Shield ground wire

I'm not sure if you're asking if the whole ground bus should be shielded, or just the ground for the input network and fist gain stage. If the latter, I found this in Merlin's grounding chapter:






I have some extra Mogami W2549 that I'm planning on using for my build. Sounds great for mics, so I figure a few inches in an amp shouldn't hurt anything.

 

Good questions, and I would add my own question. What’s more important, shielding the ground wire or grounding the cathode at its power supply filter capacitor?

 

The shield is connected to the chassis, which is of course grounded. So, in essence the shield is also grounded and is usually connected to a ground point for convenience. Any potential coupling from other signals is shunted to ground or becomes "zero referenced" or cancelled out.

Differential signals, or "balanced" signals in audio terms, (which use wires like your Mogami Mic cable) will often also have a shield which becomes a second protection against unwanted signals. In this instance, you have a positive signal and a negative signal in the wire and you would not want the shield referenced to either signal and therefore your shield is referenced to ground - which is kind of like your "shielded ground" question above.

A shield can also be connected to a high voltage, which will reject small signals coupling though the shield, but that would be a tech's nightmare if he didn't know it was attached to B+ and accidentally touched it and the chassis at the same time! I would not recommend this!

Sometimes, a circuit will have a separate chassis ground and "signal ground." But this is not as common in guitar amplifiers.

 

The thing to remember about C above is that the ground is carrying all the current running through the amplification stage. Shielded cable is usually not meant to carry much current. A signal cable carrying ground current can create a voltage across the wire and can act as another input to that amplification stage. This doesn't have anything to do with coupling. You'll want to use a more significant wire on your ground to avoid this.

 

The purpose of a shield is to shunt RF noise to ground. You can't shunt ground to ground - it's already there.

 

Check out what gusfinley and tubegeek said above.

Keeping the power filters grounded close to the stages they are filtering is good practice and would be more beneficial than the type "C" shielding referenced above.

In practice you will find builds that violate our best practices yet still provide quiet operation.

 

That makes sense. I have always assumed that shielding could bond to the chassis at any convenient location and no location is better than another. (As long as it’s only bonded at one end.) Is that true?

 

The 'shield' should only be grounded on one end- You can see that on Merlin's diagram "C".
Diagram "B" is a schematic, not a wiring diagram.
Most of the time, I ground the shield to the pots. However, for post phase splitter master volumes, I ground the shield to the ground bus on the turret board.

Ideally, the tube cathodes should be grounded to the ground of their supply capacitor.

For low gain amps, I have the 120VAC power cord grounded near the AC connection, and the circuit grounded at or near the input jacks. Each tube's cathode ground is to its supply cap, and the supply cap grounds are daisy-chained to the circuit ground.

Higher gain amps separate the power tube grounds from the preamp tube grounds.

My amps are well-appreciated for their lack of noise.

 

Not sure what amps you are speaking of, but the older Fenders were not shielded wire. Some of the newer ones were ....starting with the SF...? . I ground the shield on one end only to the chassis.

 

Yup.

Think of a shield as a tubular extension of the chassis. It can indeed sprout out of the sheet metal at any point. And I've seen AM/FM tuners with actual metal tubing bonded to the chassis acting as a shielded route for internal wiring.

To expand on my comment above:

The shield can have a second purpose - it can isolate signals from each other to prevent unwanted capacitive or inductive coupling. So, for example, Kevin O'Connor suggests shielding heater pairs to prevent hum from coupling into signal runs. You think regular heater wiring is a pain in the tuchus? You ain't seen nothing!

But primarily a shield's purpose is to intercept RF that might use signal wiring as an antenna of opportunity. That's why it's on the outside of the cable. When it picks up the RF first (because RF comes from outside the cable), it shunts the unwanted interference to ground and the signal wiring is unaffected. Ha ha nice try WCBS.

Be mindful of the fact that in between "best practices for wiring the Space Shuttle" and "how Leo did it good enough for country swing" there is a wide range and the amount that it matters is somewhere on a spectrum. Also be aware that Leo did not have wifi, wireless mics, and cell transmissions all up in his grill to worry about either.

One thing often overlooked: a noise CURRENT only causes a meaningful noise VOLTAGE in the presence of a high impedance. (That's why long mic runs are not only balanced but low impedance.)

For a whole lot of eye-opening practical information about best practices for building or connecting gear that will behave quietly in a wide range of situations, you could do a whole lot worse than read Rane Corp's archive of technical articles. Note 110, attached, is an example but by no means the only one. Rane.com

Another interesting source of info is Small Signal Audio Design by Douglas Self.

Keeping impedances as low as possible is a huge topic in pro audio transistor/op amp design. With tubes we face inherently somewhat higher impedances and fairly high gain to boot.

On the other hand, even the tightest Champ chassis is a basketball court compared to any miniaturized PCB construction so we've got that going for us.

Over-extended bandwidth above the audio range can be much more of a hindrance than a feature. Unless you also want to pick up your text messages on your guitar amp.

A fully-surrounding metal chassis SHOULD prevent RF invasion. But RF is sneaky, some broadcasts are pretty high power, and so we belt n suspenders certain highly sensitive wiring. High impedance nodes with large gain applied to them are the important ones: the input from the jack is first on that list.

 

Are you thinking of a balanced signal? The guitar TS 1/4" type signal is not balanced.

 

There are only a few situations where both ends of the shield are grounded: One of those is lightning protection.

Unless you plan on playing alone in the middle of a field during a thunderstorm, one end of the shield grounded is the way to go!

 

Voltage is a difference of charge between two points.

This is the same for single-ended ("unbalanced") and differential ("Balanced") signals.

The difference between these, however, is that in a single-ended signal the other signal is the "zero" reference point, or what we often refer to as ground.

 

More Like This:

- The signal may or may not be shielded. Shielding does help protect coupling of unwanted signals into low-signal areas of the amp.
- In guitar applications, shield at one end only.
- The ground does not need to be shielded, since the shield is connected to ground.