Is it really that bad to use a plastic enclosure?

[mr.adambeck]

I built a buffer (based on the beavisaudio.com tutorial) into a small RadioShack plastic enclosure.  It looks really nice (painted it and all), and I'd hate to have to throw the enclosure out.  Since making it I've heard it's bad to put circuits into plastic boxes because it doesn't ground correctly or shield or something that's beyond my noobish comprehension.  Is it really a bad idea to leave it in it's case?  Thanks

[R.G.]

You leave yourself open to hum pickup and radio station pickup. This may not be a problem or may be intolerably bad, depending on where you are playing the pedal. Well shielded pedals are more immune to this.

[Mark Hammer]

I build LOTS of pedals into plastic boxes.  Shielding IS important, but:

1) You can use shielded wire, the shielding doesn't have to be provided only by the box.
2) You can use shielding in conjunction with a plastic box.  I cut out a piece of copper shim that fits the underside of the top surface of the box, and hold it in place with the various panel-mount components.  I can solder ground wires to it.  Is it "complete" shielding?  No, but it helps a lot.
3) Degree of shielding needed will partly depend on the amount of gain the circuit has.  If a circuit can pick up hum, you don't want to amplify that by x300. So, if you're going to amplify by x300, make sure there is no hum.
4) Metal boxes provide no assurances of NOT picking up noise inside the box, as many builders of hi-gain or high-frequency-clock-based circuits find.
5) The box always has to be able to withstand the weight and kinetic energy of the user.  Sometimes plstic isn't so good for this, though sometimes it is surprisingly good.

[frank_p]

Anybody know if aluminium mosquito screen would work ?  Or the meshings are too large for protecting against electromanetic fields ?

[R.G.]

A metal shield excludes all electromagnetic frequencies for which the largest dimension of the opening is 1/4 of a wavelength of that frequency.

The wavelength of a EM wave is L (usually the physics and EE boys use "lambda") = v/f, where v is the speed of the wave and f is the frequency. V for radiated waves is c, the speed of light, 3x10^8 meters per second. So the frequencies excluded are those up to where four times the biggest opening length is F = c/(4*opening).

For grins, an opening 250mm long blocks everything up to f = (3*10^8)/1M = 300MHz.

This is heavily simplified, of course. The blockage isn't perfect, and it tends to be the joints where things leak in that are the longest dimension.

But yes, screen wire, done well, keeps out everything up to high microwaves, terahertz waves, visible light, etc. The wires must be metallic and must make good contact with each other.

[sean k]

I built this thing which is a sheet of steel with three strings stretched accross it and its amped with piezos. To make use of the effect we don't want I mounted the circuit board in a hole in the steel sheet and, while I was populating the PCB, I discovered I'd make a mistake and had to cut a trace and mount a 100k resistor on the copper side. Just so happens I can pick up a particular radio station using the capacitance of my body and can almost get a tremelo going by moving my finger up and down.

So I kinda tried hard to get radio interference and only got it when my finger was within a centimetre and a half of the offending point so I did acheive it but it was very controllable... I wonder how I might be able to change the station to student radio   

I tried photographing it, the board, to show you'se but my batteries wre down and are right now charging so maybe I'll do it latter. The only caveat is that I was in the basement of a huge concrete building in the middle of a city at the time I discovered this so things may change when I fire it up latter to trouble shoot one or two other things that became apparent.

[frank_p]

I had a vague sentence floating in my head: my physic teacher was saying: "The oscillating electric field is positioned at 90 degrees from the magnetic field in the oscillatory model of light...".

The problem is of course that it's a representation: I don't have a representation of the magnetic field variating that is not in "Cartesian textbook space" (I don't know how to type it... because the image is not clear in my head).
I am just in a, "how that wave can swirl in space" - moment.  Is its orientation is really in the direction of it's representation (on paper).  And that would be why it can't pass in a channel... I must study more.

But I am hacking a topic here...  and I should pull a book on that "open path".  Thanks R.G. to bring my foots back on earth: I was in a philosophical argue with my father about continuous discourse vs fragmented one and it was getting nowhere        .  And when words seems "without sense" we still have that good old science   ... Well I hope (nature of light... )  ...

OK back to boxes:
It's because the plastic box topic brought me (again) the idea of putting the led(s) inside a polycarbonate enclosure and shielding it with a screen so we could see the light when engaged.  But I don't know if electronic stores carry polycarbonate ones.  Forget about polystyrene or acrylic, those will possibly fail the  test.

In return polycarbonate could perhaps be a good plastic choice.  I think the current Hammond ones are ABS, but ABS can't be translucent. 

(Note): Only amorphous plastics can be translucent (ex: polystyrene). In crystalline ones, macromolecules get organised and form patches of crystals that will be oriented in different ways preventing the light to pass in it (ex: polyethylene).  The slower the crystalline plastic get solid in the mold, the more there will be patches of crystals (and will be more opaque).  A crystalline plastic that freeze faster in time becomes more like an amorphous one and will tend to let more light pass trough (milky).

In boxes that are made of the technical plastic ABS:  Acrylic and styrene provide rigidity and polybutadiene act as a "dampner" preventing the propagation of cracks in the material.  Just try to break a sewer pipe on a concrete floor.  The toughness of the ABS will depend on the actual copolymer properties and the speed at witch it is molded.  When  $$$s are in the equation, material choice and production speed are important factors.  Generally the faster you produce, the more your ABS objects will be fragile.  First, the speed of the screw when loading in the injection press must turn fast and that can put a lot of sheer stress on the material. Then, when injecting, stress again.  Too much stress can break macromolecures and thus reduce the mechanical properties.  A lot of the heat that is required to melt the plastic is brought by shear stress: If there is too much heat you can induce thermal degradation too (dependant on resident time in the barrel and heat).  Then you have to compress in the mold (by adding pressure with the screw) a bit of time, so the plastic freeze in the nozzle of the injection press.  This is for forming a solid plug that will keep the holding pressure in the mold while the molded piece cool and solidifies (gate time) and at the same time the screw turns again and is pushed back to get ready for the next shot.  This lap of time (gate forming time with correct pressure) is necessary and if not taken in account can lead to defects in the piece.  Then the piece of plastic must cool down.  The more you produce fast, the more the cooling is done fast: this leaves residual stress in the material.  The box could then: be more prone to fracture, warping, twisting, etc. especially when softened by heat or fragilising by cold by the user (this could come from your circuit).  Also important is that the walls of the boxes be as thin as possible to reduce shrinking (depressions caused by the change of density while cooling).  Thin walls are more prone to fractures, and that is why ribs are most often seen in pieces (look inside an ABS Hammond box on the sides).  Thinner walls cost also less in material.

In the end you can see how a plastic box can be made more prone to failure just for the necessity of the speed of production, cost of material, and from the inherent limits of the plastic and process.  One thing that saves a lot on the durability of the piece is it's aesthetics: if everything is done too fast, all kind of defects can appear on the product.  This gives time for the plastic not to be too much stressed (and freezed that way) and vulnerable.  So the cost and solidity of a simple box is dependant on many factors from the design of the piece, to the design of the mold, the process variables selection, etc.  Making a piece of plastic is an art and decisions taken at the beginning of the product design can have impacts on the quality and price of the product (especially in the long run).

Also, in a big production context when there is buying department that decide what to buy without telling for cost reasons -ex: offgrades materials- (variation in the plastic properties = mad workers on the line), and then the quality control department reveals too much failed tests, it can really be a challenge to pin point what is going on, adjust the process variables and have consistency on the product durability.

Hope that this can put an illustration on the - Stomp the Plastic Box Test Failure- .  An ABS box with appropriate wall tickness or/and structural ribs would have no problem with the stomp force.  All that to say that my plastic Hammond boxes look like they might be some "act nice and gentle to me" types.

[ClinchFX]

As I understand it, a die cast aluminium enclosure will provide almost no shielding from the predominantly magnetic field that causes mains hum.  Yes, an AC magnetic field will create eddy currents in the walls of the enclosure, but these eddy currents create a secondary magnetic field that will be re-radiated.  Even normal sheet steel doesn't significantly attenuate magnetic fields.  Mumetal is the best way of providing a good shield.  There's some info about Mumetal at this link http://www.mumetals.com/faqs.html

Die cast enclosures can provide effective shielding from radio frequency interference, but, regardless of this, the circuitry of a pedal should be designed with careful attention to grounding, RF bypassing and power filtering - aspects that are all too often ignored.

As for the original question:
Rather than discard the plastic enclosure, if the buffer is working properly and is not picking up radio stations, there's nothing wrong with using a plastic case.  If there is a problem, you can spray the inside of the box with Electrolube NSC - nickel screening compound - or a similar product.  I've found NSC to be very effective for RF shielding.

I use die cast aluminium enclosures for robustness, not for any special shielding qualities.  I use PCBs with a ground plane, insulated jacks, and a brass bracket to mount the PCB to the enclosure so that the case is grounded to the pedal circuit at only one point.

[MikeH]

I find stomping to always be the ultimate demise of a plastic enclosure- a buffer though, doesn't require foot stomping, so in that regard you're fine.  I would however, get yourself some spray adhesive and shielding foil (you can probably just use kitchen aluminum foil) and shield the inside to lighten the possibility of interference from radio signals, and similar such noise sources.

[bassmannate]

I find stomping to always be the ultimate demise of a plastic enclosure- a buffer though, doesn't require foot stomping, so in that regard you're fine.  I would however, get yourself some spray adhesive and shielding foil (you can probably just use kitchen aluminum foil) and shield the inside to lighten the possibility of interference from radio signals, and similar such noise sources.

I was just getting ready to say that you're probably better off getting some spray adhesive and aluminum foil instead of using metal window screening! This has been done for quite a while with guitar control cavities (did it to my bass a few years ago) and it's not too hard at all.

[frank_p]

I find stomping to always be the ultimate demise of a plastic enclosure- a buffer though, doesn't require foot stomping, so in that regard you're fine.  I would however, get yourself some spray adhesive and shielding foil (you can probably just use kitchen aluminum foil) and shield the inside to lighten the possibility of interference from radio signals, and similar such noise sources.

I was just getting ready to say that you're probably better off getting some spray adhesive and aluminum foil instead of using metal window screening! This has been done for quite a while with guitar control cavities (did it to my bass a few years ago) and it's not too hard at all.

As mentioned before, the idea was to put the screen inside a transparent box and put some LEDs inside the box as to have a different looking stompbox.
Other than that, yes, metal foil is simpler and more effectve.
There is also the Krylon magnetic paint option (as mentioned on AMZ).

[KorovaMilkBar]

hmm, not sure if anyone has mentioned this, but if you use some sort of spray paint that is basically a bunch of iron particles (it might be something called Rustoleum, its a kind of primer). you spray this iron paint on the inside of the enclosure to help shield and ground your project. that might be something to look into.

[WLS]

hmm, not sure if anyone has mentioned this, but if you use some sort of spray paint that is basically a bunch of iron particles (it might be something called Rustoleum, its a kind of primer). you spray this iron paint on the inside of the enclosure to help shield and ground your project. that might be something to look into.

Yes, I've also seen such a product. Don't know exactly where since I don't use plastic, but it does exists and from what I've read the results are worth looking into. But it was not Rustoleum, it was an aluminum based flake type paint.

Bill

[doc_drop]

I have had good luck with aluminum HVAC tape. You can just apply it to the inside. To be sure about the grounding I use a small bare wire in between layers to make sure the tape conducts everywhere. I also typically clear coat it after so the joints don't pop up over time.

In general it is more work to do a plastic box like this, but I have made it work quite well. It winds up not being cheaper due to the cost of the tape, though.

[sean k]

Anyways, heres my picture of the totally open PCB that picks up a radio station.

Funnily enough the resistor you can see on the board actually isn't supposed to be there. Its three piezo preamps going to one mini mixer and while I was populating it, I couldn'tfind the schematic, and mistakenly cut the track and bridged it with a resistor as I was thinking about summing.

So the point where R5, R6, R7 and R8 meet and go to pin 2 was cut and a 100k added. So thats become my place to bring in radio and when I touch the side to pin 2 I get an oscillation which I'm thinking I could actually change with different caps or even with a low pass combo.

So I've left the PCB open to the elements and done something wrong but it still takes a concerted effort to get the radio signals going, hands on as it were.

[frank_p]

Anyways, heres my picture of the totally open PCB that picks up a radio station.

Are you sure it's not extraterrestrial living form, trying to get in touch with you Sean ?
Or a shaman lost somewhere in an other dimension, in need of help ?
I like the look: and understand now why you don't want to box it.

[sean k]

Hey frank, All of the above and then some  though it sounded suspiciously like your garden variety talkback radio station. Maybe thats the new interdimensional message system...?

[mac]

IIRC, after decades away from university, the problem of a EM wave hitting a metal surface, like a a metal enclosure, can be solve in the following way,

Jimmy C. says that for an EM wave

div.E = 0
rotxE = - dtB
rotxB = u.j + dtE/c/c
div.B = 0

and for a metal

j = S.E

dt is partail time derivative.

I hope there is no purist mathematician watching this, (they are much like tube purists!)

rotxrotxE = grad(div.E) - (grad.grad)E = - u.S.dtE - dt.dtE/c/c

since div.E = 0 then

(grad.grad)E - dt.dtE/c/c = u.S.dtE

same eqs. for B

The first term is the typical wave equation, but the second is a dissipative effect.

The solution is of the form,

A(r,t) = Ao.exp[ i.(k.r -w(k).t)]  where Ao is a complex constant vector.

I'm not going to solve it, it's quite easy to get that k has a term that depends on S,

k.k = w.w/c/c + i.w.u.S

So there is an exponential decay given by

exp(-w/c/c( -1/2 + 1/2sqr( 1 +(S/e/w)^2 )))

e = 1/u/c/c

When w.e << S then the EM wave decays quickly after a few lenghtwaves.
This is true for a lot of freqs because

S/e = 1e18 Hz.

Back to the original question, plastic and wood are ok if you put a metal cover in the inside walls.
The thicker the better.

If you like a little "mojo" use aluminum!!!

mac