ground loops

[dexxyy]

Ok, probably a dumbass question, but one that has been bugging me for ages. A ground loop is a bad thing, can create a noisy pedal yes?  So if the pedal is in a metal enclosure and said enclosure is grounded, why does the enclosure not create a ground loop???

[toddvirgil]

Hey Dexxy,

I'm no expert by any means, but wikipedia has a good article:

http://en.wikipedia.org/wiki/Ground_loop_(electricity)

It sounds like its caused by different potentials in grounds (I.E. not everything is grounded to the same "ground")

"In video, ground loops can be seen as hum bars (bands of slightly different brightness) scrolling vertically up the screen. These are frequently seen with video projectors where the display device has its case grounded via a 3-prong plug, and the other components have a floating ground connected to the CATV coax. In this situation the video cable is grounded at the projector end to the home electrical system, and at the other end to the cable TV's ground, inducing a current through the cable which distorts the picture. As with audio ground loops, this problem can be solved by placing an isolation transformer on the cable-TV coax."

So, it would seem, that since we ground everything in our box to the same point -- the DC jack ground, that we wouldn't have that problem. I'm assuming in the above example, the internal components are only connected to the coax ground and not also to the case ground (as we do with our input/output by wiring them to the DC ground).

Again, someone who actually knows this stuff can definitely chime in.

T

[induction]

There are two important issues that contribute to ground loop problems.

1. Non-idealized resistivity: Conductors like copper wire are often treated as perfectly conductive, but are actually somewhat resistive. The differentiation between conductors and resistors is a convenient but inaccurate binary approximation. In reality all materials (except superconductors) have a non-zero resistivity. In electronics, we often refer to all points connected by wire as the same point. Sometimes, we even call them 'constant potential surfaces', erroneously implying that the voltage of every connected conductor must be the same at every point. In fact, wires are just low-value resistors, and there can be a (small) potential difference between two ends of the same wire.

2. Conductive loops are antennae that can detect current in nearby wires: Faraday's law indicates that a changing magnetic field will always generate an electric field. If the magnetic field inside a closed conductive loop changes, current will flow in that loop. Ampere's law indicates that electric current always generates magnetic fields.  Combine these two statements, and you will find that your pedals (as well as all nearby electrical equipment) broadcast magnetic fields that will induce current signals in any closed loops of conductive material. These closed loops are called 'ground loops' if they are connected to ground potential. Conductive loops can pick up electronic noise even if they are not at ground potential (and 'ground' is really arbitrary anyway - only potential differences matter, not absolute values), but 'ground loops' is catchier than 'conductive loops', so that's what people talk about.

As far as pedals go, both points 1 and 2 can be cause for concern, but only point 2 really refers to ground loops. (Oddly, the Wikipedia article on ground loops only discusses the first point and not the second. I can't really complain, because I'm not willing to edit the wiki page.) Nearby electrical signals (including those from pedals, lights, computer monitors, etc.) generate time-varying magnetic fields, which can generate time-varying voltage signals on ground loops. (The effect of stray voltages is aggravated by point 1 whether the conductor in question is a loop or not.) These signals can sometimes appear as noise or hum in our signal.

Grounded, conductive enclosures are great for pedals because they form a faraday cage around the circuit that helps screen out stray electromagnetic fields that are generated outside of the enclosure. However, they will create a ground loop if they are connected to the circuit's ground in more than one place. Many of us use non-isolated input and output jacks which can create two grounded points on opposite sides of the enclosure. In this case, if the input and output jacks grounds are also connected by wire (or a series of separate wires), there will be a ground loop. If one of the jacks is grounded only by the enclosure and not by an additional wire, there is no loop.

That being said, I always use a wire to ground both jacks. This is because the enclosure and jacks are made of dissimilar metals, which means their connection point will corrode over time and may eventually lose electrical contact. Even though the enclosure and the ground wires form a loop, only signals from inside the circuit will generate appreiciable current in the loop (most external fields are screened out by the enclosure itself), and these currents will be hard to detect because they are small and are correlated to the audio signal. In practice, a ground loop formed by a multiply-connected enclosure is not a problem for pedal users.

Ground loops can be a problem in daisy chained pedals. These loops are caused by the fact that each pedal is connected by ground wires in both the signal cable and the adapter cable. Hum from these loops can sometimes be a problem because these loops are much larger than the ones from multiply grounded enclosures (a larger loop means more magnetic flux inside the loop, which means larger induced signal). They are also usually not surrounded by a faraday cage. (That being said, I daisy chain my pedals and have never had a problem with it.)

Tl;dr: Don't worry about ground loops formed by pedal enclosures. They aren't a problem.

Related issues:
- Point 1 explains why you should always attach your power filter caps as close to the adapter ground wire as possible, and take all other circuit grounds from the other side of this cap. (ie no circuit ground connections should come between the filter cap and the adapter ground).  This gives the AC ripple the closest path to ground and puts a voltage divider between the signal and the ripple.

- Point 2 explains how magnetic guitar pickups generate voltage signals with no electrical power source.

- In several places on the internet, I've read that electrical signals are shielded by electric conductors and magnetic signals are shielded by magnetic materials. This isn't quite true. Both types of signals can be shielded by electrical conductors. Magnetic materials are useful for shielding low frequency magnetic fields because skin depth decreases with frequency (in other words, lower frequencies penetrate deeper into the conductive material, so conductors of finite thickness can only shield out fields above some minimum frequency). Mains hum (60 Hz) is often low enough to penetrate shielded cable. (This isn't exactly about ground loops, but I really wanted to get it off my chest.)

- Despite it's length, this post is just barely scratching the surface of this subject. There's plenty more to learn from any good electronics or electromagnetic physics textbook, not to mention internet searches.

[toddvirgil]

Awesome.