Reflection on grounds

[jfutral]

I said in another thread that it seems proper grounding is the new black, at least to us newbies/occasional hobbyists. OK. At least to me, anyway. Or, maybe I heard it way back when, but I didn't grok right away.

It just sort of dawned on me, though, for tube amps, the concept of a ground is actually mind blowing. All those points tied to a shared (common, I get it) point, but they (with the usual exception of unwanted noise) don't interfere with each other. Kind of amazing if you think about it.

Just sayin'.

Joe

[R.G.]

There is a fundamental problem with the way "grounds" are learned in today's electronics society. We're told it's a point of zero volts that cannot be changed, then never told more about it. And almost all of the teaching about electronics focuses on voltages, at least partially because voltages are much easier to measure.

There are two problems with "ground" as it's conceived by neophytes. (a) you have to think about **currents** not voltages, and (b) all wires are resistors (or impedances, even worse) and not short circuits. Those two things are fundamentally at odds with the way electronics is normally taught.

[printer2]

Taught a few courses and in the beginning it is less confusing to the student to have ideal components therefor pure resistances and wires with no voltage drop. But when you move from the theoretical to the real world layout and the part limitations creep in. Like they say, you have to learn to walk before you run. Once a person has the basics down then they can consider the practicalities.

[loudthud]

I remember when I was first starting out, everything I built had hum unless it was battery powered. Even when I attempted to use Zener diodes to regulate the supplys, there was hum. Ofcourse I didn't have an oscilloscope back then, just a Radio Shack VOM.

I think the first thing I ever built that didn't have objectional hum was a line level 800Hz crossover. I just looked at it and I didn't have a clue. Just got lucky when I improvised the power supply.

[jfutral]

The idea of current, i.e. flow, is an important concept, to me anyway. My background is in theatrical lighting, so I deal with lots of 240/120v, 3 phase power, often, and the occasional 207v power. Thinking of "ground" in terms of voltage makes a bit more sense in that context. Especially when the fire marshall who wrote the electrical code is inspecting your outdoor rig and making sure you drive two 10' copper spikes at least 6' into the ground and at least 20' apart. Plus he wants all your outdoor runs to have GFCIs. Try finding 20 of those with 2p&g connectors on a Sunday. I am used to thinking in terms of watts, volts, and amps but with a slightly different application.

I would always cringe when I saw AV techs tying their ground to their neutral in a 4 wire tie-in (4 wire plus ground run). But I was always at odds with the audio techs who want to make sure their gear is independent of the lighting (especially any backline) and heaven forbid any of my power lines run too close to their audio runs. 6-12" parallel or perpendicular is fine, but not rubbing shoulders, so to speak.

"Ground" takes on a different idea in that context, more earth than ground, sort of. Reading Merlin's article talking about flow and thinking more fluid than particles really clicked for me. If I think of electronics grounds as creating ebbs, tides, and flows of electrical currents in a pond of sorts, then all the grounding principles make far more sense. For me.

Although I did learn the realities of wire resistance when my 14 gauge but still 350' run feeding my Front of House lights were always dimmer than my onstage lighting.

Joe

[Steve Conner]

Once upon a time, some employees of a small electronics company were walking home from work. They were just passing by the reservoir that supplied water to their town, when a buzzard flying over it let out a copious stream of crap.

"That's disgusting", said the receptionist, "He pooped in our drinking water!"
"It's not an issue" said the design engineer, "The poop is small compared to the water and can be neglected"
"No, it's a layout issue" said the PCB guy, "A good layout can help keep the poop separate from the potable water."
"Actually, it's a quality assurance issue" said the head of QA, "They just ship it and hope the customer doesn't notice"

And so it is with grounding in audio equipment.

[diagrammatiks]

Once upon a time, some employees of a small electronics company were walking home from work. They were just passing by the reservoir that supplied water to their town, when a buzzard flying over it let out a copious stream of crap.

"That's disgusting", said the receptionist, "He pooped in our drinking water!"
"It's not an issue" said the design engineer, "The poop is small compared to the water and can be neglected"
"No, it's a layout issue" said the PCB guy, "A good layout can help keep the poop separate from the potable water."
"Actually, it's a quality assurance issue" said the head of QA, "They just ship it and hope the customer doesn't notice"

And so it is with grounding in audio equipment.

i've been doing node/star grounding on every amp i've built since I started messing around with them a few months ago.

They've all got less hum then the production amps I've owned.

It seems really simple to me to get it right.

I'm always confused as to what the big problems are that other people encounter.

[Steve Conner]

Well, that's great! You should write us a tutorial that explains your grounding method. I've heard it said that a general solution for the ground loop problem would make a good PhD.

In production, star grounding can't always be used, because it takes up too much PCB space and adds ground wires, all of which cost money. As DIY builders we have the luxury of being able to take extra time and money to do things right, because we don't need to make a profit to stay in business.

[diagrammatiks]

i mean granted that a lot pcb amps are much more complicated then anything I build.

my fryette ultralead has so much stuff going on in there it's insane. But, that amps pretty darn quite.

A lot of companies and builders just don't seem to care...or else they are bus grounding in the vintage correct style...

the hum is magic.

[Wilder Amplification]

A good majority of hum issues are caused by randomly grounding the center tap of the HT secondary (or the (-) side of the bridge rectifier in BR setups) to some random point on the chassis. Connecting it directly to the (-) side of the first filter cap puts you ahead of the game in hum prevention.

In a non-star ground setup I join the plate and screen filter caps together, then take the screen cap negative to the chassis. I then connect the (+) and (-) sides of the supply directly to the first filter cap. This way the heavy ripple currents are pretty well gone by the time the (-) side of the supply goes to the chassis.

[Diablo]

With the importance of proper grounding, I would hope that all schematics going forward would also include a ground scheme.

[Jazz P Bass]

There seems to be an awful lot of confusion surrounding the term GROUND.
In most cases it is a misnomer.
In a VAC circuit, there is not a ground.
Think of a transformer secondary.
In a VDC circuit the proper term is the reference of the circuit.
In most cases it is ZERO VOLTS.
Not ground.
The reference can be any voltage, depending on the circuit.
In any circuit there can be specific terms for Ground.
Chassis, dirty, signal to name the most common.
The term ground in this case is for making schematics easy to read & circuits easy to design.
The idea is to keep them separate until they are finally terminated.
In most cases they are terminated together.
Hense "Star Grounding".
The truest use of the word Ground should be reserved for the equipment bond to earth.
This is a safety issue. Not a function of the circuit.

[R.G.]

A lot of companies and builders just don't seem to care...or else they are bus grounding in the vintage correct style...

I did an on-PCB star grounding system on the Workhorse amps. They were remarkably hum-free. It took some time and thinking about how to organize the circuit sections and signal flow on the board to get a logical ground star and also keep sensitive stuff away from power lines.

A good majority of hum issues are caused by randomly grounding the center tap of the HT secondary (or the (-) side of the bridge rectifier in BR setups) to some random point on the chassis. Connecting it directly to the (-) side of the first filter cap puts you ahead of the game in hum prevention.

In a non-star ground setup I join the plate and screen filter caps together, then take the screen cap negative to the chassis. I then connect the (+) and (-) sides of the supply directly to the first filter cap. This way the heavy ripple currents are pretty well gone by the time the (-) side of the supply goes to the chassis.

Yep - it's that think-of-the-currents stuff. There are huge currents in the CT of the power transformer. That stuff simply has to go in/out of the (-) terminal of the first filter cap, and nowhere else; otherwise, the I*R drop along the conductor will cause voltages which get into the signal path. The (-) side of the first filter cap then goes to the amp's signal ground point.

With an output transformer, the speaker outs don't necessarily need to connect to signal ground at all unless you're using secondary feedback. If you are using signal feedback, the output transformer secondary leads go to the output jack and the speaker return does NOT go to the signal ground, then to the OT. If the OT secondary goes to the output jack, then you can connect one side of the secondary to signal ground with a wire and not introduce signal feedback into signal ground.
The return side of the power tubes does NOT go into a ground bus, or across the chassis, but goes into a wire that returns to the first filter cap. These heavy currents need to flow on their own conductors, and not in grounds used for signal reference.

I distinguish between several things which are called "ground":
- AC power safety ground
- grounded shields for RF/EMI
- a reference voltage of zero volts
- "sewer ground", the return path for "used electricity" back to the filter caps.

These all get connected together, eventually. But how and where they're connected and separated and the paths taken to the connection are *crucial* to hum issues.

[Merlinb]

There seems to be an awful lot of confusion surrounding the term GROUND.
In most cases it is a misnomer.
The truest use of the word Ground should be reserved for the equipment bond to earth.

I think America would benefit from borrowing the British term "earth". We are fortunate to have two words, one for mains connections to planet earth (yes, earth), and one for circuit references (grounds). Every little helps...

[greekie]

I think America would benefit from borrowing the British term "earth". We are fortunate to have two words, one for mains connections to planet earth (yes, earth), and one for circuit references (grounds). Every little helps...

I guess the american version would be "dirt" ;-)