cabinet shielding question

[Randall]

Does the thin sheet metal or screening that we find under the top cabinet boards in many amps have to be electrically connected to the chassis to be effective? Or in other words, if the chassis slides in and is not actually touching the shielding, is it still effective? The reason I ask is I plan to install some thin aluminum flashing in a cab with a painted chassis. There are a few bare spots from in/out over the build, I am wondering if I should touch up these spots or not, or if the sheilding is even going to be effective.

[g1]

It needs to be connected to ground somewhere. Try something like those Hotrod/blues series of fender where they have a ground wire connecting the foil of the rear panel. I believe I've even seen coil springs used to make the connection on rare occasion.

[SoulFetish]

Does the thin sheet metal or screening that we find under the top cabinet boards in many amps have to be electrically connected to the chassis to be effective? Or in other words, if the chassis slides in and is not actually touching the shielding, is it still effective? The reason I ask is I plan to install some thin aluminum flashing in a cab with a painted chassis. There are a few bare spots from in/out over the build, I am wondering if I should touch up these spots or not, or if the sheilding is even going to be effective.

Like g1 says, it has to be grounded to act as a shield. But if it is not grounded, then it basically acts as a large capacitor for stray noise interference.

[Chuck H]

I can't remember where I saw this, but I remember a painted chassis that had a self tap screw driven into the top edge roll so that it would make electrical contact with the cabinet shield. If you're using foil I suppose this might wear through pretty quickly, but you might also apply a small plate where the screw contacts to prevent this. You might also do the same thing in reverse and put the screw into the cabinet foil shield and then sand the paint off the chassis at the contact point. These methods assume you'll take measures to safety ground the chassis through the finish anyway.

[nevetslab]

The chassis I saw in your cabinet has sufficient front/rear/side edges to engage with aluminum roof flashing. With that cut to size and stapled into place, you'll be making adequate contact with the roof shield when the chassis mtg hardware has drawn the chassis into full mounting position.

[Randall]

Yes, but nevetslab my chassis is painted. I am questioning whether to sand off the upper lip edge.

[The Dude]

It has to form a solid connection in some fashion, so yes maybe sand off part of it. I've also seen some painted chassis where they simply drill a hole and install a screw into the chassis. The screw head or heads make the ground connection to the shield.

[nevetslab]

It has to form a solid connection in some fashion, so yes maybe sand off part of it. I've also seen some painted chassis where they simply drill a hole and install a screw into the chassis. The screw head or heads make the ground connection to the shield.

Mesa uses that technique on the four corners where their chassis screws pass thru (Lone Star Chassis mounted to the roof of the combo cabinet).

[Chuck H]

I've also seen some painted chassis where they simply drill a hole and install a screw into the chassis. The screw head or heads make the ground connection to the shield.

Mesa uses that technique on the four corners where their chassis screws pass thru (Lone Star Chassis mounted to the roof of the combo cabinet).

Really glad you guys are discussing this. I was afraid no one would mention this technique

[The Dude]

If there's any confusion, here's a pic.

[Randall]

Huh. I always wondered what was going on with those screws. I just figured it was another oddball Mesa thing. (Just looking at a Mesa gut shot makes me anxious). Anyway, my build doesn't have the tolerances for those screws.

[Chuck H]

Use smaller screws.

Or... Tap a countersink screw into an edge, run a strip of adhesive copper foil along some of that edge and then drive the screw down onto it.
Or... Screw a quick connect to the shield that can be connected to another in the chassis.
Or... Do it the normal way and just grind the paint off the chassis edge roll. Nobody is ever going to see it unless the chassis is pulled so why should it matter?

It's just not that perplexing a problem.

[R.G.]

Some theoretical, as opposed to practical, considerations follows.

Shielding an enclosure is an attempt to form a Faraday cage. A Faraday cage is a conductive enclosure that surrounds the circuits, and it works because impinging electromagnetic fields cause current flow that cancels the impinging field. How perfect the cancellation is is a variable, depending on the conductivity of the metal enclosure, the frequency of the external field, and the strength of the external field. A perfect conductor forming a complete shield makes the internal field be zero. Real conductors that have openings are less than perfect.

In practice, openings in Faraday cages allow leaks, as you'd expect. However, the openings act like antennas, receiving and re-radiating into the box frequencies above the frequency that has a quarter wavelength equal to the dimensions of the opening... oops, some of you just went to sleep. That's OK - the idea is that more complete coverage is better, and smaller holes and more conduction/conductive is better.

An incomplete cage is what you get when there are holes, slots, and gaps in the cage, and the smaller the holes and gaps, the higher the frequency where it starts leaking. Audio frequency EM waves are LLLLLLLOOOOOOONNNNNNNGGGG, so for something the size of a guitar amp case, you're good with long slots. Mostly. The exception is bursts of high frequency RF at power line frequencies, like fluorescents and neon bulbs, for instance. The RF bursts can get in through a slot, then get detected to power line frequency audio in the circuits. Smaller openings are always better, so the moral of this paragraph is do your best, but don't go crazy making a complete cage, then test what you have and see if you need to work harder. An entire open side may work OK. Or not, if you happen to be in a hard-case venue.

So far, so good. Now how about grounding the cage? For really low frequency radiated interference, like 50/60 Hz, tying the case to an earth ground drains away the charge/voltage caused by the wires leading to the chassis picking up and raising/lowering the whole case as a unit. No incoming wires, no induced interference, but then no signal in or out either. The case can have no local ground, but be grounded by the audio shield ground carried by the signal ground wires to someplace on the signal chain that does have an earth ground. Two-wire amps with simple guitar setups that have no safety/earth ground generally have some level of hum from the radiated powerline frequencies jerking the whole setup around, which is one reason that adding a third wire safety ground usually helps quiet two-wire amps.

If you locally ground the case/shield, it helps a lot in getting rid of radiated low frequency hum, so the more grounded the better. If sections of the outer shield are actually isolated from one another electrically, it still provides some shielding at radio frequencies but not above the major dimension of the isolating gaps.

The resistance of free space to emitting radio waves changes with frequency. That's why radio frequencies are easier to broadcast. There is an additional effect that the length of the antenna factors into the difficulty to broadcast, so high frequency (i.e. the multiple GHz on your cell phone) can have tiny antennas, but AM radio antennas have to be very long. For downright low frequencies like powerline, the antennas have to be quite long to get significant power. There are stories of farmers getting usable power by connecting to fence wires (i.e. long).

Screwing a cover on the bottom of chassis to seal the connecting wires inside is a good choice, if you have a chassis like a box with no bottom. In open U shaped chassis, yeah, putting aluminum on the inside of the case can help.

[nevetslab]

Some theoretical, as opposed to practical, considerations follows.

Shielding an enclosure is an attempt to form a Faraday cage. A Faraday cage is a conductive enclosure that surrounds the circuits, and it works because impinging electromagnetic fields cause current flow that cancels the impinging field. How perfect the cancellation is is a variable, depending on the conductivity of the metal enclosure, the frequency of the external field, and the strength of the external field. A perfect conductor forming a complete shield makes the internal field be zero. Real conductors that have openings are less than perfect.

In practice, openings in Faraday cages allow leaks, as you'd expect. However, the openings act like antennas, receiving and re-radiating into the box frequencies above the frequency that has a quarter wavelength equal to the dimensions of the opening... oops, some of you just went to sleep. That's OK - the idea is that more complete coverage is better, and smaller holes and more conduction/conductive is better.

An incomplete cage is what you get when there are holes, slots, and gaps in the cage, and the smaller the holes and gaps, the higher the frequency where it starts leaking. Audio frequency EM waves are LLLLLLLOOOOOOONNNNNNNGGGG, so for something the size of a guitar amp case, you're good with long slots. Mostly. The exception is bursts of high frequency RF at power line frequencies, like fluorescents and neon bulbs, for instance. The RF bursts can get in through a slot, then get detected to power line frequency audio in the circuits. Smaller openings are always better, so the moral of this paragraph is do your best, but don't go crazy making a complete cage, then test what you have and see if you need to work harder. An entire open side may work OK. Or not, if you happen to be in a hard-case venue.

So far, so good. Now how about grounding the cage? For really low frequency radiated interference, like 50/60 Hz, tying the case to an earth ground drains away the charge/voltage caused by the wires leading to the chassis picking up and raising/lowering the whole case as a unit. No incoming wires, no induced interference, but then no signal in or out either. The case can have no local ground, but be grounded by the audio shield ground carried by the signal ground wires to someplace on the signal chain that does have an earth ground. Two-wire amps with simple guitar setups that have no safety/earth ground generally have some level of hum from the radiated powerline frequencies jerking the whole setup around, which is one reason that adding a third wire safety ground usually helps quiet two-wire amps.

If you locally ground the case/shield, it helps a lot in getting rid of radiated low frequency hum, so the more grounded the better. If sections of the outer shield are actually isolated from one another electrically, it still provides some shielding at radio frequencies but not above the major dimension of the isolating gaps.

The resistance of free space to emitting radio waves changes with frequency. That's why radio frequencies are easier to broadcast. There is an additional effect that the length of the antenna factors into the difficulty to broadcast, so high frequency (i.e. the multiple GHz on your cell phone) can have tiny antennas, but AM radio antennas have to be very long. For downright low frequencies like powerline, the antennas have to be quite long to get significant power. There are stories of farmers getting usable power by connecting to fence wires (i.e. long).

Screwing a cover on the bottom of chassis to seal the connecting wires inside is a good choice, if you have a chassis like a box with no bottom. In open U shaped chassis, yeah, putting aluminum on the inside of the case can help.

Applying this to Randall's question......best results should be sanding off the painted flanges so there's as much flush contact surface as possible with the aluminum roof shield (roof flashing). Mesa's approach, while it does ground the roof shield, leaves the gap created by the head thickness of the four screws. Fender's approach, copied by many, closes up the chassis with the foil shield. It's major downfall is the adhesive fails, and all too often, the first time you attempt to remove their chassis (from the sloped-front chassis amps of combo amps and head amps), that front edge pulls off the poorly adhered front edge of the roof foil that's glued to the Tolex covering flap, and rolls it up until it stops the chassis from coming out all together! Recovery is a PITA!

[Gingertube]

Slightly Off topic or an extension of it:

The GROUNDED Faraday Cage thing is why when using isolated input jacks with the cold (cable shield) side tied to 0V at the input tube it is a good idea to also add a 10nF Ceramic cap from the input jack cold side (shield) to chassis ground right near the input jack.


This adds an RF Earth connection and by doing so extends the Faraday Cage (at least at RF frequencies) out along the length of your guitar cable.


Cheers,
Ian


P.S. Just remembered from the EMI/EMC Course I got sent on for the day job. The max hole size recommended was 1/32 of the wavelength of the highest frequency you want to block. (this to provide a decent amount of attenuation at that highest frequency).

In the day job (Airborne Survey) I needed to know this to keep the Air Traffic Control VHF Radio (124-125 MHz) out of our laser receiver and to keep the Satellite Phone (2.4 GHz??) out of the GPS Receivers.
AND
To keep any RF "crap" we generated in our gear IN, so as to not upset aircraft circuits. When we designed and built the latest system, there was a 4 hour ground test and a 4 hr flight test purely devoted to EMI/EMC, we had to go through to get Airworthiness Approval sign off, before we could start airborne survey work.