Ginsly
Well-known member
Ah! Ok didn’t consider that.
Shielded Wire Advice?
- Thread starter Ginsly
- Start date
There's is a thought process of isolating the ground of the output of audio devices and and forcing the inputs to do all the grounding. This works because you only need to ground a shield once.
It creates a better star ground because all individual grounds meet back at the power source and only the power source. Think of an upside down tree. On the tips of the twigs are the 3pdt then the PCB then the jacks. Adjacent are the jacks and cable(only "touches" one jack), then DC connector, then PSU. There's a bunch of branches(pedals) all like that. And they all never touch until the PSU.
The down side is if something breaks...
Shield is compromised at one end of the cable, jack gets dirty, etc. Then you could loose your shielding and pick up noise.
Ground loops on pedal boards are more of a myth, IMO. For ground loops, you need distance and usually things operating on different circuits.
That's why you see them on DIs a lot.
FOH is operating on a different circuit than the stage. That means if you connect grounds of the pedal board and the mic cable to fog, you've completed a loop that goes AAALLLLLLL the way from the pedal to the wall to the main breaker through the mains backplane to the FOH breaker to the FOH mains power plug and through its power supply. And, as we discussed, you only need one end of a shield grounded. Hence, the ground lift.
It creates a better star ground because all individual grounds meet back at the power source and only the power source. Think of an upside down tree. On the tips of the twigs are the 3pdt then the PCB then the jacks. Adjacent are the jacks and cable(only "touches" one jack), then DC connector, then PSU. There's a bunch of branches(pedals) all like that. And they all never touch until the PSU.
The down side is if something breaks...
Shield is compromised at one end of the cable, jack gets dirty, etc. Then you could loose your shielding and pick up noise.
Ground loops on pedal boards are more of a myth, IMO. For ground loops, you need distance and usually things operating on different circuits.
That's why you see them on DIs a lot.
FOH is operating on a different circuit than the stage. That means if you connect grounds of the pedal board and the mic cable to fog, you've completed a loop that goes AAALLLLLLL the way from the pedal to the wall to the main breaker through the mains backplane to the FOH breaker to the FOH mains power plug and through its power supply. And, as we discussed, you only need one end of a shield grounded. Hence, the ground lift.
If you have a ground loop on a pedal board, it's probably not the board.
1)make sure it's on the same circuit as what it is feeding(amp, PC, etc)
2)if all else fails, lift the ground on the output of the board. The other end *should* be grounded, if by nothing else a chassis earth ground.
Getting way off topic but I feel it's all (semi) relevant.
1)make sure it's on the same circuit as what it is feeding(amp, PC, etc)
2)if all else fails, lift the ground on the output of the board. The other end *should* be grounded, if by nothing else a chassis earth ground.
Getting way off topic but I feel it's all (semi) relevant.
I've heard of people sticking hampsters up their butt too
But I wouldn't do it.
If you really wanted to recycle something I suppose you could find some old RG coax cable
DailyDovetails
Well-known member
I have been thinking about this whole shielding/ground loop thing a fair amount recently. It at first seemed a little abstract but I think the noise comes down to the following areas.
Antenna. Although possible none of our equipment other than say a cable which is improperly wired makes a very good antenna. I think the best practice for minimizing this type of noise is to keep the unshielded lengths of wire as short as possible.
Capacitive noise. This is the noise that happens when you have paths with dissimilar voltage close to one another. Like when your input and output wires are close or you have a wire close to your board. This can be minimized by either increasing the distance between points or shielding.
Ground loops. This is when there are multiple paths to ground and current is able to flow between the different paths creating noise. This one was the most abstract to me. I try to think of it like the different ground areas are ponds of water and I am trying to drain them using single streams to one location. So this would be having a grounding plan and avoiding excess paths.
The pedals I have been building recently have been really quiet. I often turn up the gain and volume up all the way even to hear what the noise floor might be. I shield the input and output jacks, make sure all the pots, switches ect have continuity to the case, and make sure the top and bottom of the case have continuity. I sometimes wire the shield to the jack, the footswitch board or both, for me it hasn’t made much of a difference.
The one additional thing I want to try is to print a plastic washer to isolate the output jacks from the case.
What I wonder sometimes about ground loops is if the power supplies we use and the guitar cable create a ground loop. I wonder if the best practice would actually be to wire all the pedals and amp to a single ground and terminate the cable shield to that point.
Antenna. Although possible none of our equipment other than say a cable which is improperly wired makes a very good antenna. I think the best practice for minimizing this type of noise is to keep the unshielded lengths of wire as short as possible.
Capacitive noise. This is the noise that happens when you have paths with dissimilar voltage close to one another. Like when your input and output wires are close or you have a wire close to your board. This can be minimized by either increasing the distance between points or shielding.
Ground loops. This is when there are multiple paths to ground and current is able to flow between the different paths creating noise. This one was the most abstract to me. I try to think of it like the different ground areas are ponds of water and I am trying to drain them using single streams to one location. So this would be having a grounding plan and avoiding excess paths.
The pedals I have been building recently have been really quiet. I often turn up the gain and volume up all the way even to hear what the noise floor might be. I shield the input and output jacks, make sure all the pots, switches ect have continuity to the case, and make sure the top and bottom of the case have continuity. I sometimes wire the shield to the jack, the footswitch board or both, for me it hasn’t made much of a difference.
The one additional thing I want to try is to print a plastic washer to isolate the output jacks from the case.
What I wonder sometimes about ground loops is if the power supplies we use and the guitar cable create a ground loop. I wonder if the best practice would actually be to wire all the pedals and amp to a single ground and terminate the cable shield to that point.
Ginsly
Well-known member
Huh - so even when you grounded both sides of the Input and Output wires it didn't make a difference one way or the other? In that case, why would you need to look into plastic washers to isolate the Output jack from the case?
Since you've tried it in different permutations, which one do you typically go with?
Ginsly
Well-known member
That's a helpful analogy, thanks!
DailyDovetails
Well-known member
I had been doing the separate shield and ground but now I am mostly grounding the shield at the jacks just because it’s neater. The isolating washer is just a best practice type thing that I was considering messing with.
Some of the other things I have run across that are grounded this way is a mixer I recently rebuild which is pretty quiet after rebuilding.
Also EMG pickups are grounded that way. They look like they aren’t but if you pull back the heat shrink the black wires go to the shield.
Stickman393
Well-known member
Hmmm.
Reading through this thread. Have thoughts.
The way I think about a Faraday cage (which is what we're doing with shielded cable) is that you're bonding the cage to the most massive and "potentially" (in the literal, electrical sense) stable point as is reasonably accessible.
Signal ground serves this purpose well.
Earlier someone said that only grounding one end of the shield can create an antenna. This is counter to my understanding of how Faraday cage works.
Look...I didn't go to school for this. I've kinda had to piece together random bits and scraps of information from my own personal study. So take the following with an enormous grain of salt:
The way I think about this: electromagnetism exists all around us. Certain frequencies can influence the electrons in a conductor. Wrapping that conductor in a conductive material that is bonded to a massive object makes it much more difficult for electromagnetism to influence both the conductive material that wraps the signal conductor, and the conductor *itself*.
This is more of less effective based on the amplitude and frequency of the EM. It gets over my head real quick.
Here's the problem with bonding to ground on both sides of the shield: you've now connected to two different points, and those two different points could have different base potentials, as well as two different paths of different resistances to those same points. That causes electron flow in close proximity to your conductor, which generates EMF, which influences the signal that you are supposed to be shielding from interference.
Not only that, but it also magnifies the impact of external EMF. Because the sheild's potential is now no longer tied to a single anchor point, it's potential is floating between two surfaces. This induces current, which is called a "ground loop"
Like I said: don't take my word for it. This is just the way I've constructed this particular model in my head, and I'm open to changing it.
But...I do have professional experience in this regard. In HVAC we have a communication protocol called "BACnet" (Building Automation Control Network) It's generally run with a 22awg/2 conductor shielded cable.
Things get *bad* when folks don't follow good shielding practice. That means making the sheild a complete chain back to a single earth ground point. If folks don't do that: Devices have a tendency to pop in and out of existence, invariably leading to a programmer complaining that the device they're trying to connect to isn't talking to them.
I ask em if they tried flowers and apologizing. It generally doesn't go over well.
Anywho, with that in mind: shielding is great. But most of the time, not necessary. Why? Because your enclosure is typically a Faraday cage. The only time it really becomes necessary is when something in your pedal is generating something that is interfering with your signal. Like, the grand orbiter generates a particularly strong LFO that can easily fuck with your shit. That's a good use case.
But for...like...a dirt box? Nah dude. Don't bother.
Reading through this thread. Have thoughts.
The way I think about a Faraday cage (which is what we're doing with shielded cable) is that you're bonding the cage to the most massive and "potentially" (in the literal, electrical sense) stable point as is reasonably accessible.
Signal ground serves this purpose well.
Earlier someone said that only grounding one end of the shield can create an antenna. This is counter to my understanding of how Faraday cage works.
Look...I didn't go to school for this. I've kinda had to piece together random bits and scraps of information from my own personal study. So take the following with an enormous grain of salt:
The way I think about this: electromagnetism exists all around us. Certain frequencies can influence the electrons in a conductor. Wrapping that conductor in a conductive material that is bonded to a massive object makes it much more difficult for electromagnetism to influence both the conductive material that wraps the signal conductor, and the conductor *itself*.
This is more of less effective based on the amplitude and frequency of the EM. It gets over my head real quick.
Here's the problem with bonding to ground on both sides of the shield: you've now connected to two different points, and those two different points could have different base potentials, as well as two different paths of different resistances to those same points. That causes electron flow in close proximity to your conductor, which generates EMF, which influences the signal that you are supposed to be shielding from interference.
Not only that, but it also magnifies the impact of external EMF. Because the sheild's potential is now no longer tied to a single anchor point, it's potential is floating between two surfaces. This induces current, which is called a "ground loop"
Like I said: don't take my word for it. This is just the way I've constructed this particular model in my head, and I'm open to changing it.
But...I do have professional experience in this regard. In HVAC we have a communication protocol called "BACnet" (Building Automation Control Network) It's generally run with a 22awg/2 conductor shielded cable.
Things get *bad* when folks don't follow good shielding practice. That means making the sheild a complete chain back to a single earth ground point. If folks don't do that: Devices have a tendency to pop in and out of existence, invariably leading to a programmer complaining that the device they're trying to connect to isn't talking to them.
I ask em if they tried flowers and apologizing. It generally doesn't go over well.
Anywho, with that in mind: shielding is great. But most of the time, not necessary. Why? Because your enclosure is typically a Faraday cage. The only time it really becomes necessary is when something in your pedal is generating something that is interfering with your signal. Like, the grand orbiter generates a particularly strong LFO that can easily fuck with your shit. That's a good use case.
But for...like...a dirt box? Nah dude. Don't bother.
Last edited:
Ginsly
Well-known member
This is the conclusion I came to a while back, and is essentially what RG Keen had said in some ancient post somewhere. I just keep seeing people do it in fuzz circuits, and I'm trying to see if anyone has actually lowered their noise floor this way (outside of LFO interference and the like). Seems like some people might have seen results, but more often than not I'm gathering that it's a "couldn't hurt" precaution in high-gain dirt. I'll go ahead and try it on a fuzz soon - I'll have to A/B it, which will require using non-grounded In/Out wire and then grounding the shield at the jacks.
Stickman393
Well-known member
Hmm. Speaking of fuzz boxes:
Imma try to work this out in my head as I type...cause...it helps.
So, lots of fuzz boxes use a positive ground. This can be be a little tricky when intermingling with other pedals. It's helpful to know that voltage, as it is expressed, is a measure of the "distance" between two points.
Basically, on a negative ground pedal, we use ground as our reference point. The power supply will generate a voltage that is 9 volts above that ground reference point.
In a positive ground pedal, ground is still our reference point: we've just hooked the "positive" side of our power supply to the reference point. This means that our power supply is forced to generate a voltage that is *below* that ground reference point.
When you connect signal ground between two pedals: one with a positive ground, and one with a negative ground, you aren't impacting the stability of ground itself. It's simply the middle point between the two operating voltages.
I imagine there is a use case, though, if one doesn't use jacks that directly connect ground to the enclosure. If the circuit ground is isolated from the enclosure, it may make sense to use shielded wire between the input and the output...though I would think the wiser choice would be to simply bond the circuit ground to the enclosure.
I dunno. Maybe there's a consideration with germanium transistors that I'm not aware of. I could see someone going down a rabbit hole of "but this one is a positive ground! I gotta do extra shielding because noise!". Even if, fundimentally, there's no real issue there.
But I could be missing something. Like I said: I am just a humble Stick Man who holds the equivalent of an AA because of a five year apprenticeship program. I know some things about electrical engineering, but the gaps in my knowledge base are vast.
But I also know that it is quite common for inexperienced EEs to over-engineer the devices they design. Which is apparently how the DOD Gonkulator got made.
Imma try to work this out in my head as I type...cause...it helps.
So, lots of fuzz boxes use a positive ground. This can be be a little tricky when intermingling with other pedals. It's helpful to know that voltage, as it is expressed, is a measure of the "distance" between two points.
Basically, on a negative ground pedal, we use ground as our reference point. The power supply will generate a voltage that is 9 volts above that ground reference point.
In a positive ground pedal, ground is still our reference point: we've just hooked the "positive" side of our power supply to the reference point. This means that our power supply is forced to generate a voltage that is *below* that ground reference point.
When you connect signal ground between two pedals: one with a positive ground, and one with a negative ground, you aren't impacting the stability of ground itself. It's simply the middle point between the two operating voltages.
I imagine there is a use case, though, if one doesn't use jacks that directly connect ground to the enclosure. If the circuit ground is isolated from the enclosure, it may make sense to use shielded wire between the input and the output...though I would think the wiser choice would be to simply bond the circuit ground to the enclosure.
I dunno. Maybe there's a consideration with germanium transistors that I'm not aware of. I could see someone going down a rabbit hole of "but this one is a positive ground! I gotta do extra shielding because noise!". Even if, fundimentally, there's no real issue there.
But I could be missing something. Like I said: I am just a humble Stick Man who holds the equivalent of an AA because of a five year apprenticeship program. I know some things about electrical engineering, but the gaps in my knowledge base are vast.
But I also know that it is quite common for inexperienced EEs to over-engineer the devices they design. Which is apparently how the DOD Gonkulator got made.
Ginsly
Well-known member
This is a great point, and although I've only ever used metal jacks that have direct contact with a bare-metal enclosure, I've seen plenty of people use plastic jacks. And come to think of it, I'm mostly seeing shielded wire in Ge fuzz builds - not exclusively, but it's where I've seen them the most.
Stickman393
Well-known member
Yeah, it could be that folks are purposefully trying to avoid grounding the enclosure. For...reasons? Doesn't make a lot of sense to me unless we're talking about the propensity for humans to overthink stuff.
Lucky, I am not human, so that tendency isn't a problem. It's a superpower.
Lucky, I am not human, so that tendency isn't a problem. It's a superpower.
I believe there's merit in using shielded cable in HIGH gain circuits in addition to clocked circuits.
It's all as insurance and not "required"
