Phantom ref shared by PSU signal ref

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
abbey road d enfer said:
We have a very different conception here of what earthing does. Indeed voltages created by electrostatic fields are "dumped" to earth. Actually they are not, they are prevented to develop by the equipotential surface.
I do not recall using the word dumped. Fault currents need to be routed to safety earth  and for this reason chassis needs to be connected to safety earth. I agree electrostatic interference does indeed rely on screens and chassis acting as equipotentials.

Ian
 
ruffrecords said:
I do not recall using the word dumped.
I didn't suggest that. My word.

Fault currents need to be routed to safety earth  and for this reason chassis needs to be connected to safety earth.
Agreed. We are not talking about fault currents in the commonly accepted meaning, though. We are talking essentially about currents that result from the equipment being immersed in magnetic fields and currents resulting from leakage of transformers and Cy capacitors. Currents resulting from accidental contact or leakage are outside the scope of the subject.
 
abbey road d enfer said:
I didn't suggest that. My word.
OK
Agreed. We are not talking about fault currents in the commonly accepted meaning, though. We are talking essentially about currents that result from the equipment being immersed in magnetic fields and currents resulting from leakage of transformers and Cy capacitors. Currents resulting from accidental contact or leakage are outside the scope of the subject.
Except safety rules mandate that all exposed metal parts, e.g. chassis, be connected to safety ground which impacts the options for managing  other system aspects such as EMC and f gound loops.

Cheers

Ian
 
ruffrecords said:
Except safety rules mandate that all exposed metal parts, e.g. chassis, be connected to safety ground which impacts the options for managing  other system aspects such as EMC and f gound loops.
I don't deny it. I just fail to see how that influences the decision of where to connect the 48V minus.
 
abbey road d enfer said:
I don't deny it. I just fail to see how that influences the decision of where to connect the 48V minus.

Referring to the schematic I posted, what about the loop it creates if you also connect the two chassis together via the power supply to mixer cable?

Cheers

Ian
 
ruffrecords said:
Referring to the schematic I posted, what about the loop it creates if you also connect the two chassis together via the power supply to mixer cable?

Cheers

Ian
Magnetic (ground?) loops in a chassis/safety ground, that are ignored by active input common mode rejection are generally of no consequence. Of course less is always better.

This is why talking about grounds gets so convoluted.

JR
 
ruffrecords said:
Referring to the schematic I posted, what about the loop it creates if you also connect the two chassis together via the power supply to mixer cable?

Cheers

Ian
If you look at my schemo on post #15, you can see that the loop is slightly changed, like it would be changed by using a different wall socket, a different mains cable of a different PSU cable.
 
Assuming one had infinite conductors between all pieces, what would an 'ideal' arrangement look like?  Is it a simple matter of having an extra 48V return wire that lead back to the PSU 0V common connection?  In this ideal case, there is no need to intermingle 48V return and chassis at any point, no?
 
Matador said:
In this ideal case, there is no need to intermingle 48V return and chassis at any point, no?
Assuming a metal chassis is being used, pin 1 should be connected directly to the chassis over the shortest possible length of wire. The length of that wire determines the wavelength of RF that can be radiated inside the enclosure. Otherwise, analog ground and chassis ground should only be connected together at the ground of the filter cap of the power supply. If that power supply is separate, then you can use a separate wire for chassis ground vs analog ground. After that, CMRR of inputs is probably what limits what the noise performance of the device can be.

The external supply for my channel strips uses a through hole PCB mount 9 pin CPC connector. Of the 9 pins, 3 are connected to the ground plane. Those conductors run out to each unit where one is connected to the ground plane of the power input PCB / CPC connector. Then a wire from the ground plane of the remote unit is connected to the chassis. The mic XLR input has a metal spike that connects pin 1 to the chassis. From what I know about E so far, that is what I understand to be "ideal".

Although, the way Ian does it where the branch point is at the remote unit chassis is fine too because all analog references that impact CMRR are downstream of that point.
 
Is this acceptable? Rather than use the PSU as the com/earth point, the frame? 
(Recall my PSU treats phantom/com same)

pOyHYQN.jpg
 
As Abbey has pointed out, the resistance of the dc cables if very low so the force shields are probably unnecessary. There was a time in days of yore when a large star ground wars bolt was fitted as standard to most Neve battleships and a techitickle note was supplied showing how it should be anchored to the nearest planetoid.

Neve-Technical-Earth.png


Cheers

Ian
 
squarewave said:
Otherwise, analog ground and chassis ground should only be connected together at the ground of the filter cap of the power supply.
Wrong. Terrible misconception. This connection must be done at the reference point of the power supply, i.e. in most cases the reference point of the regulators. For a non regulated CLC or CRC supply, it's the  ground of the last smoothing caps. The "dirty" connection between the center tap of the xfmr and the first smoothing caps must be kept as an single antenna with no loop.
 
boji said:
Is this acceptable? Rather than use the PSU as the com/earth point, the frame? 
(Recall my PSU treats phantom/com same)

pOyHYQN.jpg
For safety, the connection to earth should be done at the PSU. Then it's perfectly correct to run a wire from the PSU earth to the mixer chassis.
 
squarewave said:
This is what I would consider to be "ideal" for an audio device in a metal chassis with an external supply.
See attachment regarding "Chassis and 0v ground and chassis ground converge at filter cap ground".
 

Attachments

  • earthing 0v.jpg
    earthing 0v.jpg
    53.1 KB · Views: 26
No, that little drawing about caps was perfect for a guy like me, as I was going to ask.    ;D ;D

Thanks Square and Abby!

And Ian for humoring my space opera drawing.
 

Latest posts

Back
Top