Help me choose the best of 3 buss grounding options

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Dear people,

Because a picture is worth a thousand words I have attached a visual representation of my questions below.

In a mock up amp I have made mini star grounds per stage which connects directly to a buss bar as in option 1 on the picture. Amp is very quiet hum wise.

Now for layout reasons I am considering connecting the mini star grounds to the buss bar through a short wire, as in option 2 on the picture.
Would this be detrimental because the local grounds now share a single wire before connecting to the bus bar? Would the connecting wire need to be thick?

For saving chassis space I am also considering option 3. Yes, in this option chassis is used as a ground connector, but the grounds are ordered in mini stars before connecting to chassis and the chassis grounds are ordered in a straight line. Chassis is used as the buss bar.
What would be the downside of this option?
Could a possible problem be that the fields of the power transformer interfere with the chassis ground reference?
 

Attachments

  • Buss grounding options.png
    Buss grounding options.png
    19.9 KB · Views: 138
What dotneck said. But the biggie is : where does the current loop? (Current only moves in loops) Where are you going to get large currents messing with small currents?

Kirchhoff may be long gone but his rules, well, rule. Faraday goes further and says keep your current loops apart.
 
Thanks Dotneck and Thoglette.

I am trying to contain the current loops within the sub stars, this is why I do not totally understand why option 3 would be so bad, other than that the currents in the shield ( chassis) are able to interfere with the others.

Any other knowledge about why I should choose option 1 over 2, or are they equal?
 
Sorry, I've been off the net for a while (finishing a project and then recovering, part of which is to stay away from keyboards).

You answered your question in #4: the chassis is also a shield, which is not ground, if working, by definition.

Don't ground at the input jack - this forces all the high-level signals to travel past your lowest level signal, potentially causing noise problems. Remember that the currents at the input jack are many millions of times smaller than the currents at the output jack.

(These days one can purchase an isolated input jack easily: use one)

If #2 can be done well it will further isolate the stages but it does so at the likely cost of higher earth impedance.
Again, remember that current travels in loops - no current actually travels "to ground", rather it travels _via_ ground back to a power supply decoupling capacitor before magically appearing at B+ again (for the purposes of AC analysis decoupling capacitors are almost short circuits and DC voltages don't exist)
 
Thanks Thoglette. Sorry, I have been away for a while too.

With "ground at the input" jack I mean that I make the connection to chassis at the input jack. In Blencowe's book I read "no circuit current flows in this connection under normal circumstances. It is merely a voltage reference, not a return for current."

Is this false?

Another reason for making the ground connection to chassis at the input jack is that it reduces the ground loop when a mains powered pedal is plugged in the input, but since hardly anyone uses these anymore maybe this is old fashioned.

I thought that 2 would be worse than 1 because the currents are higher in 2 in the connecting wire, thus creating a higher voltage drop. Is this what you mean with "higher earth impedance"?
Can you elaborate why 2 further isolates the stages, I do not see why.
 
Jim, good question and I am left thinking about the genesis of my standard practice.

I've a background in telecoms*; pro-audio and largish transistor amps, none of which I'd get my ground reference at the input: in fact where ever possible I'd be floating the input.

I need to go and think about it. If Blencowe's saying "input" there may be a good reason for it.

* Once I had an install where the same bit of gear was misbehaving in a dozen or so substations at one site. Each piece of kit had a blown zero-ohm resistor on the signal-earth-to-chassis path. Why? That resistor was the only link between the main and back up power supply "grounds", which meant that every time a site moved away from "ideal" normal circumstances (motor start up, 2,000A/20kV relay trip) several thousand amps tried to traverse the resistor. And suddenly "the b****y network is playing up again"
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.