Author Topic: EMI case shielding  (Read 13146 times)

0 Members and 1 Guest are viewing this topic.

Offline peperTopic starter

  • Contributor
  • Posts: 16
EMI case shielding
« on: March 27, 2012, 02:05:06 pm »
Hi,

What factors determine the required properties of shielding for a case?  Specifically I want to add shielding to a wooden PC case but I'm having trouble finding sufficiently reliable information.  I've found other forums where people have asked the same question but the answers vary from 'completely unnecessary' to 'like being able to start a microwave with the door open'.  The answers are generally not backed up with any specifics (or links to further information) so I thought I'd ask here where I feel the people are a bit more knowledgeable.

I'm hoping someone could give me a few pointers in the context of a PC case and the shielding normally found therein.

How much effect does the thickness have?  Is aluminium foil too thin to have much of an effect?

This is a one-off, not something to be mass-produced so it won't be submitted anywhere for testing but cases aren't normally shielded for fun so I'd like to give it my best effort without going overboard (or at the other extreme wasting time doing something that will have little or no effect).

Many thanks in advance!

Regards,

Richard.
 

Offline jahonen

  • Super Contributor
  • ***
  • Posts: 1054
  • Country: fi
Re: EMI case shielding
« Reply #1 on: March 27, 2012, 02:50:53 pm »
The most difficult aspect is the filtering of the signals passing through the enclosure. That is probably the most difficult issue, making EMI-tight enclosure is pretty simple, just put everything in a conductive case and seal it tight. Unfortunately, that is usually quite useless since one generally wants to interact with the device inside :) One cable can ruin your otherwise good EMI shielding, noise conducts to outside world with it.

Enclosure itself should not have any long slots. And you would want the inside surfaces to be as conductive as possible. This maximizes the reflection loss. The conductive layer does not need to be very thick due to skin effect at MHz+ frequencies.

Adding aluminium foil to existing enclosure does not work at all (just because it is difficult to make it seamless) and can just make things worse instead.

For good illustration how one makes a proper shielded box for electronics, take a look inside microwave/RF equipment, they are probably quite adequately shielded :)

Regards,
Janne
 

Offline sacherjj

  • Frequent Contributor
  • **
  • Posts: 993
  • Country: us
Re: EMI case shielding
« Reply #2 on: March 27, 2012, 02:54:32 pm »
Adding aluminium foil to existing enclosure does not work at all (just because it is difficult to make it seamless) and can just make things worse instead.

Copper foil can work fairly well, as you can run a solder bead down it to complete the Faraday cage.
 

Online Mechatrommer

  • Super Contributor
  • ***
  • Posts: 11640
  • Country: my
  • reassessing directives...
Re: EMI case shielding
« Reply #3 on: March 27, 2012, 03:01:52 pm »
apart from janne's answer on very good technical points...

1) What factors determine the required properties of shielding for a case?
2) Specifically I want to add shielding to a wooden PC case
3) but I'm having trouble finding sufficiently reliable information.

1) fear factor
2) what?
3)because it depends... on what you want to do... you want to shield inside out, or outside in?
i can run my PC naked if i want to. did that many times and never have a single problem ;)
so what actually the problem is?
Nature: Evolution and the Illusion of Randomness (Stephen L. Talbott): Its now indisputable that... organisms “expertise” contextualizes its genome, and its nonsense to say that these powers are under the control of the genome being contextualized - Barbara McClintock
 

Offline peperTopic starter

  • Contributor
  • Posts: 16
Re: EMI case shielding
« Reply #4 on: March 27, 2012, 04:37:21 pm »
Hi,

Thanks for the replies so far.

The idea I had was using aluminium foil and conductive tape along any joins.

I understand the problem about holes and slots but I'm curious about how this is handled in a normal PC case.  If I put (for example) a DVD drive in, I'm creating a big hole.  Even if you're only talking about the endplate on the tray, it's still a big hole.  Is this where the manufacturer argues that the case alone meets shielding requirements, but if you remove a blanking plate to put in a DVD drive (even though that's what the bay is there for), it is a modification beyond their control?


1) What factors determine the required properties of shielding for a case?
2) Specifically I want to add shielding to a wooden PC case
3) but I'm having trouble finding sufficiently reliable information.

1) fear factor
2) what?
3)because it depends... on what you want to do... you want to shield inside out, or outside in?
i can run my PC naked if i want to. did that many times and never have a single problem ;)
so what actually the problem is?


1) So why are normal cases shielded?  Is it purely to meet requirements for sale?
2) It's an HTPC.  I couldn't find a case that met my requirements, so I made my own.  That's a different topic though.
3) I would imagine a normal case is shielded to stop the computer potentially interfering with other devices.

The company I work for makes high-end head units for cars.  I remember being told about a problem in one of the test cars where the radio reception did not meet specification.  Apparently it was because the aerial was built into the rear window and a break-out tab had been removed on the head unit case to gain access to a debug connector and then not covered again.  The head unit is in the dash board looking out between the driver and passenger seats at the back window.  The hole in the case was enough to cause sufficient RF to 'leak' out and interfere with the effectiveness of the aerial in the back window.  I think it was only an intermittent problem when the radio signal was weak anyway but the shielding was obviously necessary in this case.

I'm not desperate to add shielding to the case for the sake of it.  If the amount of RF emitted from a normal PC is minimal and the effort required to shield it substantial, I'm happy enough to leave it.  My problem is just that I don't have any experience in this field to be able to balance "it's not necessary" against "so why are PC cases shielded?".

Regards,

Richard.
 

HLA-27b

  • Guest
Re: EMI case shielding
« Reply #5 on: March 27, 2012, 05:23:12 pm »
Hi peper

Wooden case is a neat idea, hope it looks as good as it sounds. To answer your questions, here are a few thoughts...

- Your case would probably work without any shielding whatsoever. If (actually when) it does consider your problem solved and move on.

- The minimum thickness of the shielding material is determined by the skin depth at the frequency you want to isolate, the higher the frequency the smaller the skin depth becomes. Skin depth is also affected by the conductivity of the shielding material.

- Holes, slots, discontinuities and especially conductors passing though the shielding material severely compromise the shielding effectiveness to the point of rendering it non-existent. This is why I think even metal PC enclosures have no shielding effectiveness whatsoever therefore board manufacturers do not depend on it, they design enough shielding in the PCB layout. This is probably why you can do without it.

- Aluminium foil is not a good shielding material because aluminium always has an oxide layer on the surface which is non conductive. When two pieces of shielding material meet at a seam it is crucial to have electrical contact along the entire seam. With aluminium there is no way to be sure. Your best bet are materials which you can solder, copper foil, tin foil etc. are good.

- Leslie Green's book Analog Seekrets which is highly acclaimed on this forum, and is available for download on his site has a chapter (14.4) on the topic.

- The Circuit Designer's Companion by Tim Williams deals with shielding and grounding very comprehensively. Here is a short excerpt
Quote
Aluminium is used throughout the electronics industry as a light, strong and highly
conductive chassis material ? only silver, copper and gold have a higher conductivity.
You would expect an aluminium chassis to exhibit a decently low bulk resistance, and
so it does, and is very suitable as a conductive ground as a result. Unfortunately, another
property of aluminium (which is useful in other contexts) is that it oxidises very readily
on its surface, to the extent that all real-life samples of aluminium are covered by a thin
surface film of aluminium oxide (Al2O3). Aluminium oxide is an insulator. In fact, it is
such a good insulator that anodised aluminium, on which a thick coating of oxide is
deliberately grown by chemical treatment, is used for insulating washers on heatsinks.
The practical consequence of this quality of aluminium oxide is that the contact
resistance of two sheets of aluminium joined together is unpredictably high. Actual
electrical contact will only be made where the oxide film is breached. Therefore,
whenever you want to maintain continuity through a chassis made of separate pieces of
aluminium, you must ensure that the plates are tightly bonded together, preferably with
welding or by fixings which incorporate shakeproof serrated washers to dig actively
into the surface.
 

Offline peperTopic starter

  • Contributor
  • Posts: 16
Re: EMI case shielding
« Reply #6 on: March 27, 2012, 07:26:44 pm »
Wooden case is a neat idea, hope it looks as good as it sounds. To answer your questions, here are a few thoughts...

- Your case would probably work without any shielding whatsoever. If (actually when) it does consider your problem solved and move on.

Hi,

I'm quietly pleased with how it's shaping up considering I'm doing it the old-fashioned way (i.e. no laser cutter or CNC).  I still need to do the front panel and then paint it (I've seen some nice looking cases that emphasize the wood but that wouldn't really fit in with the rest of my stuff and you can get away with a few imperfections when it's painted  :) )

I'm less concerned about the computer not working (I know it'll work without shielding), I was more concerned about it affecting other equipment.

- Aluminium foil is not a good shielding material because aluminium always has an oxide layer on the surface which is non conductive. When two pieces of shielding material meet at a seam it is crucial to have electrical contact along the entire seam. With aluminium there is no way to be sure. Your best bet are materials which you can solder, copper foil, tin foil etc. are good.

- Leslie Green's book Analog Seekrets which is highly acclaimed on this forum, and is available for download on his site has a chapter (14.4) on the topic.

- The Circuit Designer's Companion by Tim Williams deals with shielding and grounding very comprehensively.

That's great stuff, thanks!  I've just downloaded Analog Seekrets.  I'll have a read as soon as I get chance (looks like I'll need to read some of it several times before it really makes sense, but that's ok!).

I think then there is probably not much point trying to shield the case as I probably wouldn't be able to do effectively anyway.  If nothing else, the holes on the sides for the fans would probably nullify my efforts immediately.  The fan guard is almost certainly too coarse.  I tried using mesh but the air being sucked through it made it too loud.

Anyway, thanks for the brief explanation and the book tips!

Regards,

Richard.
 

Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: EMI case shielding
« Reply #7 on: March 27, 2012, 07:35:07 pm »
Nice work on the wood! Looks very professional.

HLA-27b

  • Guest
Re: EMI case shielding
« Reply #8 on: March 27, 2012, 09:29:08 pm »
I'm quietly pleased with how it's shaping up considering I'm doing it the old-fashioned way (i.e. no laser cutter or CNC).  I still need to do the front panel and then paint it (I've seen some nice looking cases that emphasize the wood but that wouldn't really fit in with the rest of my stuff and you can get away with a few imperfections when it's painted  :) )


Looks gorgeous and reminds me of Lee Nielsen for some reason. No wonder robenz appreciates it, I'm sure he has many precision tools each in it's own wooden case  ;D
 


Offline robrenz

  • Super Contributor
  • ***
  • Posts: 3035
  • Country: us
  • Real Machinist, Wannabe EE
Re: EMI case shielding
« Reply #10 on: March 28, 2012, 01:31:01 pm »
Looks gorgeous and reminds me of Lee Nielsen for some reason. No wonder robenz appreciates it, I'm sure he has many precision tools each in it's own wooden case  ;D

HAL-42b,  Several trees gave thier lives to encase all the machinist/Metrology tools I have. ;D

Offline peperTopic starter

  • Contributor
  • Posts: 16
Re: EMI case shielding
« Reply #11 on: March 28, 2012, 04:32:11 pm »
Thanks for the kind words on the case!  I've spent a silly amount of time designing and building it.  I'm not a craftsman but I'm satisfied that it's as good as I am capable of making it, so it's very nice to hear appreciation :)

Regards,

Richard.
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16284
  • Country: za
Re: EMI case shielding
« Reply #12 on: March 28, 2012, 06:18:31 pm »
A simple way to keep the fan noise down is to place a spacer to the guard to move it away from the fan. I used an old dead fan, and cut out the guts leaving only the outside, and long screws. Got rid of a lot of noise for such a simple mod. The guard allows some of the turbulent flow to straighten so that the fan gets a more even flow of air, and you do not have the noise created by the blades chopping up the flow through the guard slots. As well use a wire guard, and cut out the pattern of holes in the case that is normally there - you get massively increased airflow from removing what is essentially a 75% restriction.
 

Offline CampKohler

  • Contributor
  • Posts: 17
Re: EMI case shielding
« Reply #13 on: March 30, 2012, 02:13:43 am »
Shielding a PC is not done to make it work better (unless there is a radio/TV transmitter nearby that needs to be excluded from the PC's circuits); it's to prevent the PC from interfering with other equipment. If you live in an isolated farmhouse and find it doesn't bother your use of other devices, then none is needed*. If you live near other people, the latter may come looking for you if they are on the receiving end of your interference. (A portable radio with a ferrite antenna on the AM band makes a nice direction finder.)

A slip-in peripheral may indeed create a hole in the shielding. A PC gets tested for RFI approval by running it in all of its saleable configurations (with or without a CD drive, for example), and then on top of that, has the cables moved around to find the worst possible position. Then if it passes, it passes; if not, back to the drawing board. An empty PC case, on the other hand, has nothing to test, so I don't think there is a way to "approve" them. They probably just get designed as if they were going into a system that would require approval (if you're lucky).

I see nothing wrong with aluminum foil. Oxide is of course going to coat the surface, but so what? If that is not conductive, then the RF will pass through it and "find" the aluminum below. Else all the aluminum equipment cases in the world wouldn't work either, because they would get "all oxidey" too. Oxide does present one problem, and that is preventing the intimate joining of different pieces of foil if you can't do the shielding with one large piece. Poor connections is the reason we gave up on aluminum house wiring. You could fold the edges over several times to increase thickness and then fasten those together with small machine screws and nuts or with backing washers, Pop rivet them together. Or even inexpensive small eylets could be used as rivets. Copper or brass would be better on this aspect because it could be easily soldered at the seams. Copper shielding tape (conductive adhesive?) can also be used for joining or as shielding by itself for difficult areas (like a the inside of a CD-player front plate).

The increased conductivity of copper theoretically makes it a more effective shielding material, but I don't know if you could even detect the difference, either by measuring or testing with target devices. You might need to be shielding a high-power transmitter to tell the difference. Everyone has seen zillions of those little tinned-steel shielding boxes in TV sets, game consoles, etc. and we all know how crappy steel is vs. copper, yet they are designed in. I think it just makes us feel good to know that there is copper in there when, practically speaking, an old tomato soup can might be close enough to perfection.

That CRC shielding spray Baliszoft suggested looks a lot easier to use on a wooden case instead of lining it with metal. You would need to figure out how to connect to it to bond all the shielded parts together. And I'll bet it is expensive, just because they can. (No pun intended, but enjoy it anyway.)

For holes nullifying the shielding, it depends upon the frequencies you are trying to block. If you want to really get serious use inch-thick copper honeycomb, or the equivalent made by joining lots of tubes together to simulate honeycomb. Nothing says "cool" like a shiny big exhaust stack sticking out, especially if you do a nice neat soldering job.

You can buy ferrite cores that snap shut over external cables (like handcuffs) to block RF from using them as antennas. (I always save them if I am junking something that has them.) I think RS sells them. If you dismiss them, then go to bed thinking about the fact that your power cable is UNSHIELDED! And that it plugs into a receptacle fed by UNSHIELDED house wiring. That's it, your whole house has become the enemy. Sleep tight.

Bond all shielding together in as many places as possible.

Whatever you do, get that battery-powered transistor radio out, tune to a quiet spot, run some software and test before and after your efforts, keeping all other conditions the same. If you didn't do any good, tell us and we will all write you notes of commiseration and say, "There, there, now."

--------
*Exception: Keeping radiation in is important if you don't want spies reading your data, say if you're moonlighting for the CIA or something like that.
« Last Edit: March 30, 2012, 02:29:36 am by CampKohler »
 

Offline SeanB

  • Super Contributor
  • ***
  • Posts: 16284
  • Country: za
Re: EMI case shielding
« Reply #14 on: March 31, 2012, 10:31:00 am »
I used a conductive spray years ago to use a radio near an Apple II, spraying the inside of the case with a graphite loaded spray. Grounding was by gluing a few copper strips into the case before spraying, and soldering wire onto them after cure, but before spraying. Did reduce the RFI quite a lot, but the spray was expensive ( but the solvent smelt like banana so I loved it) and really only use once, as the seals would dissolve in the can after a few months. Wooden cases are easy to line with foil, and you can solder it with silver loaded solder if you cut a tab and solder onto that, after coating liberally with flux. Stick down with spray glue, or you can spray the wood with ordinary paint and place the foil onto the wet paint, which will stick it down well.  Matt side to the paint/glue to provide better adhesion.
 

HLA-27b

  • Guest
Re: EMI case shielding
« Reply #15 on: March 31, 2012, 03:57:16 pm »
Ok here is an idea,

- take clear varnish from hardware store
- take powdered graphite from the same store
- mix the two together until the paint is saturated with graphite
- paint the inside of the case and let dry
- repeat the above step until you run out of paint or patience

Would this be an effective RF shield?
What if this is how they make the stealth bomber paint and keep it secret just because it is so embarrassingly simple?  ;D
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf