Ground Control...

[mydementia]

Hi guys.
I've been building for a while now and keep reading different things on grounding technique.  Most recently I read that you should run battery neg to the board ground, then run board ground points to the input jack ground, pot grounds, and the output jack (this is how the GGG projects are set up) - but the guy also mentioned isolating the output jack from the metal enclosure...seemed strange to me. 

Another technique was to solder all the pot grounds to the pot body counting on the enclosure to provide ground...

I'm looking for recommendations/opinions on this...is any ol ground good, or should we be smarter about it in these low power stomp circuits?

Thanks in advance.
Mike

[R.G.]

You read different things about grounding because grounding is one of those things that seems like it ought to be simple but isn't.

"Ground" as used by us USA'ns and "earth" as used by others, is taken to mean a reference voltage, usually a place of zero volts. Everything else is measure with reference to this agreed-upon reference point. This made sense when electrical equipment was big and ugly enough that it could always be connected to a metal rod driven into the ... um, ground. The entire planet is a big and conductive enough object that it's quite hard to have voltage differences across it.

When we moved electrical equipment up onto wooden tabletops, "ground" got kind of fuzzy. We could decide to connect any point in the circuit to a ground rod, and the equipment still worked the same. So we changed the meaning of "ground" to mean "a reference potential of zero volts".

The fly in this ointment is that our wire conductors are not perfect - they are actually resistors.

If I have a sensitive input stage which uses power from a battery, and a tough, manly, high current output stage that carries an unrelated signal and power it from that same battery, what happens?

It depends on how you hook up the ground wires, and where current flows. If both the input stage and the output stage have a separate wire to the battery "ground", then the currents flow through isolated resistances, and there is no cross-contamination of the signals. If the both flow through the same wire, the high currents from the output stage cause a voltage to appear across the non-perfect copper wire, and the input stage may amplify this as though it were a signal. Worse, if our "ground" wire forms a loop, the constant sleet of 60 cycle radiation, radio wave hum, and magnetic fields passing around us can induce a signal into that "ground" wire, and the sensitive input stage will happily amplify that junk, not being able to tell it from the real signal coming in from its input.

It's hard for us to find these things out because what matters is where the currents flow more so than where the voltages are. The induced voltages are tiny, and hard to find. And we're trained to look at voltages.

The theoretically perfect method for grounding for audio is star grounding. In this method, you designate one point as "ground" and every other place where a signal must be referenced to ground or current is drawn back to ground has its own wire. So no currents from different stages flow in the same wires, and there can be no cross-contamination of ground currents. We want these wires to not form pickup loops for radiated junk to eliminate that possibility too. Every electrical device in the world would be wired this way except for the fact that it's impractical, difficult, or impossible for any complex circuits because of the huge number of ground wires.

But we do as best we can. The story in grounding is to group together low current grounds or sensitive stages, high current stages or stages with cancelling ground currents, and shields so that they don't cross-contaminate. For effects, the number of parts to be grounded is small enought that we can usually get away with simplified all-in-one ground busses that don't show up many bad effects. But there are some rules of thumb.

I read that you should run battery neg to the board ground, then run board ground points to the input jack ground, pot grounds, and the output jack

This is good practice. It makes the PCB be a biggish, blurred star ground point and gives you most of the advantages.

but the guy also mentioned isolating the output jack from the metal enclosure...seemed strange to me.

Strange, but also good practice. The metallic box acts as a shield. You want this shield grounded in exactly one place to your circuit. If there are more grounding points to the external shield, all that radiated junk can be induced in it. The smaller the box and circuit the less this matters, and sometimes we can get away with grounding murder, but it's not good practice. The external box is attached to ground by the input jack unless that is isolated, and you only want one, so isolating the output jack is a good idea.

Another technique was to solder all the pot grounds to the pot body counting on the enclosure to provide ground...

Modestly good idea because the enclosure is a big chunk of metal and therefore a lowish resistance path. But a quite bad idea otherwise because the enclosure is the RF shield, and should not be used to carry signal currents. Soldering the pot grounds to the pot bodies and hoping for ground through the body makes the signal currents flow in the shield in the best case, and if the pot attachment nut gets loose gives you a dirty, intermittent ground over time. Overall, bad idea.

Any ole ground is NOT a good idea. We should be smarter. And are.

[Peter Snowberg]

Aluminum boxes are a very poor thing to count on for connections because aluminum tends to have a thin layer of oxide on the surface which is a great insulator.

[mac dillard]

Well, I'm new and dumb to pedal building but have built a few tube type amps..I always try for a form of Star Grounding. It is hard to beat , in my opinion..Read some of Randall Akins's articles on grounding. Pretty good stuff.

[idlechatterbox]

Interesting thoughts on grounding. I am a tinkerer, at most, with the stompboxes, and have had only a bit of success in building tube amps, but two things come to mind:

1. whatever drawbacks it does have (and I agree there are some), the star method has this advantage. It provides a great visual feedback cue. You can without grabbing a schematic or trying to remember the value of this or that simply glance at the star and immediately tell if a wire has come disconnected, and do a quick "head count" of all the various components that need to be grounded, assuring that they are, or that they need to have a lead running from them to the star. That's maybe a pessimistic way of looking at it, and I suspect that people really confident of their soldering or design capabilities might not need the emotional comfort that a star with all its little "arms" connected gives.    That said, there are functional disadvantages, as some others have noted.

2. I wonder if the way that we think of grounding is very much a cultural trait or effect. It's as if the ground side of the circuit is the poor step child. It's even called the "negative" side and is often denoted with black wire (black hat = villian in the movies), and a minus symbol. No wonder grounds have self-esteem problems. There's also a sense that it's the side you don't have to worry about. After all, as someone pointed out above, all you have to do is usher the "ground" out by way of a wire stuck in the ground or attached to a water pipe. The positive side of the circuit is then made to seem like the "real" side. That's probably arbitrary and not very accurate, but I guess it's so ingrained in the way that we learn about or think about electronics. We think nothing (well, almost nothing) about just attaching the stray black wire to the side of the case (if you've done this, you don't have to admit it. I have, even using a metal binder clip as my "solder" while I worked on some other part of the circuit!). It's as if we figure that the electrons will find their way home, so why bother. We'd never take a fat red wire and just connect a few things together on the + side of the circuit.

Nice discussion, and great forum overall. When I'm brave enough I'll post some jpgs of my DIY projects. For now I'll just go back to lurking and occasionally soap-boxing.