Star Ground Confusion

[mcgruff]

I may be the wrong person to attempt to answer this.. but the circuit only needs to refer to itself: this "common circuit ground" doesn't have to connect to the chassis. The circuit would work just fine.

However, all wire screens should be connected to the common ground (at one end only) and the chassis is in effect another screen so it too should be connected to the common ground.

The safety "ground" - the connection from chassis to true earth via the mains socket earth wire - stops you getting shocked in the event of a wiring fault which connects the chassis to a high voltage line. Any current would find it easier to flow through the safety ground to earth rather than through you to earth. The safety ground has nothing to do with the normal operation of the circuit.

[Enzo]

grounds can be earthed at various places on the chassis, the important thing is watching the current paths through all this groundage.

[Chuck H]

Excellent. But already understood. What I'm considering is a 0V reference for circuits!!! Not being an electronics engineer, I'm having some trouble figuring what constitutes 0V other than an earth ground. Which in every case I've seen is ultimately the chassis!!! In more modern designs the chassis is "earthed" via the third prong (green plug lead). On older amps it's still the chassis, only the chassis floats on the neutral lead.

The way I see it, as it relates to noise rejection, it comes down to circuit resistance. signal leads need to do one of two things... 1) Have a lower resistance path to any 0V reference than high current leads, or... 2) Be separated by some resistance from high current leads that have a low resistance to a proper "earth".

Either one of these scenarios has worked at times for me. On amps that are complicated by intermittent circuits like tremolo and/or reverb it can become quite confounding. Especially when additional transformer noise and high gain circuit leads are in play.

Perhaps I'm missing the fundamental principals. I make it a point to include grounding options in final designs specifically for the events where I may need to change a ground reference for noise. It would be much easier if I understood the missing pieces.

[Enzo]

sure, unless it is a wireless battery deal like a wireless mic or guitar transmitter, the earth will be the ultimate ground reference for the circuit commons. The point is that connection ought not to be carrying noise currents. And that is why Marshall and some others connect the common to the chassis through a little pair of diodes, a cap, and a 10 ohm resistor. All in parallel. The circuit is thus earthed, but unless the noise currents are more than that little circuit, they won;t flow through that connection.

[dai h.]

Excellent. But already understood. What I'm considering is a 0V reference for circuits!!! Not being an electronics engineer, I'm having some trouble figuring what constitutes 0V other than an earth ground. Which in every case I've seen is ultimately the chassis!!! In more modern designs the chassis is "earthed" via the third prong (green plug lead). On older amps it's still the chassis, only the chassis floats on the neutral lead.

wouldn't the "0V reference" be the circuit ground line? If you don't have the chassis, it will still work (but not work well and be noisy). If you don't have the safety ground, it still works, but won't be as safe.

The way I see it, as it relates to noise rejection, it comes down to circuit resistance. signal leads need to do one of two things... 1) Have a lower resistance path to any 0V reference than high current leads, or... 2) Be separated by some resistance from high current leads that have a low resistance to a proper "earth".

so I guess you're talking about "brute forcing" and separating impedances (like star connections), both useful methods. But for 2) if you mean the safety ground as a "proper earth", then that connection isn't necessary for the circuit to work, it's for safety. Airplanes, ipods, satellites, space stations, etc. don't have a "proper earth" (a wire that connects to physical earth) and still work. You use the local low impedance conductive area (which would presumably have to be the metal or conductive chassis). Also, for a skyscraper, the local power "0" is nearby steel framework (closest massive conductor). (That is, (as I understand) you don't run a separate wire all the way to ground.)

So... Should the input be nearest the earth ground or should the high current power supply be nearest the earth ground??? I've seen PCB designs where the power supply is nearest and the preamp is grounded via a buss with a small value capacitor from the input side to the chassis. I'm not sure what this achieves, but I'll hope for an explanation and education on it.

I think that if the circuit grounds are wired correctly, then you can have different (say either near input for power supply end)
circuit ground to chassis connections (not part of the circuit but metal chassis can't be left floating and the circuit ground to safety ground connection needs to be made). (Not completely clear to me but) seems to be different advantages and disadvantages. If the input jack side is the connection (and chassis isn't used for other circuit ground/return connections), seems good from a noise perspective since the chassis is basically just a shield and the return at the jack doesn't have much going on there at all in terms of current flow (and also from helping to keep RF out of chassis). The other end is less clear (re: advantages). Soldano SLO100 is wired this way, with that end to chassis (and safety ground close by IIRC). Leakage currents have closer path to home (AC source)??

re: the small cap (I presume this is from ground to chassis), my understanding is that it's for AC terminating at high freqs, or shunting RF on the cable impedance (the shield = length of conductor RF can develop a voltage over) to chassis (or large low impedance ground area--maybe a plane or mesh). For the purpose of not getting RF into the audio and intermodulating it (which creates non-musically related tones). (Perhaps there will be a trend in the future for more RF-proofed wiring using such a cap, 360 deg. shield terminations, RF filters or whatever.)