In defense of 'star' grounding

[Munjibunga]

Originally Posted by 1958Bassman

True balanced audio has a pair for the signal, floated above chassis ground potential, often about 50 Ohms above. I suspect that the twist is reducing the noise. Look into 'common-mode noise'- this is exactly why twisted pair works and it has to do with the magnetic field caused by the current in the wires. The field from one wire cancels the other.

Using quad wire reduces the noise even more than simple twisted pairs because each half of the signal has its own twisted pair.

[SGD Lutherie]

Originally Posted by 1958Bassman

OK, but did you consider the fact that the cover if a pot is supposed to be grounded, to allow blocking noise?

+1

That's why they are metal. If they were totally ungrounded they would add noise.

Ever wonder why a Strat has un-shielded wires from the pickups? It's because twisted pair wires reject certain kinds of noise well enough that braid isn't needed. It's the same reason phone and network cabling uses twisted pair- it's not used for instrument cables because they're handled frequently and this could un-do the twist, rendering any benefits useless.

Except Strat pickup wires aren't always twisted. You can see that Gibson went overboard to make their guitars quiet, as a marketing point. They went with humbuckers with metal covers. Had metal boxes in the control compartment, and used coax cable. Fender on the other hand made very little effort to shield anything.

But a twisted pair can help reduce noise while having lower capacitance than a coax cable. Of course any small runs of coax in the bass or guitar will have their capacitance swamped by the cable run to the amp, if they are passive instruments.

[SGD Lutherie]

Originally Posted by Munjibunga

Using quad wire reduces the noise even more than simple twisted pairs because each half of the signal has its own twisted pair.

Back in the early 80s I used to wire up pickups that way. I used two shielded twisted pair cables. I still have a DiMarzio Super Distortion here I wired like that.

[Eight_Stringer]

Uncertain if this information will help, in what constitutes a conductor with a "ground" connection at some point.

Transmission line theory is a special subject of mine, skip this if moderate technical details are not a required reading, grin.

College department i was working with had one Transmission line demonstration unit, very nice it was, very visual and all that. Yet nothing beats hands on is there, why we play bass and not just watch.
Time Domain Reflectometry ( TDR ) equipment was expensive in the day, no problem, designed one for them. 8 students set ups consisting of, basic audio signal generator, student grade Oscilloscope ( CRO ) rolls of transmission cable, balanced and unbalanced, coax included. 3 position switch functioning as open circuit, Zo termination ( matched to transmission cable, 50, 75 or 300Ohms ) and short circuit. Drive the transmission cable ( TC, even audio cables are transmission cables ) with signal generator ( SG ) correct Zo termination at TC end, CRO terminated at SG drive point. Signal looks fine, then we open circuit the opposite end of the TC. Observed waveform changes markedly, ditto when we short circuit the opposite end. At no time does the drive signal disappear. We have standing waves at audio frequency on the TC, at various time intervals on the TC, we have near zero volts and at other points we have twice the applied signal level.

When a short is applied we have zero volts on the short yet we have maximum signal current. When open circuit , we observe maximum applied signal voltage and zero ( for the most part ) current. At no time do we have a passive condition on the TC when driven, short included. You can place a short at the centre of the TC and you will still observe a signal downstream ( going away from the SG ) of the short circuit, ie past the short circuit. There is no short circuit in the true meaning of the words as understood by most people, of moderate technical level.

If you measure the waveform shape and use the known physical properties of the TC ( velocity factor for one ) you can actually determine the position of a "fault" for the TC, all from the comfort of the SG position. We did not run out and dig up cable faults for telephone lines with faults, without using basic TDR technologies of the day, accuracy of a few feet in lengths up to 20 miles was common.

For David, your a hard man to convince, not my job to do that, rather the intent is to give all that are interested, a chance to make the best of solutions, tried and true over many many decades. You have your solutions gained over many years, not suggesting what works for you does not work, rather when you point out a star type connection system is not valid, then i make engineering principles known for all to assess what may work on what basis.

An Antenna will still radiate a signal even if you put a "ground" connection at some point along the entire length, the magnitude will be effected, yet there will still be a radiated signal. For me "ground" is a variable and not an absolute fixed value.

HTH.

[Stealth]

I'm sticking to my daisy-chaining or "tree-grounding" or whatever I should call it. As long as all the ground planes and pots were connected, none of my basses or my friends' instruments I worked on ever buzzed after that kind of grounding.

[okcrum]

Originally Posted by Eight_Stringer

...
An Antenna will still radiate a signal even if you put a "ground" connection at some point along the entire length, the magnitude will be effected, yet there will still be a radiated signal. For me "ground" is a variable and not an absolute fixed value.

Word. Vonada's Engineering Maxim #1:

There is no such thing as ground.

ref: http://www.amhsys.com/?option=com_co...d=219&catid=44

[khutch]

That reference is a little light on technical details but there truly is no such thing as ground. Treating certain things as if they were ground works often enough that the myth of ground persists. If "grounding" things is the only tool in your interference fighting toolbox then you will eventually run into interference that you cannot fix.

Ken

[Troph]

I suppose "there is no such thing as ground" is a sensible aphorism for those who think ground is somehow different from "common". Colloquially in circuit design, however, the term is used interchangeably, and understood to be the same thing. No one would mistake "common" to mean a universally consistent potential. Just a locally consistent potential.

Also, I don't buy the argument that transmission line theory has any bearing on a guitar circuit grounding runs, which are too short to matter at audio frequencies. The rule of thumb I learned way back in microelectronics class was to worry about transmission line wave interference effects if the run was greater than 1/10th the wavelength of the signal. Granted, that rule of thumb was primarily for communication systems and not for pristine audio quality, so perhaps we could strengthen the requirement by an order of magnitude for a healthy safety margin.

So even at 20,000 Hz (a stretch for most guitars) the wavelength is still about 15,000 meters long. Using our extremely conservative 1/100th wavelength limit, we could safely ignore transmission line effects on audio signals out to about 150 meters with standard conductors.

[1958Bassman]

Originally Posted by SGD Lutherie

+1

That's why they are metal. If they were totally ungrounded they would add noise.

Except Strat pickup wires aren't always twisted. You can see that Gibson went overboard to make their guitars quiet, as a marketing point. They went with humbuckers with metal covers. Had metal boxes in the control compartment, and used coax cable. Fender on the other hand made very little effort to shield anything.

But a twisted pair can help reduce noise while having lower capacitance than a coax cable. Of course any small runs of coax in the bass or guitar will have their capacitance swamped by the cable run to the amp, if they are passive instruments.

The pickup wires in my Baja Tele weren't twisted, either, and it was pretty noisy. I decided to twist them and add a ground wire to the base plate on the bridge P/U and when I fired it up, I was immediately disappointed that it still picked up some noise. Then, I realized that I hadn't paid attention to what was causing it- my laptop was on, right next to the headstock. DOH! Once I turned that off, I heard that it made a definite difference in hum/buzz pickup.

[1958Bassman]

Originally Posted by Eight_Stringer

Uncertain if this information will help, in what constitutes a conductor with a "ground" connection at some point.

Transmission line theory is a special subject of mine, skip this if moderate technical details are not a required reading, grin.

College department i was working with had one Transmission line demonstration unit, very nice it was, very visual and all that. Yet nothing beats hands on is there, why we play bass and not just watch.
Time Domain Reflectometry ( TDR ) equipment was expensive in the day, no problem, designed one for them. 8 students set ups consisting of, basic audio signal generator, student grade Oscilloscope ( CRO ) rolls of transmission cable, balanced and unbalanced, coax included. 3 position switch functioning as open circuit, Zo termination ( matched to transmission cable, 50, 75 or 300Ohms ) and short circuit. Drive the transmission cable ( TC, even audio cables are transmission cables ) with signal generator ( SG ) correct Zo termination at TC end, CRO terminated at SG drive point. Signal looks fine, then we open circuit the opposite end of the TC. Observed waveform changes markedly, ditto when we short circuit the opposite end. At no time does the drive signal disappear. We have standing waves at audio frequency on the TC, at various time intervals on the TC, we have near zero volts and at other points we have twice the applied signal level.

When a short is applied we have zero volts on the short yet we have maximum signal current. When open circuit , we observe maximum applied signal voltage and zero ( for the most part ) current. At no time do we have a passive condition on the TC when driven, short included. You can place a short at the centre of the TC and you will still observe a signal downstream ( going away from the SG ) of the short circuit, ie past the short circuit. There is no short circuit in the true meaning of the words as understood by most people, of moderate technical level.

If you measure the waveform shape and use the known physical properties of the TC ( velocity factor for one ) you can actually determine the position of a "fault" for the TC, all from the comfort of the SG position. We did not run out and dig up cable faults for telephone lines with faults, without using basic TDR technologies of the day, accuracy of a few feet in lengths up to 20 miles was common.

For David, your a hard man to convince, not my job to do that, rather the intent is to give all that are interested, a chance to make the best of solutions, tried and true over many many decades. You have your solutions gained over many years, not suggesting what works for you does not work, rather when you point out a star type connection system is not valid, then i make engineering principles known for all to assess what may work on what basis.

An Antenna will still radiate a signal even if you put a "ground" connection at some point along the entire length, the magnitude will be effected, yet there will still be a radiated signal. For me "ground" is a variable and not an absolute fixed value.

HTH.

I'm sure glad the price of a decent TDR has come down- I use them for AV and security camera installs and it comes in handy for avoiding me wasting time looking for cable faults.

[1958Bassman]

Part of the problem in this, and other discussions about instrument/equipment "grounding", is in the fact that we're really discussing 'bonding' when braided/foil shield and some kind of shielding is used in a cavity. If it's not a current-carrying conductor, it's a shield or bonding conductor.

[okcrum]

Please, guys. It's a philosophy of engineering life, stated in terms of the 70s digital design world. But it also exposes some reality, vs. the theory we love to argue. DEC spawned the team that formed the core of Intel's CPU group from the 90s onward. You know how that worked out.

SGD Lutherie is fundamentally correct. You see, there are two types of fields (and interference, and shielding) here, electrostatic (think sparks from walking the rug in winter), and electromagnetic.* There are also two ways these can interfere with our gear, by conduction (along a cable) or radiation (power transformer leakage or local power wiring).

Radiated electromagnetic interference is why single coil pickups and ground loops hum, and why humbuckers and twisted pair cables reduce hum. Hum, as I stated earlier, is sometimes buzz too. You can usually tell this is what you are hearing if you can turn while holding your bass and the level of hum/buzz changes or disappears. It is inductively coupled, like a transformer.

Radiated electrostatic interference is what SGD Lutherie is talking about here. This is capacitively coupled, and is what the shielding in your bass is there to defeat. At audio wavelengths, the choice of grounding technique (presuming it has been properly implemented) inside your bass will likely make little or no difference in the level of noise you hear from your signal chain. Why? Because it is 40-50 dB below the signal level being made by the bass itself. The ONLY case where it might have an effect would be if you had an active bass with an insane (90-100 dB) amount of gain in the preamp.

Conducted interference is a whole different subject, not as relevant to this discussion.

HTH.

* For reasons not explained here, these fields are actually related in a fundamental way and can be converted one to the other.

[Meddle]

Originally Posted by DiabolusInMusic

If you have a noise that goes away when you touch metal/strings/bridge you have a shielding issue.

This happens with every bass I've ever encountered. If it didn't happen then there would no purpose for the ground wire to the bridge.


I think star grounding became a fad because most Google searches for 'shielding my guitar' takes you to one horrible looking html bombsite of a page, written by some 'electronics guru' who insists on star grounding Telecaster and Stratocaster guitars.

[SGD Lutherie]

Originally Posted by Meddle

This happens with every bass I've ever encountered. If it didn't happen then there would no purpose for the ground wire to the bridge.


I think star grounding became a fad because most Google searches for 'shielding my guitar' takes you to one horrible looking html bombsite of a page, written by some 'electronics guru' who insists on star grounding Telecaster and Stratocaster guitars.

Yes, a fad like orange drop and paper-in-oil tone caps.

I'm assuming you are talking about that guitar site where they show to float the shields with a cap? There was someone over at the music electronics forum that did that and was asking why their instrument was much nosier than when they started!

A lot of that buzzing you get when you don't touch the strings is based on the outlet your amp is plugged into. If you have bad grounds, you get buzzing. And when you have light dimmers and other kinds of noisy lighting, you get buzzing. That kind of buzzing is hard to fix.

But really, no matter how you wire your bass, do it neatly with minimum unshielded wire runs, and use adequate shielding. Magnets on single coils need to be grounded, and it helps to wrap some grounded copper foil around the coil, but don't let the two ends connect as a loop. They need to be insulated from each other. Otherwise you can attenuate some of your high end due to eddy current loading.

[Cadfael]

Originally Posted by SGD Lutherie

Yes, a fad like orange drop and paper-in-oil tone caps.

Oh Dave ...
I like you more with every post!

BUT ...
In a German guitar forum we managed to destroy the "shim are sh*t" myth! And when someone wants new pickups, we try to get the right strings first in 90% of all cases!

It was very funny to see the change. It took some years - but it worked!

I am a real Fender Vintage fetishist (Leo Fender + CBS time).
But I also try to bust as many myth as I can, put Leo "down to his place" and put him even higher ...
I wrote a story about Fender colours - and all my research showed that ash or alder were chosen because of the colour (besides the fact that Leo liked the colour Blond for instruments).

During a "English in the office" course I learnt the English KISS formular (Keep It Short & Simple). To my mind, this is also good for instrument wiring.
No stars +++ but as many as needed and as less as possible (also a metal worker slogan) ...

[warnergt]

Originally Posted by okcrum

...You see, there are two types of fields (and interference, and shielding) here, electrostatic (think sparks from walking the rug in winter), and electromagnetic.*

There are three types of fields here:
1) Electrostatic (electric field)
2) Electromagnetic (interrelated electric and magnetic fields e.g. radio waves)
3) Magnetic fields

Shielding can block (1) or (2).
Humbuckers cancel the effects of magnetic fields.
Sometimes it is called electromagnetic induction but
only because the magnetic field induces an electric potential
in the conductor.

Single coil pickups hum primarily from picking up magnetic
fields; not electromagnetic interference as you claim.
Magnetic fields are strongest near a current source
(e.g. transformer, power lines).

[warnergt]

Originally Posted by SGD Lutherie

On a car the whole cassis is also ground, which is why ground connections are made to it. When those connections get flakey, as on an old car (like the 1980 BMW I used to have), then current will find the next lowest potential, which is often not ground.

So if a ground wire is grounded, how is it an antenna for noise? It that were true, then your grounded shields (like on your coax cable) and strings would also pick up noise. But they don't when they are grounded. That's why they are shields. If you unground them, they will be an antenna for noise.

SGD Lutherie, you just don't get it. You're spreading tons of
misinformation here and don't even realize it. You are
performing a terrible disservice to the TB community. You're
on a one man crusade to rewrite the physics of noise. It's
time to stop.

Anyone who installs car stereos will recognize these two
paragraphs of yours as total nonsense. And it fits the entire
theme of what you're spewing here. A car chassis ground is
fine for headlights and taillights and motors but, for audio, it
is complete noise. When your amplifier is separate from your
music source, you have to ensure that your audio signal
doesn't get referenced to the car chassis ground or you will
have noise that will make your car stereo sound like trash.
It is not uncommon for custom car stereo installations to use
isolation transformers to isolate audio signals from chassis
ground. Standard stereo systems typically use differential
signals that are not referenced to ground.

Grounds can be extremely noisy. Cars chassis make an
excellent example of this. You have trouble comprehending this.

[Troph]

Originally Posted by warnergt

...
When your amplifier is separate from your
music source, you have to ensure that your audio signal
doesn't get referenced to the car chassis ground or you will
have noise that will make your car stereo sound like trash.
...
Grounds can be extremely noisy. Cars chassis make an
excellent example of this. You have trouble comprehending this.

Why are we talking about cars here? Noise due to ground coupling (also called "common impedance" coupling) is simply not a factor for guitar electronics. Ground coupling noise is only interesting when there are two pieces of separately grounded equipment which share a signal line, like between the dash unit and the trunk sub amplifier.

In our case, the guitar is not separately grounded; it is grounded entirely by the shielding of the signal line, and therefore cannot introduce additional ground coupling noise into the amplifier. All of the current which flows across the coupled shield is entirely due to "signal" flow (by signal I mean guitar output).

The noise we are (or should be) discussing in this thread is due to inductive coupling, where the noise sources are the dynamic magnetic fields caused by AC currents in nearby devices and wires.

I think that proponents of star grounding are trying to argue that small conductor loop paths in the ground wiring inside a guitar are susceptible to induced currents, which despite occurring on the grounded side of the circuit can cause hum/buzz somehow. Star ground topology is an attempt to eliminate such loops, which makes induced currents less likely to occur, or so goes the hypothesis.

I'm suspicious for several reasons:

1. It isn't clear to me how the tiny induced currents in such loops could cause any appreciable voltage drop. The resistance in the tiny conductor runs is extremely small... almost immeasurably small.

2. Second, the grounded potential of these wires would resist the induced current, which is why we ground all of the shielding in the first place!

3 Third, I haven't ever seen measurements of this effect which show it to be a significant factor.

[okcrum]

Originally Posted by warnergt

There are three types of fields here:
1) Electrostatic (electric field)
2) Electromagnetic (interrelated electric and magnetic fields e.g. radio waves)
3) Magnetic fields

Shielding can block (1) or (2).
Humbuckers cancel the effects of magnetic fields.
Sometimes it is called electromagnetic induction but
only because the magnetic field induces an electric potential
in the conductor.

Single coil pickups hum primarily from picking up magnetic
fields; not electromagnetic interference as you claim.
Magnetic fields are strongest near a current source
(e.g. transformer, power lines).

At what frequency, then, does your electromagnetic field become a purely magnetic field?

Or should I have said just electric and magnetic fields? There are only two kinds. We've known that since the 1800s.

I'll repeat the part you missed about single coil interference: It is inductively coupled, like a transformer.

[SGD Lutherie]

Originally Posted by okcrum

Or should I have said just electric and magnetic fields? There are only two kinds. We've known that since the 1800s.

I'll repeat the part you missed about single coil interference: It is inductively coupled, like a transformer.

+1

I said all this in an earlier post! And he's repeating it? I haven't been following any of this though.

I said (as you did) that electrical fields cause that high pitched buzz. Shielding helps that. Humbuckers don't, because it's not inductive noise.

Magnetic fields are the 60Hz hum you hear, and humbuckers fix that, but they don't fix the electrical field noise.

So you need both.

Of course we say "electromagnetic" because they are two aspects of the same phenomenon, but are two different things. You can't have one without the other though.

Some people think of the two as being on different frequencies, but they behave as you would expect of two different phenomena.