XLR input to amp module connections?

[Feanor]

For folks like me with minimal electronics background some ultimate clarification would be useful regarding the connection of XLR input pins to an amplifier module, whether by Hypex, ICEpower, Purifi, Pascal or whoever.

Assuming the amp module supports balance input, (as most but not all), do. It's pretty obvious that Pin 2 of the input XLR connector ought to be connected to the module's V+ input pin, and Pin 3 to the module's V- input pin. But what of ...

• the XLR's Pin 1, and also
• the module's input Gnd (ground) pin?

RANE and other sources make it clear the the XLR Pin 1 ought to be connected to chassis ground and apparently nowhere else ...
.

.
But my observation of both Hypex and ICEpower modules, amp modules had a input Gnd pin, (as well as V+ and V- input pins). Further I've observed that standard wiring sets for these provide a shielded cable consisting of a pair wires connecting to the V+ and V- pins and a shield connected to the Gnd pin.

It's totally obvious that the respective wires ought to connect XLR pin 2 to the input V+ pin, and the XLR pin 3 to the V= pin. But where (if anywhere) does one connect the shielding at the far end of the cable from module input?

Apparently one does NOT connect the cable shielding to the XLR pin1 according to RANE et al. One then does one do with it?

Depending on this answer, then also, what is the purpose of the module's input Gnd pin? Is it purely to drain interference collected by the cable shielding back to the amp module rather than allow it to affect the V+ and V- wires?

How to connect a single-end, RCA, input to a module with V+, V-, and Gnd inputs is another question, but not one I'm posing at the moment.

 

[sergeauckland]

XLR pin 1 on the equipment should be connected to the chassis and nowhere else.

However, on the cable, pin1 should be connected to the shield at both ends in my view. In all my years in broadcasting, I've never seen it done any other way. Connecting the shield at one end only may be OK for short domestic lengths, but where cable lengths are 10s or 100s of metres, I think it Good Practice to connect the shield at both ends.

This is particularly important with microphone cables where the metal case of the microphone should be grounded through pin 1.

Any currents in the shield due to earth potential differences should be rejected by the balanced and differential input.

How to connect a single-ended output to a differential input is easy, it's the other way round that's hard!

S.

 

[AnalogSteph]

I was just scratching my head about the recommendations on input hookup given in ICEpower datasheets the other day. The 100AS2 one explains it better than the 700AS2 IMO, but still it isn't obvious how they'd want you to connect shield in a pure, 100% balanced power amplifier. (Looks like that one would need a buffer for sure. And different input impedance for +in and -in? Oh c'mon.) I'd just follow AES48 recommendations to keep shield currents away from everything else and do just that.

EDIT: Looks like the smaller modules like the 100AS2 and 200AS2 have different input impedance on IN+ and IN- of their "balanced" inputs (2.8k/1.8k and 3k/1.8k, respectively). Great.

 

[March Audio]

XLR pin 1 on the equipment should be connected to the chassis and nowhere else.

However, on the cable, pin1 should be connected to the shield at both ends in my view. In all my years in broadcasting, I've never seen it done any other way. Connecting the shield at one end only may be OK for short domestic lengths, but where cable lengths are 10s or 100s of metres, I think it Good Practice to connect the shield at both ends.

This is particularly important with microphone cables where the metal case of the microphone should be grounded through pin 1.

Any currents in the shield due to earth potential differences should be rejected by the balanced and differential input.

How to connect a single-ended output to a differential input is easy, it's the other way round that's hard!

S.

This.

I think we need a sticky as the question is asked on a weekly basis.

 

[March Audio]

For folks like me with minimal electronics background some ultimate clarification would be useful regarding the connection of XLR input pins to an amplifier module, whether by Hypex, ICEpower, Purifi, Pascal or whoever.

Assuming the amp module supports balance input, (as most but not all), do. It's pretty obvious that Pin 2 of the input XLR connector ought to be connected to the module's V+ input pin, and Pin 3 to the module's V- input pin. But what of ...

• the XLR's Pin 1, and also
• the module's input Gnd (ground) pin?

RANE and other sources make it clear the the XLR Pin 1 ought to be connected to chassis ground and apparently nowhere else ...
.
View attachment 79709
.
But my observation of both Hypex and ICEpower modules, amp modules had a input Gnd pin, (as well as V+ and V- input pins). Further I've observed that standard wiring sets for these provide a shielded cable consisting of a pair wires connecting to the V+ and V- pins and a shield connected to the Gnd pin.

It's totally obvious that the respective wires ought to connect XLR pin 2 to the input V+ pin, and the XLR pin 3 to the V= pin. But where (if anywhere) does one connect the shielding at the far end of the cable from module input?

Apparently one does NOT connect the cable shielding to the XLR pin1 according to RANE et al. One then does one do with it?

Depending on this answer, then also, what is the purpose of the module's input Gnd pin? Is it purely to drain interference collected by the cable shielding back to the amp module rather than allow it to affect the V+ and V- wires?

How to connect a single-end, RCA, input to a module with V+, V-, and Gnd inputs is another question, but not one I'm posing at the moment.

The module input cable screen should be connected to chassis elsewhere, not at pin 1.

see page 9 specifically but its all good info.

https://www.diyaudio.com/archive/bl...d1460406090-bruno-putzeys-micropre-g-word.pdf

"At some point the problem became so prevalent that the AES had to enshrine the obvious into a standard. Called AES48, it patiently explains that the shield should be connected to the chassis via the shortest possible route and that connections between the PCB ground and the chassis should be made elsewhere. "

 

[bigjacko]

However, on the cable, pin1 should be connected to the shield at both ends in my view.

I thought Don said that connecting both ends of pin 1 will result in ground loop, connect to only one end will still get shielding but no ground loop. Which one is the correct one?

 

[AnalogSteph]

I thought Don said that connecting both ends of pin 1 will result in ground loop, connect to only one end will still get shielding but no ground loop. Which one is the correct one?

This is what you have to do with a Pin 1 Problem afflicted input. It would be good to know in which context he said that.

 

[gene_stl]

There are also some different american and european (or is it japanese) conventions about xlr connectors.
I will look and see if I can find those. Maybe someone else knows the conflict off the top of their head.

 

[bigjacko]

This is what you have to do with a Pin 1 Problem afflicted input. It would be good to know in which context he said that.

There is a thread called ground loops 101, here is the link https://www.audiosciencereview.com/forum/index.php?threads/ground-loops-101.7162/ .

 

[March Audio]

There are also some different american and european (or is it japanese) conventions about xlr connectors.
I will look and see if I can find those. Maybe someone else knows the conflict off the top of their head.

I think it may be more accurate to say there are "global misunderstandings" re xlr connections

 

[sergeauckland]

I thought Don said that connecting both ends of pin 1 will result in ground loop, connect to only one end will still get shielding but no ground loop. Which one is the correct one?

Ground loops aren't a problem with differential and balanced inputs and outputs as any hum is rejected. RF is a problem, made worse by only connecting the screen at one end. A balanced system's CMRR gets much worse as frequency rises, so there's more benefit to be had by connecting the screen at both ends than might possibly by lost through hum which should be rejected.

S

 

[Feanor]

...
"At some point the problem became so prevalent that the AES had to enshrine the obvious into a standard. Called AES48, it patiently explains that the shield {i.e. via Pin 1} should be connected to the chassis via the shortest possible route and that connections between the PCB ground and the chassis should be made elsewhere. "
...

That, then, seems to nail the correct connections pretty succinctly. .The shield on the cable from the amp module to the input XLR should be connected directly to the chassis at the nearest practical point -- definitely not to Pin 1.

 

[AnalogSteph]

There are also some different american and european (or is it japanese) conventions about xlr connectors.
I will look and see if I can find those. Maybe someone else knows the conflict off the top of their head.

Do you mean pin 2 hot (now standard) vs. pin 3 hot (found on older equipment, especially 1980s and earlier)? That's not really a super major issue.

There is a thread called ground loops 101, here is the link https://www.audiosciencereview.com/forum/index.php?threads/ground-loops-101.7162/ .

Specifically, he says:

Ground loops can still form but do not impact the signal. However, it is possible for ground currents to couple internal to the component through various “sneak” paths in the circuits.

The "sneak paths" he is referring to would be what's otherwise called a Pin 1 Problem. These should not occur if the equipment is constructed to be AES48-2005 compliant.

That, then, seems to nail the correct connections pretty succinctly. .The shield on the cable from the amp module to the input XLR should be connected directly to the chassis at the nearest practical point -- definitely not to Pin 1.

Mind you, it might find itself right next to the Pin 1 to chassis connection there.

This is pretty much what the Hypex UcD wiring guidelines are suggesting, too. Would be interesting to know what those modules are actually doing with it - use a resistor similar to ICEpower or have a dedicated chassis ground that currents can just disappear in without bothering anything else à la AES48 (with a Bruno Putzeys designed amplifier, I wouldn't be surprised to see the latter).

 

[Speedskater]

Neil Muncy (RIP) first wrote about the 'pin 1 problem' in 1995. It took about a decade for pro equipment manufactures to understand AES48. Some audiophile equipment manufactures still haven't got the message.

 

[gene_stl]

I having "been away" from audio missed the whole thing and only noticed it in the manual of one of my new pieces of gear either the SchMarantz AV 7005 or the Oppo or maybe a power amp. I don't remember which But I was going to drill down all the way before making connections.
For me 2005 is recent. Could we have references explaining the misunderstanding for those of us that are new to this annoyance. TIA

 

[misterdog]

This.

I think we need a sticky as the question is asked on a weekly basis.

Though in my opinion when we are on here chasing the last 0.00001 % THD we should just bin the 1940's technology of single ended RCA as it offers nothing other than nostalgia.

I binned it years ago. When I saw vendors claiming that some carbon/unicorn loaded RCA cable could work magic for only £ 300/Metre while the musicians/studios were using Van Damme star quad for less than £ 2 /Metre.

RCA/phono was cutting edge in the days of the gramophone.
The live pin makes contact first, if someone designed that today we would all roll around the floor laughing.


Don's tin hat awaiting incoming.

Those who still swear by cross ply tyres will be along to swear that RCA is 'good enough'.

 

[sergeauckland]

Though in my opinion when we are on here chasing the last 0.00001 % THD we should just bin the 1940's technology of single ended RCA as it offers nothing other than nostalgia.

I binned it years ago. When I saw vendors claiming that some carbon/unicorn loaded RCA cable could work magic for only £ 300/Metre while the musicians/studios were using Van Damme star quad for less than £ 2 /Metre.

RCA/phono was cutting edge in the days of the gramophone.
The live pin makes contact first, if someone designed that today we would all roll around the floor laughing.


Don's tin hat awaiting incoming.

Those who still swear by cross ply tyres will be along to swear that RCA is 'good enough'.

As someone who espouses the 'Good Enough' philosophy, I completely agree that the RCA connector is an abomination. The live pin making contact first is about the most stupid thing ever. For unbalanced connections, the BNC is my favourite, hell, even the Belling-Lee coax connector is better.

Cross-ply tyres were also OK when cars didn't do more than 50mph!

S

 

[misterdog]

Bruno Putzeys, John Westlake and Tom Christiansen et al. all design balanced by default, they know far more than I ever will regarding modern audio engineering.

The 'it's only necessary with long cable runs or a noisy environment' brigade that beat me for years are slowly quietening down. The fact that a £ 10 quad core pro audio balanced cable is as good as it gets, swung it for me years ago.

Spend your money on quality engineering not foo cable.

 

[waynel]

Bruno Putzeys, John Westlake and Tom Christiansen et al. all design balanced by default, they know far more than I ever will regarding modern audio engineering.

The 'it's only necessary with long cable runs or a noisy environment' brigade that beat me for years are slowly quietening down. The fact that a £ 10 quad core pro audio balanced cable is as good as it gets, swung it for me years ago.

XLR FTW!

 

[misterdog]

Must try that AHB2 one day.....