Guitar pickup theory #5 : shielding, hum-cancelling and flux concentrators

Since I am STILL waiting for my parts orders to arrive, I took the time to run a few tests on shielding materials’ noise reduction in practical situation (i.e., sitting in front of my computer workstation).  This is just another off-the-cuff post, so it will be light on math and explanation, but hopefully it will be helpful.  The tests are also easily repeatable by anyone looking to experiment on their own with just a few alligator clips, a freebie plugin, and some metal from the hardware store. WARNING: This post will be graph-heavy, and some of them will be quite cluttered, but hopefully they illustrate the trends.

I ran the tests on a Duncan JB humbucker and a 5k DCR staggered alnico pole single coil (with an EMG81 tossed in as a reference at the end).  The materials tested were thick aluminum foil and 1/16″ mild steel.   The foil was used in 2 configurations: single sheet wrapped around the entire pickup, and 5 sheets.  The steel was used in 2 ways: a gapless enclosure around the entire pickup, and a U-channel guide around the bottom and sides of the pickup with the top open.  The pickups under test were run into my test setup… 1Meg input, unity-gain opa2134 opamp, battery-powered buffer, and then into a 96kHz/32-bit float sound card.  I used Voxengo SPAN plugin, hosted in LogicX for the FFT measurements.

BACKGROUND:  guitar pickups are typically unshielded, and have quite leaky magnetic flux patterns.  This leads to quite a bit of extraneous noise pickup from surrounding EMI sources, particularly 60HzAC, 120Hz rectified noise, EMI from electronics and all their upper harmonics.  I won’t bore anyone with the background of the development of hum bucking pickups and various noise-cancelling techniques since they are well-known, but the subject of shielding is rarely discussed in a testing format (apart from control cavities) and is even more rarely empirically demonstrated.

LET THE GRAPHS BEGIN!

For reference, here is an unshielded humbucker vs an unshielded single-coil:

SHIELDBLOGHumvsSin

As expected, lower range noise is much lower (12-18dB) on the humbucker, and upper harmonic noise is generally lower on the humbucker, although you can see the increased level at their respective resonant frequency peaks.  This is pretty typical comparison of unshielded pickups.

Here is a graph of the same pickups, but this time they are both shielded by a single layer of foil (I ran the foil both grounded and ungrounded and there was little difference, although it was slightly better when grounded):

SHIELDBLOGHumvsSinShield

There is a general trend for both pickups that the low freq noise level drops 6-12dB or so because of the shielding.  Another trend worth noting is that, as we would expect, the foil is more effective at higher frequencies (12-24dB) since the magnetic interference at lower freq passes though the low-permeability foil, and the penetration depth becomes less significant in relation to the thickness of the foil.

Moving on to steel.  Here are graphs showing the difference between an unshielded pickup, non-ferrous foil shielded, and steel shielded.  First, the single coil (white line represents the noise level of the interface with input cable tip…errr alligator clip… grounded):

SHIELDBLOGSCshielding

And now the humbucker:

SHIELDBLOGshield hum

The graphs are a bit busy, but as a general trend, you can see that the steel redirects the magnetic fields around the pickup better than the foil at low frequencies, and lowers the low-free noise about 12dB more, while the foil works better at higher frequencies.  Additionally, adding 4 more layers of foil reduces low freq noise about 1-6dB, and high freq noise an additional 12-18dB beyond the single layer.

So let’s look at our best and worst case measurements so far: unshielded single-coil vs foil+steel-enclosure shielded humbucker to illustrate the effect of the combination of common mode cancellation plus shielding:

SHIELDBLOGHumDblShVsSC

The effects are obvious.  We have more than 12-18dB reduction in the low freq noise, and 30dB reduction in the high freq noise.

 

NOW WE SWITCH GEARS:

Instead of just shielding, let’s try plugging up a few holes in the leaky flux path inherent in guitar pickups.  To this end, we’ll use a steel U-channel to guide the flux from the bottom and sides of the pickup, up to the top, leaving the face of the pickup and pole pieces exposed.  This differs from the steel box shielding test because we are not using steel to redirect external fields from the pickup, but rather placing the steel IN the pickup’s flux path to guide it to a more useful place, and prevent the flux lines from radiating outward where it can pick up all kinds of nonsense.  Typically this is called a ‘flux guide’ or ‘flux concentrator’ in the world of magnetics:

SHIELDBLOGHumDblShVsSCVsFluxGuide

Here we can see an additional 12dB or so improvement in low-frequency noise even over the gapless steel enclosure (which is convenient since a pickup fully enclosed in steel would not exactly function).

So as a final graph, here is a graph of our best measurements (flux concentrator/guide + foil) , as well as our baseline interface noise, and the low frequency noise from an EMG81 active pickup for reference:

SHIELDBLOGthickshieldVsFluxguide1

So we can see that there is a significant noise reduction across all relevant bands using just a few dollars worth of metal.  We have even managed to improve the noise performance of a standard passive pickup beyond the noise performance of an active pickup.  Of course, merely adding a bunch of metal will definitely change the sound and output level of the pickup (as well as making it too large to fit in a pickup cavity… derp), but new designs can quite easily be configured for much better noise performance.

On a related note, here is the rig I put together to shield pickups for sweep measurements, etc.  It is made from two aluminum tubes formed from wine rack parts (~1/8″ avg thick interlocking plates), with a cardboard liner, and a foam bed to isolate it from vibrations.  The two are isolated from each other with foam, and the entire rig is sitting on wood to isolate it from my workbench.  There is a steel plate underneath to provide some magnetic shielding from the mess of wires underneath my workbench. Here is the difference between the stock Duncan humbucker outside the tube shield vs inside:

shriedtube

 

Hopefully my parts will come in soon and I can put up a more technical post.

Advertisements

About alexkenis

Guitarist, philosopher, tinkerer

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: