Carnhill VTT2326 Transformer

[ruffrecords]

Crikey, so much repeated discussion about grounding. Please read Grounding 101.

Cheers

Ian

 

[Emmathom]

Thank you Ian ! I will Indeed read this paper with attention

 

[Emmathom]

Hi Ian !
Thanks for your link. I think we both agree in a certain way...
Grounding is a vast subject... which would deserve a dedicated thread as everyone has their own version and depending on the case different solutions can be recommended.

Taking up your "Mixer Grounding 101" (which is interseting) I drew 3 diagrams.

The first concerns the wiring of a potentiometer placed between 2 stages (tubes in this example). We must admit that the pot's shell is mechanically connected to châssis (so to ground).
The second drawing concerns the wiring inside a preamp (or any symetrical audio device).

One thing not to forget is that we are not always able to use our own XLR cables (where pin1 and shell pin should not be linked). I mean that we can be in a situation where a sound engineer hands you a microphone (or line) symetrical cable and you have to make do with it. However, in some cables the "shell" pin is linked to pin1. And this can happen quite often.

What does it mean? Well, on the side of your device, pins1 are connected to the chassis via your ground bar but also mechanically at each connector by the shell itself. Does this can generate ground loops ? Yes I suppose... How can you avoid it ? I'm afraid you can't (unless you use XLR châssis connectors isolated like plastic ones...)

The third drawing concerns a mic input transformer wiring. Its secondary "-" is connected to the first "-" current point concerned. It is the "-" of the first stage anode power filtering cap (or drain/source/emitter/collector in case of transistor). The shell and/or the electrostatic shield is connected to "-" secondary.

The fourth drawing is about a way to connect châssis ground to safety earth. Using a (strong) diodes bridge isolates châssis from earth until you reach 2x0,6v (diodes breakdown voltage). From 1,2v a volatge leak connects to safety earth. Another way is to connect châssis ground to safety earth thru a low R value in // with a X2 100nF capacitor (10Ω@3w + 100nF X2). I've used these two configurations without troubles.

Any input is more than welcome (we're here to debate & learn from each other)

 

[ruffrecords]

Hi Ian !
Thanks for your link. I think we both agree in a certain way...
Grounding is a vast subject... which would deserve a dedicated thread as everyone has their own version and depending on the case different solutions can be recommended.

Taking up your "Mixer Grounding 101" (which is interseting) I drew 3 diagrams.

The first concerns the wiring of a potentiometer placed between 2 stages (tubes in this example). We must admit that the pot's shell is mechanically connected to châssis (so to ground).

It is important to distinguish between chassis and analogue 0V. The body of the pot is mechanically connected to chassis. No other connection to it is necessary. There is no need for a twin screened cable. These are unbalanced signal inside a device which is not covered by grounding 101.

The second drawing concerns the wiring inside a preamp (or any symetrical audio device)

One thing not to forget is that we are not always able to use our own XLR cables (where pin1 and shell pin should not be linked). I mean that we can be in a situation where a sound engineer hands you a microphone (or line) symetrical cable and you have to make do with it. However, in some cables the "shell" pin is linked to pin1. And this can happen quite often.

What does it mean? Well, on the side of your device, pins1 are connected to the chassis via your ground bar but also mechanically at each connector by the shell itself. Does this can generate ground loops ? Yes I suppose... How can you avoid it ? I'm afraid you can't (unless you use XLR châssis connectors isolated like plastic ones...)

I do not see the problem. Pin 1 should be connected direct to chassis. Interference currents induced in the screen of any connecting cable are routed via the chassis to earth - the important thing is they do not flow along ony signal grounds (analoge 0V). If the shell of the connecting cable is also connected to its pin 1 that just acts as an additional screen. No loop is created.

Note the wiring together of pins 1 is not strictly necessary. I do it only on mic input XLRs as a separate means of connecting the 0V of the phantom power. I would typically connect it to the centre of the string and not to the chassis common point.

The third drawing concerns a mic input transformer wiring. Its secondary "-" is connected to the first "-" current point concerned. It is the "-" of the first stage anode power filtering cap (or drain/source/emitter/collector in case of transistor). The shell and/or the electrostatic shield is connected to "-" secondary.

You are into unbalanced territory again which is not covered by Grounding 101. However, as a rule, you should aim to ensure high output currents do not flow in low level input circuits.

The fourth drawing is about a way to connect châssis ground to safety earth. Using a (strong) diodes bridge isolates châssis from earth until you reach 2x0,6v (diodes breakdown voltage). From 1,2v a volatge leak connects to safety earth. Another way is to connect châssis ground to safety earth thru a low R value in // with a X2 100nF capacitor (10Ω@3w + 100nF X2). I've used these two configurations without troubles.

In the EU (and probably the US too) it is a requirement for there to be a very low ohmic connection between any exposed metal parts and safety ground. Both the schemes you mentioned do not provide this and are therefore non-compliant. Personally, if a high voltage every gets connected to a chassis don't want anything between it and safety ground.

Cheers

Ian

 

[Emmathom]

It is important to distinguish between chassis and analogue 0V. The body of the pot is mechanically connected to chassis. No other connection to it is necessary. There is no need for a twin screened cable. These are unbalanced signal inside a device which is not covered by grounding 101.

Ok but I thought that it would "protect" audio signal going forth and back from circuit to pot in front panel...

I do not see the problem. Pin 1 should be connected direct to chassis. Interference currents induced in the screen of any connecting cable are routed via the chassis to earth - the important thing is they do not flow along ony signal grounds (analoge 0V). If the shell of the connecting cable is also connected to its pin 1 that just acts as an additional screen. No loop is created.

Ok

Note the wiring together of pins 1 is not strictly necessary. I do it only on mic input XLRs as a separate means of connecting the 0V of the phantom power. I would typically connect it to the centre of the string and not to the chassis common point.

Understood

You are into unbalanced territory again which is not covered by Grounding 101. However, as a rule, you should aim to ensure high output currents do not flow in low level input circuits.

In the EU (and probably the US too) it is a requirement for there to be a very low ohmic connection between any exposed metal parts and safety ground. Both the schemes you mentioned do not provide this and are therefore non-compliant. Personally, if a high voltage every gets connected to a chassis don't want anything between it and safety ground.

Cheers

Ian

Thank you for your clear answers

 

[ruffrecords]

Ok but I thought that it would "protect" audio signal going forth and back from circuit to pot in front panel...

Screens only stop electrostatic interference fields. If you simply use a regular single core screen cable with the screen connected to )V analogue you get wxactly the same screening effec

Thank you for your clear answers

No problem!

Cheers

ian

 

[Dan Kennedy]

Here's the basic idea of the supply I use, omitted are input voltage selection, fusing and a capacitance multiplier used in the production stuff. But the general idea is here.