Pedalboard PS Enclosure Material?

[RDV]

Should a PS for Pedalboards be in a aluminum box or can non-metallic materials be substituted? I'm making one from scratch with a 300ma tranny.

RDV

[cd]

A metal, so the transformer doesn't radiate hum into your pedals, especially a wah.

[R.G.]

Steel. Well, actually, soft transformer iron, but you won't find any of that.

Shunts any stray field. Also put a copper "belly band" strap around the transformer.

[RDV]

Steel. Well, actually, soft transformer iron, but you won't find any of that.
Shunts any stray field. Also put a copper "belly band" strap around the transformer.

Now I know that you're a rancher down there in Austin, but I'm from AR and I know that a "belly band" goes on a horse(or dog) to hold back the old penis. Are saying a grounded(to the DC ground) copper sheild around the transformer will block magnetic fields that will mess with nearby effects? I wouldn't mind using a steel enclosure, but they're so hard to drill without a press(especially the 1/2"ers for the DC jacks. I don't want to half-ass it, but will a 1590S Hammond not work if I still copper-shield the Tranny?

Thanks

RDV

[RDV]

Also, is 1000uF(low-leakage Xicon) enough initial filtering(not counting ripple filters at the regulator) for a 300uA tranny?

RDV

[Michael Allen]

I used 1000uF's on my supply. It was 12v tranny at 1.5A and It works well for me. It should be enough for 300mA.

I made good use of a computer power supply case. Good metal and good size as well....

[RDV]

O.K. R.G., I've found scant info on the internet about "belly shields" as they apply to transformers(and a lot about dogs & horses). I'm assuming that it would be installed hemispherical, somewhat like the rings of Saturn?

RDV

[R.G.]

A belly band in transformer jargon is a thickish copper strip wound around the core and windings on the outside of the whole mess and soldered into a shorted ring. The heavy, low resistance copper shorted turn represents a very large load to any magnetic flux that cuts the ring. Therefore, the flux is semi-confined inside the ring.

[RDV]

This is the transformer I'm using: http://www.mouser.com/catalog/specsheets/114137.pdf Since I'm going to stuff the whole mess into a 1590S Hammond, would I be well served just copper-taping the entire interior? Would I then ground the copper to the DC ground?

RDV

[petemoore]

I don't understand...
 I would think enclosing the tranny in a Hammond Box [aluminum] that's connected to circuit ground would contain the electrons the tranny shoots out...or...shield all RF coming from the tranny...

[RDV]

http://www.lonestaramps.com/users/RDV/gitrdone.wav

RDV

[petemoore]

:twisted:  :lol:    :shock:  :?

[R.G.]

I would think enclosing the tranny in a Hammond Box [aluminum] that's connected to circuit ground would contain the electrons the tranny shoots out...or...shield all RF coming from the tranny...

This is where understanding magnetic shielding gets messy, and your statement is a good example of the most common reasoning.

We're used to dealing with electric fields and their appurtenances, charge carriers, conductors and insulators. The charge carriers of e-fields are electrons and/or positively charged ions. There is a whole panoply of conductors from quite high conduction (silver and copper) to almost no conduction ( teflon, mineral oil, and glass) which we call insulators. The range of conductances is huge, many orders of magnitude.

Magnetic fields don't have the same simplicity. There are no "charge carriers" of magnetic energy analogous to electrons; there are no magnetic monopoles known yet. The best m-field "conductors" are soft iron and special stuff like mu-metal. These "conductors" are at best tens of thousands of times (i.e. 10E4) times better than free space. Contrast that to the easily 10E12 range of conductances and you see that the range of stuff that can "short out" m-fields is limited.

Ordinary transformer iron has a relative permeability (magnetic conductivity) of maybe 15 thousand. That means it conducts m-field 15 thousand times better than free space. So in a transformer, where you're trying to keep the m-field inside the iron, you have the equivalent of free space "shorting" the magnetic flux around the outside of the core by paralleling infinitely many loops of free space outside the core. Each is 15K times worse at conducting m-field, but there are infinitely many of them, so the result is significant leakage of the m-field from the core.

There is no m-field insulator better than free space, so there is no magic cardboard shield we can wrap around a transformer to keep the flux in.

What we can do is to make it hard for the flux to get out. If you have m-field cutting across a loop of wire, an electric current is induced into the loop. The laws of physics (...er, I mean Mother Nature) say that the induced current causes a resultant m-field that opposes the m-field that caused the current flow. Don't ask me why. Mother Nature just said so.

The degree to which the conductor makes a repellent m-field that keeps an external m-field out depends on the amount of current, and our friend Georg Ohm tells us what the current flow is - I = V/R where V is the induced voltage caused by the impinging flux density and R is the resistance of the current loop. If you have a superconductor (that is, R=0) then you will induce an infinite current by causing a magnetic field to cut a conductor loop. Experiments show this to be true: you can suspend a magnet forever over a superconducting disk. This has heretofore required a disk of metal dunked in liquid helium, or equivalent, but it does work in the real world. Shorted turns of conductor force magnetic fields to be entirely inside or outside them.

Back at our normal transformer, the hum inducing leakage is m-field leakage. It's not electrons shooting out, and it's not electromagnetic radiation (RF). It's magnetic field flux lines. M-field comes out where the iron "conducting" the m-field is discontinuous, at the joints of the E-I stack. This is one reason toroids are so low-hum: they don't have corners or gaps. By putting a low resistance band around the transformer, the induced current in the shorted turn forces flux back into the transformer rather than letting it fly off into the rest of the universe. The lower the resistance of the shorting ring, the more flux forced back into the transformer. That is why copper foil doesn't help all that much - you want *LOW* resistance.  And it *MUST* form a shorted turn outside the transformer.

But copper's not perfect as an electrical conductor. There is some residual flux that gets past the copper belly band, a consequence of copper's non-zero resistance. You could try another belly band, but most people then use iron to soak up some more flux. This is kind of an iron case around the belly banded transformer. End bells are a clumsy way to do some of this, but good isolated transformers use entire cases. In fact, good audio input transformers that want to keep external m-field radiation out will use multiple layers of copper sheet and iron to get high magnetic isolation. See the Jenson web site.

A Hammond box helps some, but only if you can seal the joint around the edges *CONDUCTIVELY* as well as the box conducts. The box oxide hinders the conduction, too.

If you're getting the idea that preventing magnetic field radiation from transformers is hard, you're right.

Could be worse. There is *NO* known way to short out or shield out a "DC" magnetic field short of a superconductor. Conductive shielding works by the AC nature of the leaked magnetic field.

[RDV]

Wow! Git-er-done indeed. That's yet another incredible post.

Thanks, R.G.

RDV

[Phorhas]

R.G, you prove to me again and again thatall that I know is tha I know nothing


So much to learn....

[Hal]

wow i knew all of that from physics, but didn't put it together myself...

could some sort of active magnetic field destroyer be made....something with a small b-field sensor, which would actively create an out of phase current, almost like noise cancelling  headphones do?  This way, resistance of the copper wouldn't be a problem...