Isolated stompbox power supplies sharing a board, share the ground plane? Or fully isolate?

[kato]

Hey folks, I'm working on a board to handle the 9 separate secondaries on the Weber WPDLXFMR-1 Transformer.

Normally I would assume all grounds should be connected, but if the goal is true isolation to reduce hum, I wonder if it's better keep those ground planes separated as well.

These are 11vAC to 9vDC through an LM317 with a trimmer to set the voltages.



Any thoughts?
FWIW, I self etched a test board with separated grounds and it's been working fine. Before I commit to PCB manufacture, even a small batch, I figured I should ask the collective wisdom of GDIY. Thanks.

 

[moamps]

It makes no sense to use a 9 secondaries transformer if you are going to connect common/ground outputs together in power supply.

 

[abbey road d enfer]

Actually, does it make much sense using floating supplies for pedals that share a common ground circuit?
As much as I concur with the idea of using separate outputs in order to prevent cross-pollution*, I believe that separating grounds is more a marketing exercise than a practical benefit.

A valid reason is when mixing pedals with positive and negative ground.

*I could never succeed in making a Line 6 Roto Machine and Echo Park working together. Whatever the pedal board configuration, it would be noisy as hell. Spurious and whistles, not plain noise. Powering them from two different supplies did not improve the situation enough to make it usable.
I let the Echo Park go, because there is no lack of substitutes, and kept the Roto Machine, which is pretty irreplaceable, at least in the pedal format.

 

[JohnRoberts]

Back last century while at Peavey I made a wall wart replacement transformer that had 6 separate secondaries. These were AC and floating outputs. This product was done for convenience rather than ground isolation, but it surely accomplished that too.

JR

 

[Bo Deadly]

Definitely floating. All of the grounds are going to be connected together through the 0V of the plug jacks. But there's no reason to have them also connected together at the supply. That just creates big loops in which currents could be induced from nearby magnetic fields and equate to noise in 0V. And you might have positive ground devices like PNP germanium pedals or the negative side of a bipolar supply.

Note that even though the polarity of a positive ground device wouldn't be different, the barrel connector might be. So you might want to make the connectors on the boards such that they easily be swapped around to work with various cables. Then you just make the right cables for whatever it is you want to do. Although you do need to use some kind of connector system that can fit through your panel hole(s). For this I would use molex PicoBlade with a cable gland as described in this post.

 

[Newmarket]

I take it this is for effects / pedals or similar that typically run off 9V or 12V supplies and using simple unbalanced connections ?
For this you definitely want to keep the 'Grounds' separate. No point in having the multiple secondaries otherwise. The signal 0V / screen connections will provide the common reference.

 

[kato]

It makes no sense to use a 9 secondaries transformer if you are going to connect common/ground outputs together in power supply.

Ha, thanks. That's what I reasoned. Noise from one pedal could creep its way back and pollute the waters for everything else.

A valid reason is when mixing pedals with positive and negative ground.

Yeah, thanks for bringing that up. I don't have any pedals with positive ground. But a buyer recently bought my Octavia clone on Reverb. The main selling point for him was that I used a "modern" negative ground scheme. Apparently, some of the commercial repros use positive ground for authenticity and they don't play well with his pedalboard.

Back last century while at Peavey I made a wall wart replacement transformer that had 6 separate secondaries. These were AC and floating outputs. This product was done for convenience rather than ground isolation, but it surely accomplished that too.

Cool! Always good to hear Peavey stories from the golden era of music gear made in America. I saw that reality show episode where Hartley dressed in disguise and asked his employee what she thought of him. Facepalm. Officially the end of an era right there.

Definitely floating.

…you might want to make the connectors on the boards such that they easily be swapped around to work with various cables.

Thanks for verifying Bo. Good logic.
I plan to use terminal blocks for the DC side so everything can have a custom connector. I ended up getting one of those extruded enclosures for this project, which should be just barely big enough to fit everything. We'll see.

I take it this is for effects / pedals or similar that typically run off 9V or 12V supplies and using simple unbalanced connections ?
For this you definitely want to keep the 'Grounds' separate. No point in having the multiple secondaries otherwise. The signal 0V / screen connections will provide the common reference.

Gracias!! I appreciate all the feedback. Yes, this is for bog-standard, single-ended 9v stompboxes.

Thanks for putting your heads together for me fellas.

 

[Newmarket]

I think I looked at getting one of those transformers (or similar) some years ago. Until I saw they were from North America an no European mains option.

 

[abbey road d enfer]

Definitely floating. All of the grounds are going to be connected together through the 0V of the plug jacks. But there's no reason to have them also connected together at the supply. That just creates big loops in which currents could be induced from nearby magnetic fields and equate to noise in 0V.

A big loop? Think it over.

 

[abbey road d enfer]

I take it this is for effects / pedals or similar that typically run off 9V or 12V supplies and using simple unbalanced connections ?
For this you definitely want to keep the 'Grounds' separate. No point in having the multiple secondaries otherwise.

Is there actually a point of having multiple secondaries when all voltages are identiacl and same polarity?

 

[moamps]

Is there actually a point of having multiple secondaries when all voltages are identiacl and same polarity?

Definitely. To avoid (multiple) ground loops.

 

[abbey road d enfer]

Definitely. To avoid (multiple) ground loops.

What does a ground loop mean in a pedal board? Where there is no connection to earth.
The situation is similar to a mixing console where power rails are distributed in buckets and a common audio ground runs across.
It would matter if the voltage delivered by the supply was very dirty.

 

[JohnRoberts]

Is there actually a point of having multiple secondaries when all voltages are identiacl and same polarity?

Not make this all about Peavey but late last century we sold many truckloads of wall wart powered SKUs. I basically standardized on a 16VAC 1A wall wart that effectively powered several tens of different SKUs. The beauty of an AC output wall-wart is the flexibility in how to convert to different usable voltage rails. The 16VAC was ideal for generating typical +/-15V rails. The AC supply also supported voltage doublers to step up to make phantom voltage. In one application where I needed the efficiency benefit of full wave rectification I developed a floating supply, and used an active circuit to force the rails to be symmetrical. The 5V digital supply rail current draw was compensated for by a dedicated pass device sucking current from the - rail to balance the digital supply draw out. Inside that SKU the active line level electronics had to settle for +/- 12V rails but that worked adequately for the -10dBV nominal 0VU target.

Last century several manufacturers depending on lump/wall warts (they were attractive for lower agency approval costs) came up with their own different approaches each with different pros and cons. I am comfortable with my decisions to standardize on AC WW power. Peavey's multiple winding transformer never won market success because of the cost (Peavey customers loved the SKU just not the price).

In hindsight I made plenty of late night scribbles about designing an active multiple output power supply with fully floating outputs. Using linear technology this involves wasting lots of power. This might work better using modern switching supply technology but that was not a practical option for me last century.

JR

 

[abbey road d enfer]

I am comfortable with my decisions to standardize on AC WW power.

Of course you wouldn't want to power separate units from the same AC source.

 

[moamps]

What does a ground loop mean in a pedal board? Where there is no connection to earth.
The situation is similar to a mixing console where power rails are distributed in buckets and a common audio ground runs across.
It would matter if the voltage delivered by the supply was very dirty.

Have you finally finished editing the post?

 

[JohnRoberts]

What does a ground loop means in a pedal board? Where there is no connection to earth.

More likely ground contamination, than a magnetic winding loop, but many guitar players using multiple pedals experience ground related noise problems.

In principle the only solid safety ground bond should be via the guitar amp chassis. I am not versed in guitar amp input jack practices but suspect the input jack on many is chassis grounded.

Sundry wall-wart powered pedals strung together with who know what power supply designs. Most are probably 2 wire (so double insulated) but who knows what they might be dumping into signal ground?

For today's TMI (veer) a couple years ago I did some research into an idea I was pursuing combining a cap coupled ground lift with a GFCI protected line cord. My design thesis was to cap couple the ground through a decent sized cap (I don't recall my actual value but probably a 0.1uF or so of the safety caps designed to fail short circuit). My concept was to prevent the hard ground bond at 60/120Hz, while still protecting human safety. I sent my prototype to James Brown to test on his personal rig. At the time James was selling his own guitar pedals, now he is working at Fender. He tested my prototype on his personal rig, with multiple pedals and it was silent.

Many guitar players get shocked either by external mains voltage sources (like from hot chassis mixers via the microphone) or from guitar amps with energized chassis. I sized my cap coupled safety ground to deliver sub lethal current to typical mains voltage exposure. The icing on the cake of my guitar player safety device, was to also sense current flowing in the safety leg and use that to disconnect all three mains connections with a 3 pole relay. I had less than zero chance of getting this UL approved, because I didn't bother submitting it (the UL likes their safety ground bonds).

My general advice now for practical guitar player safety is to cap couple all the grounded metal parts inside the physical guitar. A properly sized cap will be adequately quiet and prevent human exposure to lethal currents in case of exposure to mains voltage coming from any direction. /TMI]

JR

PS: I am sensitive to this because while I was working at Peavey we got sued over an electrocution death when a player got between two Peavey amps, where one was plugged into a RPBG (hot safety ground) outlet. This mis-wired outlet effectively energized one guitar amp chassis. The other properly grounded amp provided the electrocution path. We won in court but the musician is still dead.

 

[JohnRoberts]

Of course you wouldn't want to power separate units from the same AC source.

I never powered multiple units from a single WW even though the 1A probably could.

JR

 

[abbey road d enfer]

Have you finally finished editing the post?

My mind is much slower than my hands.
Here we have a saying: I understand quick but I need a looooong explanation.

 

[JohnRoberts]

my mind gets slower every day....

JR

 

[kato]

Is there actually a point of having multiple secondaries when all voltages are identical and same polarity?

I'm not entirely sure. I figure there must be a reason Weber created a transformer with 11 identical secondaries.

I know that not all stompbox supplies on the market do this. I wrote to the folks at Walrus audio to ask if their Aetos product uses separate secondary windings, to which they replied, “The Aetos does have secondary windings for each output.”

I have to assume there is a potential noise-related issue this solves. Otherwise, why would they spend extra for the custom transformer. Hopefully not just for the privilege of saying they do so.