Page 1 of 1

#1 Two-chassis phono stage - grounding question

Posted: Wed Jan 12, 2022 6:27 pm
by Thermionic Idler
So I have one metal chassis containing the phono stage circuitry with step-up transformers, and there will be another containing the power supply. An umbilical cable will connect the two with the ±12V supply.

The chassis with power transformer and power supply board will obviously be connected to safety earth because it will have mains cabling in it.

On the face of it, the chassis containing the phono stage circuit is all low-voltage and therefore, on the face of it, shouldn't need to be connected to mains ground. However, I was re-reading today about the issues Ant had with the Cole phono stage and the primary to secondary short that Nick discovered, and whether I should allow for that scenario and ground the whole lot to mains earth (and deal with whatever noise issues that may cause). How usual is it that mains finds its way onto a secondary winding?

#2 Re: Two-chassis phono stage - grounding question

Posted: Wed Jan 12, 2022 6:55 pm
by Nick
How usual is it that mains finds its way onto a secondary winding?
Once will do it.

Take the earth to the second box via a cable that can stand a few amps (15 I think for CE, 24 for USA). Earth both boxes, second one via via the umbilical. Connect the signal ground to safety ground via a 100R 3W resistor with 100nf cap across it at the output RCA's. Connect the Ground Terminal on the phono to the case ground.

At least that's what I do.

#3 Re: Two-chassis phono stage - grounding question

Posted: Wed Jan 12, 2022 9:26 pm
by izzy wizzy
I have wooden 2 box phono. I still take mains earth to the audio chassis for earthing/grounding purposes.

#4 Re: Two-chassis phono stage - grounding question

Posted: Fri Jan 14, 2022 1:09 pm
by Thermionic Idler
Thanks both.

#5 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 2:08 pm
by Thermionic Idler
Nick wrote: Wed Jan 12, 2022 6:55 pm
Take the earth to the second box via a cable that can stand a few amps (15 I think for CE, 24 for USA). Earth both boxes, second one via via the umbilical. Connect the signal ground to safety ground via a 100R 3W resistor with 100nf cap across it at the output RCA's. Connect the Ground Terminal on the phono to the case ground.

At least that's what I do.
Hi Nick - I do have a follow-up question on this. I have XLR3 outputs on this, not RCA's, as my linestage is fully balanced (basically everything except the tonearm is now on XLR connectors). I 'pseudo-balanced' the phono stage's signal output by connecting a 100R resistor between output pin 3 'cold' and signal ground, to match the 100R output impedance of the phonostage. Pin 1 is connected to the chassis ground.

Would it still be necessary/recommended to float the signal ground with the 100R/100nF in this scenario? I can't visualise any way a ground loop could form in the phono stage circuit with the signal earth connected directly to the chassis, say at the tonearm earth terminal?

Many thanks in advance.

#6 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 3:50 pm
by Nick
I don;t follow you description of the output, but I don't see how it would make a difference. You are still connecting grounds together that may or may not be at differing potential, so you may still get ground current being developed.

But you are only making it to work for your system so the restrictions are far less than if it was for general consumption.

#7 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 4:12 pm
by Thermionic Idler
Nick wrote: Thu Jan 20, 2022 3:50 pm I don;t follow you description of the output, but I don't see how it would make a difference. You are still connecting grounds together that may or may not be at differing potential, so you may still get ground current being developed.

But you are only making it to work for your system so the restrictions are far less than if it was for general consumption.
I'll have a go at drawing it out - it'd probable help for my own visualisations anyway

#8 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 5:25 pm
by Thermionic Idler
Nick wrote: Thu Jan 20, 2022 3:50 pm I don;t follow you description of the output, but I don't see how it would make a difference. You are still connecting grounds together that may or may not be at differing potential, so you may still get ground current being developed.

But you are only making it to work for your system so the restrictions are far less than if it was for general consumption.
OK, so here's a quick and dirty schematic which hopefully clarifies things. I thought it would also be helpful to sketch out what happens downstream in the linestage - if my understanding of AMB's schematics is right (they are quite tricky to read) there are separate left and right signal ground planes which the Pin 1's or RCA shells on the input sockets connect to (somewhat at odds with AES48) - they are then floated off the chassis (connected to AC safety ground) by the 10R/100n resistor/cap pairing.

So my question is more around where and how to connect the PCB ground plane in the phonostage - marked with "???" in the diagram. Hopefully that makes sense.

Image

#9 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 6:06 pm
by Thermionic Idler
Maybe I've just answered my own question - thinking this through, the phono stage's ground plane is referenced to the linestage input cold side, and the active output goes to the hot side - the linestage only cares about the difference between hot and cold. So I don't actually need to connect the ground plane to anything else, doing so may inject more noise - I think. The interconnect shields are connected to the case via Pin 1 in accordance with AES48, and so is the turntable earth wire.

I've been doing a lot of reading up on the "Pin 1 problem". I hate trying to solve grounding riddles.

#10 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 6:31 pm
by Nick
Yep, XLR's solve some problems but introduce their own.

#11 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 7:38 pm
by Daniel Quinn
Surely the problems they solve only arise with long lengths of interconnect. Thus for most people they don’t solve anything.

#12 Re: Two-chassis phono stage - grounding question

Posted: Thu Jan 20, 2022 8:43 pm
by Nick
Daniel Quinn wrote: Thu Jan 20, 2022 7:38 pm Surely the problems they solve only arise with long lengths of interconnect. Thus for most people they don’t solve anything.
I would agree with that.

The only thing where I would say there is a small benefit of balanced is if it is done that way through the system and you end up with a valve push pull power output stage where there is no need for an additional phase splitter. Buts that;s very much an edge case.

#13 Re: Two-chassis phono stage - grounding question

Posted: Fri Jan 21, 2022 11:01 am
by Thermionic Idler
Nick wrote: Thu Jan 20, 2022 8:43 pm
Daniel Quinn wrote: Thu Jan 20, 2022 7:38 pm Surely the problems they solve only arise with long lengths of interconnect. Thus for most people they don’t solve anything.
I would agree with that.

The only thing where I would say there is a small benefit of balanced is if it is done that way through the system and you end up with a valve push pull power output stage where there is no need for an additional phase splitter. Buts that;s very much an edge case.
Oh yes, it's certainly true that with the short lengths of interconnect in a typical domestic audio system, there's barely any difference in actual noise rejection between balanced and unbalanced. My argument in favour of balanced operation however, is more to do with the physical separation of the signal return from the interconnect shield. If that standard was consistently applied throughout the whole system, you'd never have to chase down random hums and buzzes from ground loops originating from the AC safety ground ever again.

One of my reasons for going balanced was that I was utterly sick to the back teeth of trying to solve system hums with the commercial gear I had at the time. Plus I wanted to locate the monoblock power amps near the speakers, and have long interconnects running to them rather than long speaker cables. I once went from 3m speaker cables to 7m and heard a noticeable degradation. And balanced operation just has a logic and elegance to it that I like.

Unfortunately the downside is in the phrase I used above - "if that standard was consistently applied throughout the whole system". With commercial stuff, you have no idea how it's been applied, and just one rogue source can royally cock up your carefully crafted grounding arrangement. Maybe I should just route unknowns through an isolation transformer.

Of course if your entire system is Class II then the ground loop problem also goes away - I think? Never quite sure of these things.

I'm off to read Broskie's article again - it's quite a good one.

Edit - I like this article too.