Strategies for internal shielding of low-level passive signals

In building a few Pultec MEQ-5 inspired EQs for our studio, I was poring over internal photos of the originals.

I noticed that there's a metal shielding plate near the power supply (see below).

However, I also noticed that all of the wires running from the inductors and capacitors in the filter enclosure on the rear of the chassis are unshielded.

Does anyone have any insight into why this may have been the chosen strategy, as opposed to shielding the wires coming from the filter enclosure?

More to the point--would there be any disadvantage to my shielding groups of wires with braided metal sleeve, grounded on one end, until they terminate at the front panel controls--in lieu of the metal fabrication that would be needed to replicate the pictured shield?

Could it be that the pictured shield is a high permeability alloy designed to guard against low frequency EMI from the power transformer?

Thanks for any input!

It is more likely to be an electrostatic filed to prevent the high E-field around the power supply from coupling to the filter circuits.
It may be overkill because this EQ is a low impedance device so it is much less susceptible to this type of interference than, say, the gain make up input tube. As built there is a small amount of stray capacitance between the wires in the EQ. If you screened them individually this capacitance would disappear and be replaced by a larger capacitance to ground due to the cable shield. When the top plate is fitted the whole EQ is fully shielded so shielding inside is probably not necessary.

Cheers

Ian

That sort of noise can be elusive to characterize. For one it probably depends greatly on layout. I would think the power section would emit more electromagnetic noise than electrostatic. Whatever the case, even a small increase in physical distance between the PS and the sensitive bits will reduce noise. The worst case would be unshielded inductors next to the PS. A better case is shielded inductors (like the external case in the orignal) with a PS a few inches away. So if you have control over such things, that is where I think you will get a performance improvement.

But the first thing I would do is devise a nice rig for viewing the spectrum. Its difficult to fix something if you can't really see it in the first place.

ruffrecords said:

It is more likely to be an electrostatic filed to prevent the high E-field around the power supply from coupling to the filter circuits.

Click to expand...

I see, interesting!

It may be overkill because this EQ is a low impedance device so it is much less susceptible to this type of interference than, say, the gain make up input tube.

Click to expand...

That seems logical. I've asked others who have built this circuit using this layout, and nearly all omitted that shield and said they didn't notice any issues with noise.

The only reason I hesitated to just take their word for it is that my monitoring environment is extremely revealing (a Northward Acoustics FTB room), and sometimes things that aren't objectionable in typical monitoring environments are readily noticeable in that space.

As built there is a small amount of stray capacitance between the wires in the EQ. If you screened them individually this capacitance would disappear and be replaced by a larger capacitance to ground due to the cable shield. When the top plate is fitted the whole EQ is fully shielded so shielding inside is probably not necessary.

Click to expand...

I see. Since most of these leads connect to capacitors of rather larger values (up to 1.5µF, though the smallest is 1nF), I suppose they may have decided the significance of capacitance when looming the wires as they did was insignificant?

My remaining questions are:

[list type=decimal]
[*]Is shielding from electrostatic noise unlikely enough to make a difference, even in a very revealing monitoring environment, that it's not worth bothering? (may be unanswerable without experimentation, but I welcome informed conjecture)
[*]Would screening that entire bundle with a large braided metal sleeve be likely to serve as an effective substitute for the shield panel near the power supply of the original
[*]Would the capacitance to ground of screening the entire bundle as one be likely to create other issues/unintended consequences?
[/list]

Just doing some back-of-the-envelope math, even assuming a rather large capacitance to ground like 200pF it seems most likely that any stray low-pass effects would likely be well above the audio band.

Of course if the best course is truly to just omit the screen and do nothing, I'll absolutely listen to reason. Really that would be easiest of all for me (I'd just rather not be faced with extensive modifications after the build!)