and/or soldered) to the shield wall where it enters. You'll need to carefully splice the USB cable to get it in the narrow tube. Maybe add a layer of common braided shielding right up to your test equipment and clamp it, using a hose clamp, to the test equipment connector ground. If there is any USB cable left exposed at either end, then add some (appropriate mix) ferrite snap-on cores.
You should be able to prevent external RF getting in - at least to a large number of dB (probably at least 60dB, perhaps much better). It should be a very simple matter to confirm the isolation you've achieved using your test equipment.
VB1BLL,
This is an interesting argument you are presenting here. What you are saying is that if you feed a wire through a “waveguide beyond cutoff” then no signal will be able to get in this way. I am not convinced by this, although I have no experimental evidence to back it up.
If you put a wire down the middle of a (circular) waveguide-beyond-cutoff then aren’t you just making a coax cable? Whilst a free wave can’t propagate very far down a narrow waveguide, surely a quasi-TEM wave could, just as it does down a coax. It may “confuse” the RF a bit, but I would think the stuff will still get in to a rather large extent. After all, why would people bother making these expensive power feedthru boxes for the mains if you could simply use a cheap and cheerful waveguide solution.
I wasn't exactly thinking of 'waveguide beyond cutoff' in this case due (as you point out) to the cable. However, the more of the USB cable that can be covered, the less signal will be picked-up by that which remains exposed. Also, the braided shielding and ferrite that I mentioned are important for those parts of the cable that are not covered by the copper tube.
The approach given should be sufficient for adequate, but not perfect, isolation.
I'm still curious about what exactly would require this sort of approach...
You will get very good shielding by covering the box with an unbroken sheet of aluminium or copper and even better if both inside and outside are covered. You seal the seams for RF with EMC tape (eg. Chomerics tape).
The problem with any shield is access panels and cable entry points. Any break in the shield will 'leak', even the tiny slit a few inches long between overlapping sheets of metal can have significant RF leakage. In a temporary installation it's not too difficult if you are willing to reseal openings with EMC tape each time the seal is broken, otherwise you'll need beryllium copper finger-stock material or conductive gaskets at the openings.
Cable entry must go through properly installed EMC grounding connectors and filters / shielding used to block RF from that path.
Attention to detail is important, eg. a length of double shielded coax cable may have 60dB shielding when properly terminated with coax connectors, but terminate one end of the shield with a 2" pigtail wire instead and the overall shielding drops to around 20-25dB.
Wow, now we're getting somewhere. Thank you for your contributions. A question, though.
Would you consider a signal leak through the cable (when shielded as you are describing, or as near as possible) to be negligible or something reasonably noticeable?
I’m just trying to get a feel for how devastating a leak of that nature would be or if I would even be able to pick it up on networking equipment. (assuming, again, that I’m using one of the said methods of shielding the wire)…
Running a wire through a long thin metal tube changes it from being a waveguide beyond cutoff (which gives nearly total RF blocking) to another form of a coax cable which gives very little (unless you use it and terminate it as coax).
There are so many variables associated with this USB cable, what it's connected to and how. What 2.4GHz field strength you need to block and what you can tolerate in the chamber and do you need ventilation, that you need some local professional assistance.
The shield on the USB cable (yes you need a shielded one) should be connected so its complete circumference makes contact with the outer shield on the box. Some ferrite beads over the cable will add loss to any 2.4GHz signal carried on the shield. After that if you need any more suppression it becomes hard.
Getting "complete" shieldieng -- or even approximating it
for all practical purpose --is not a trivial task.
The shield of an external cable shouldn't be connected
to the internal side of the shield of the box on the inside
If this is just for a temporary use, you may consider
removing it to a quiet environment, else you may
use fiberoptic cable to communicate and filter the (&%##@
out of the power. This seems to be the simplest and
most predictable solution -- the best solution is to
avoid the need to solve the problem.
