Attaching preamp in/out signal grounds at chassis vs. on PCB

Status
Not open for further replies.
forr, that link seems to focus on grounding for safety in a worst case fault scenario. I am only concerned with lowering the noise floor and increasing dynamic range. Can you link something that claims that tying all in/outgoing grounds together at chassis has the effect of lowering noise?
You have already proven to yourself that way #2 is valid and can be better than way #1. For fair comparisons, more tests done by various people are of course needed. I have not measured way #2 vs # way 1 yet but I use it without having my 3-ways active system ever suffering from any hum or other parasitic effects. I have no link for such experiments but they can easily be carried out.
 
Last edited:
Leach discusses the two alternatives of audio grounding the input Signal Return from the PCB to Main Audio ground or from input RCAs to Main Audio ground (MAG).

Some years ago I experimented with what turned out to be more than a hundred aternative wirings and none were perfect in a stereo amplifier.
The most resistant to picking up hum, used Leach's RCA to Main Audio Ground as the voltage reference.
A stereo amplifier that has a single MAG can use a single RCA to MAG wire. This allows one to cross couple all the RCA barrels.

I recall posting this information. I have no idea how many adopted it.
D.Self showed something similar, but there is a significant difference in his implementation. As far as I recall, the Self implementation was not as hum tolerant as the Leach version.

BTW.
the single voltage reference wire from RCA to MAG must follow closely the two sets of twisted pair input cables, or two sets of coax input cables.

If the single wire follows a different route it includes a LOOP AREA between the input returns and the voltage reference wire. That loop will impose a current around the LOOP and that current will introduce an interference voltage in the coax/twisted pair input signal return route.
 
Last edited:
Interestingly enough, an RCA ground to chassis connectionb (i.e. essentially star ground at back panel) also is what mass-market amplifiers/receivers tend to have. I guess it's also because they tend to have a lot of connections and mounting all of these with insulation would be a pain, but if hum rejection is better on top of that, nobody's going to complain.
 
The in ground should connect to the ground at the in end of the PCB. The out ground should connect to the ground at the out end of the PCB. This assumes that the PCB uses a ground bus or ground plane. The signal ground should follow the signal.

Where this connects to the chassis is another issue entirely. When working with audio grounds you should ignore the chassis as much as possible. The chassis is a safety ground, not an audio ground.

Like this?
cb50a4032f9a44427eb6c89821b28657.png

Just consider this one audio channel to keep things simple.

One thing that I don't understand though is what you mean by ignoring the chassis. I don't have to connect the chassis to the ground because it isn't being used as a safety ground in my case but in past experiments, I've found that connecting the chassis to the PSU star point reduces the noise floor by about 5dB.

Does anyone have a take on the best point to attach chassis to ground for maximum RF interference?
 
In a preamp, the electric potential difference at the input has only to be multiplied by a precise quantity to give the electric potential difference at the output.

With unbalanced (RCA) connectors, both input and output potential differences become voltages refered to a strictly identical potential called "ground". This is theory.

In fact, there is always some distance between the input ground and the output ground. The path between them is submitted to all kinds of parasitics and some unwanted voltage develops across the impedance of this path.

It is legitimate to try to minimize this unwanted voltage as much as possible. This can be done by making the path between the RCA input ground and the RCA output ground to present the lowest impedance as possible. The metal chassis seems to be the best candidate to ensure this function.
 
Leach discusses the two alternatives of audio grounding the input Signal Return from the PCB to Main Audio ground or from input RCAs to Main Audio ground (MAG).

Some years ago I experimented with what turned out to be more than a hundred aternative wirings and none were perfect in a stereo amplifier.
The most resistant to picking up hum, used Leach's RCA to Main Audio Ground as the voltage reference.
A stereo amplifier that has a single MAG can use a single RCA to MAG wire. This allows one to cross couple all the RCA barrels.

I recall posting this information. I have no idea how many adopted it.
D.Self showed something similar, but there is a significant difference in his implementation. As far as I recall, the Self implementation was not as hum tolerant as the Leach version.

BTW.
the single voltage reference wire from RCA to MAG must follow closely the two sets of twisted pair input cables, or two sets of coax input cables.

If the single wire follows a different route it includes a LOOP AREA between the input returns and the voltage reference wire. That loop will impose a current around the LOOP and that current will introduce an interference voltage in the coax/twisted pair input signal return route.


Andrew can you share a link to Leach's article?
 
Like this?
cb50a4032f9a44427eb6c89821b28657.png

Just consider this one audio channel to keep things simple.

One thing that I don't understand though is what you mean by ignoring the chassis. I don't have to connect the chassis to the ground because it isn't being used as a safety ground in my case but in past experiments, I've found that connecting the chassis to the PSU star point reduces the noise floor by about 5dB.

Does anyone have a take on the best point to attach chassis to ground for maximum RF interference?
SAFETY:
Do not connect the Chassis to the RCA barrels.

Connect the Chassis (either directly, or via a Disconnecting Network) to the Main Audio Ground to provide protection for a Mains Live Fault. This connection MUST be capable of passing Fault Current approaching kA (1000A). This connection is NOT Audio.

AUDIO:
Connect a two wire to the input socket and take those two wires all the way to the Receiver circuit. This must be a close coupled pair.
Connect the other input socket using a close couple two wire pair to it's Receiver circuit.
Add on the Common RCA Return wire and pass it allong the same route as the two sets of two wire signal connection to the Receiver circuits and from there to the Common grounding point you have chosen for your voltage reference.
 
Last edited:
shredhead said:
One thing that I don't understand though is what you mean by ignoring the chassis. I don't have to connect the chassis to the ground because it isn't being used as a safety ground in my case but in past experiments, I've found that connecting the chassis to the PSU star point reduces the noise floor by about 5dB.
Design and build your circuit so that it works correctly as an audio circuit with no connection to the chassis. As a separate matter (of safety), connect the chassis to the incoming mains ground. Carefully choose the best place to join the audio ground to the chassis - this will often be at an input or output ground although for some reason most people choose to do it at the power supply. Never use the chassis as part of the audio ground system and especially not part of the PSU ground.
 
Design and build your circuit so that it works correctly as an audio circuit with no connection to the chassis.

I am sorry that, like AndrewT, you only make recommendations but do support them on a purely technical point of view to justify why using chassis as ground reference should be avoided.

As a separate matter (of safety), connect the chassis to the incoming mains ground.

To a point very near the IEC socket.

Carefully choose the best place to join the audio ground to the chassis -

This seems to rely on the old star grounding approach.
A more recent approach thinks in terms of equipotentiality of all ground connections.

this will often be at an input or output ground although for some reason most people choose to do it at the power supply.

Linsley Hood did that, he was plain wrong.
 
forr said:
This seems to rely on the old star grounding approach.
A more recent approach thinks in terms of equipotentiality of all ground connections.
Who mentioned star grounding? I usually use a bus ground. 'Ground follows signal' is a bus ground.

"Equipotentiality of all ground connections" is not "a more recent approach" but a complete myth. Equipotentiality of grounds is what newbies often believe, but it can only be true when no currents are flowing and no magnetic flux links the circuit. Such a circuit would not do very much!

I am sorry that, like AndrewT, you only make recommendations but do support them on a purely technical point of view to justify why using chassis as ground reference should be avoided.
A ground reference must, by definition, be a point. A chassis is not a point.

Star grounds try to approximate a point, but in fact are usually a short uncontrolled bus - in the worst cases involving the chassis too!

Bus grounds give up on the idea of a single ground reference, and instead use the output ground of each stage as the input ground reference for the next stage.

I was trying to keep things simple. A thread answering a particular question is not the best place to teach all the many details of audio grounding.
 
Who mentioned star grounding? I usually use a bus ground. 'Ground follows signal' is a bus ground.

"Equipotentiality of all ground connections" is not "a more recent approach" but a complete myth. Equipotentiality of grounds is what newbies often believe, but it can only be true when no currents are flowing and no magnetic flux links the circuit. Such a circuit would not do very much!

I am not precisely a newbie. The concept of equipotentiality is a valid approach even if it can never be fully achieved.

A ground reference must, by definition, be a point. A chassis is not a point.
By definition, ground reference is a potential, somewhat conceptual, and not a physical point.

Star grounds try to approximate a point, but in fact are usually a short uncontrolled bus - in the worst cases involving the chassis too!
So it is a case to wonder why success can be achieved when intentionally involving the chassis.
 
As I said, the only way to get equipotential is to have no currents flowing. Without equipotential all points are different, so you have to pick just one as the reference. Most grounding problems come from forgetting this, or believing that somehow it doesn't apply.

What is sought is the avoidance of parasitic voltages due to current flowing across the impedance of the paths. Minimizing the impedance between all points is the way to approach the idealness of equipotentiality.
 
What is sought is the avoidance of parasitic voltages due to current flowing across the impedance of the paths. Minimizing the impedance between all points is the way to approach the idealness of equipotentiality.
Not the only way.
Identifying the most susceptable circuits and minimising the interference currents for those circuits, gives more likelyhood of minimising the total effects of the interference signals.
 
Last edited:
forr said:
Minimizing the impedance between all points is the way to approach the idealness of equipotentiality.
As we can't yet have room temperature superconducting circuits the best option is to control where the currents go. RF circuits use ground planes for a completely different reason: they are minimising inductance.

As equipotentiality is unattainable, why not design in a away which does not require it?
 
Andrew can you share a link to Leach's article?

I searched through the current Lo Tim article and could not find the paragraphs.
What he wrote must appear in earlier write ups. There have been a few and I have followed them for about 15years.

These are the two paragraphs from the current Lo Tim that address the input stage coupling.
If you are sure that the hum is due to an internal ground loop, the procedure for breaking this loop is as follows:

Turn the amplifier off and wait for the power supply to discharge. Do not perform this procedure with the amplifier on.
Cut the wire to the central ground on the input side of one circuit board.
Solder a short circuit jumper wire between the ground lugs on the two input jacks.
The circuit board with the cut ground wire is now grounded back through its input ground lead to the ground of
the other circuit board. Use an ohmmeter to verify the new ground connection before turning the amp back on.
The circuit board has two ground leads, both of which connect to the central power supply ground. One lead grounds the signal reference points for the diff amp input stage. The other grounds the power supply decoupling capacitors and provides a ground reference for the protection circuit. R51 connects the two ground leads together on the circuit board. This resistor is small enough to look like a signal short circuit between the two grounds but large enough to force the currents in the two grounds to flow to central ground through the separate wires. This helps to prevent hum induced by power supply ripple currents in the ground system.
 
Status
Not open for further replies.