Help! New PC, new noise! SOLVED

This problem has been bugging the past 6 months or so because I built a PC with one of those cases a couple years ago. The noise problem wasn't as bad when I used active Fishman Fluence pickups, but since I moved back to passive pups, I get more noise. It's not bad enough to really annoy me when I'm just practicing/jamming, but if I do any recording with high gain, it does make me move to another part in the room while I track that part.

I've been thinking about building a DIY EMI shield that I could just put over/around the PC while I'm recording. Do you think something as simple as aluminum foil wrapped around a cardboard "wall" would make a significant difference? Do I need to make a complete faraday cage around the PC?

You're answer is probably, "I don't know... give it a try!" but I'm just curious since it sounds like you're knowledgeable on the topic.
I've thought about covering the PC case window with copper foil tape because I have a large roll of it for shielding guitars with but I think I'd have to scrape off some of the case paint to get good conductivity because the paint acts as an insulator.
 
USB has a flaw in its design in that it causes a ground loop because both ends are grounded. You can use a USB Isolator to break the ground loop.
Just out of curiosity: Do Thunderbolt 3 cables also have both ends grounded? I seem to have the same problem with my UA Apollo X8, and, while TB3 isn't exactly USB-C, it sure has a lot of commonalities.

Are there isolators for TB3 cables that fix this issue, if it indeed exists for TB3 cables?
 
Last edited:
There's your problem. Glass does not provide EMI shielding. Computers with glass windows are the dumbest thing ever invented. All the EMI goes right through the window and into your guitar pickups. You'll notice that Dell, HP, Lenovo, etc. don't sell computers with windows. Why? Because they wouldn't pass FCC/CE emissions tests.

Do a search as this has been discussed here many times before.
I'd expect that to be a large part of EMI noise I get using an Apple iMac Pro. It's a big sheet of glass that's just above the motherboard. Maybe I'll give it to my SO and get a loaded M1 Mini and separate monitor.
 
Can you share how you fixed it and what cable you modified and how?

I've tried disconnecting the ground on the USB A side, USB B side and both and neither made much of a difference. I've tried removing the ground from my power conditioner that everything is plugged into. I've tried using a APC Back-UPS BX1000 and disconnected from the power and ran off of battery. I've turned off all the lights in the house and even turned off the main circuit breaker to the house, so there was no power to anything and just ran off the UPS battery.

One thing that helped was to move my PC's from the right side of my desk to the left side and keep all my audio gear on the right side. I'm right handed, so when I sit at my desk the guitar pickups where directed at my PC's that used to be on the right side and were picking up a lot of noise. Now that I've moved my PC's, the noise has been reduced but I'm still getting a really high frequency noise above 2k. I've tried using a low pass filter as the first thing in my FX chain but it makes it lose definition and sounds muffled.

This basically explains how I fixed the issue: https://gearspace.com/board/music-computers/1298283-solved-usb-ground-loop-noise.html

It's not ideal but it will do for a couple of days until my buddy fixes a cable for me. I do use a humbuster cable from the Axe Fx to the amplifiers, and that already helps but the ground switch on my loadboxes also always needs to be depressed to prevent a loud feedback loop. I would advise moving your PC as far away as possible and on the floor instead of on the desk.
 
This basically explains how I fixed the issue: https://gearspace.com/board/music-computers/1298283-solved-usb-ground-loop-noise.html

It's not ideal but it will do for a couple of days until my buddy fixes a cable for me. I do use a humbuster cable from the Axe Fx to the amplifiers, and that already helps but the ground switch on my loadboxes also always needs to be depressed to prevent a loud feedback loop. I would advise moving your PC as far away as possible and on the floor instead of on the desk.
@Guitarjon Thanks! I've actually read that post before. Do you tape over the ground on both ends on the cable or just the USB A side?

Also, I wonder if the +5v should be taped over because it's not needed when a device uses it's own power supply?
 
@Guitarjon Thanks! I've actually read that post before. Do you tape over the ground on both ends on the cable or just the USB A side?

Also, I wonder if the +5v should be taped over because it's not needed when a device uses it's own power supply?
While the 5V may not be used by the device for power it's still necessary. The peripheral uses the 5V as a detect. 5V is connected to the peripheral's Vbus pin which tells the peripheral that a host is connected.

Disconnecting the ground is dangerous. While it can solve noise due to a ground loop it increases the risk of an overvoltage on the Vbus and data pins which can damage the USB controllers (this is why the spec calls for ground at both ends). The correct way to solve a USB ground loop is to make sure the peripheral and PC are plugged into the same power strip. Keep all power cables bundled close together as possible.
 
Last edited:
Last edited:
I found this device: https://ifi-audio.com/products/idefender-plus/ It has USB 3.0 support so it should have enough speed. Also this one: https://ifi-audio.com/products/isilencer-plus/

Now I just need to figure out which one to get as they're not cheap here. Here's a guide on the differences: https://www.listeningpost.co.nz/new...isilencer-what-difference-__I.443010__N.25567

A little technical for me but I think the iDefender+ should do the trick?

EDIT: going for the iDefender+!

yes, that would be the one i would buy. The only reason I put this option so far down the list is that its the one option that costs money.

This is type of issue is a problem that has been around for a while, and is particularly common amongst laptop users when plugged in. I had been planning for some time to make a design based upon the ADI component I mentioned above however the more recent arrival of these devices had put my project onto a "whenever" schedule as while still a technically interesting project, there would now be no money in it.

I will agree with @FractalAudio that simply removing an earth from a cable may not be the smartest solution and isn't recommended, however for a home based setup the potential difference of grounds in two locations in the same room should not be sufficient to cause any dramatic issue.. (but notice i said"should not", not "won't ") and while a USB isolator does exactly the same thing, the excursion potential will be clamped internally so it is the safe (smartest) way to fix the problem.
 
I found this device: https://ifi-audio.com/products/idefender-plus/ It has USB 3.0 support so it should have enough speed. Also this one: https://ifi-audio.com/products/isilencer-plus/

Now I just need to figure out which one to get as they're not cheap here. Here's a guide on the differences: https://www.listeningpost.co.nz/new...isilencer-what-difference-__I.443010__N.25567

A little technical for me but I think the iDefender+ should do the trick?

EDIT: going for the iDefender+!

Let us know how that works out. Thanks for sharing the comparison. I was about to order the iDefender+ and iSilencer+ from Amazon to see which one works better but I'm going to wait to see how the iDefender+ works for you.
 
This problem has been bugging the past 6 months or so because I built a PC with one of those cases a couple years ago. The noise problem wasn't as bad when I used active Fishman Fluence pickups, but since I moved back to passive pups, I get more noise. It's not bad enough to really annoy me when I'm just practicing/jamming, but if I do any recording with high gain, it does make me move to another part in the room while I track that part.

I've been thinking about building a DIY EMI shield that I could just put over/around the PC while I'm recording. Do you think something as simple as aluminum foil wrapped around a cardboard "wall" would make a significant difference? Do I need to make a complete faraday cage around the PC?

You're answer is probably, "I don't know... give it a try!" but I'm just curious since it sounds like you're knowledgeable on the topic.
I would get some 12" × 12" self-adhesive copper foil squares from StewMac or similar, find a screw somewhere in the case to attach a ground wire, solder the wire to the copper foil, and make it disconnectible via a spade connector inline.
 
Last edited:
It would be really interesting to find out why some computers emit this noise and some do not.
It has to do with poor grounding and poor system design. Ground currents are flowing in the motherboard and/or chassis.

In a well-designed system currents return from a source on a dedicated, local return path. In a poorly designed system (i.e. typical cheap consumer PC) return currents flow over multiple return paths. This causes differences in ground potential at different points in the system. Now when you connect a cable to a different system there will be a potential difference between the two systems and current will flow in the ground of the cable. This current causes an electromagnetic field which goes right into your pickups.

For example, consider a PC with a typical low-cost consumer hard drive. This hard drive has poor power supply filtering (to save $$$). When the drive is active it pulls current from the power supply. The current has a noise component due to the poor filtering. Ideally this noise returns on the ground wire but, often, it returns via a different path, typically the motherboard or chassis. Due to the finite resistance of the ground this noise current induces a noise voltage on the ground wrt to earth. Same thing happens with signal sources, often times worse. A signal current is sent to a receiver and the return current takes some circuitous path back to the source through the chassis, etc. Remember, current always flows in a loop (Kirchhoff's law). If you source current it has to return somehow.

Now you connect this PC to an Axe-Fx (which is designed correctly and whose chassis is very close to earth potential) via a cable. The potential at the PC side is noisy wrt to the Axe-Fx side. This causes noise currents to flow in the cable ground. Now any current that flows in the cable has to return somehow. Some of that current returns by going through the Axe-Fx power cable ground and back into the PC via its power cable ground. Now you've created the infamous "ground loop". A loop with current flowing in it creates a magnetic field (Ampere's Law). Your pickups are magnetic transducers and convert that magnetic noise field into an electric noise voltage.

When it comes to USB, sometimes the easiest solution is to try a different port. The ports on the back of the PC will be at a different ground potential then those at the front/top/whatever.

USB is inherently flawed in this regard. It specifies the cable be connected to system ground at both ends. But you can't remove ground safely as you risk damaging the controller ICs.

This is one reason optical protocols were invented. They break the galvanic connection between systems. Ethernet works well too because it is differential signaling with a fault-protected PHY and no ground connection between systems.

Now all this is different than the $%(* stupid case window BS. Cases with windows are the dumbest idea ever and should be illegal.
 
Last edited:
It has to do with poor grounding and poor system design. Ground currents are flowing in the motherboard and/or chassis.

In a well-designed system currents return from a source on a dedicated, local return path. In a poorly designed system (i.e. typical cheap consumer PC) return currents flow over multiple return paths. This causes differences in ground potential at different points in the system. Now when you connect a cable to a different system there will be a potential difference between the two systems and current will flow in the ground of the cable. This current causes an electromagnetic field which goes right into your pickups.

For example, consider a PC with a typical low-cost consumer hard drive. This hard drive has poor power supply filtering (to save $$$). When the drive is active it pulls current from the power supply. The current has a noise component due to the poor filtering. Ideally this noise returns on the ground wire but, often, it returns via a different path, typically the motherboard or chassis. Due to the finite resistance of the ground this noise current induces a noise voltage on the ground wrt to earth. Same thing happens with signal sources, often times worse. A signal current is sent to a receiver and the return current takes some circuitous path back to the source through the chassis, etc. Remember, current always flows in a loop (Kirchhoff's law). If you source current it has to return somehow.

Now you connect this PC to an Axe-Fx (which is designed correctly and whose chassis is very close to earth potential) via a cable. The potential at the PC side is noisy wrt to the Axe-Fx side. This causes noise currents to flow in the cable ground. Now any current that flows in the cable has to return somehow. Some of that current returns by going through the Axe-Fx power cable ground and back into the PC via its power cable ground. Now you've created the infamous "ground loop". A loop with current flowing in it creates a magnetic field (Ampere's Law). Your pickups are magnetic devices and turn that magnetic noise field into an electric noise voltage.

When it comes to USB, sometimes the easiest solution is to try a different port. The ports on the back of the PC will be at a different ground potential then those at the front/top/whatever.

USB is inherently flawed in this regard. It specifies the cable be connected to system ground at both ends. But you can't remove ground safely as you risk damaging the controller ICs.

This is one reason optical protocols were invented. They break the galvanic connection between systems. Ethernet works well too because it is differential signaling with a fault-protected PHY and no ground connection between systems.

Now all this is different than the $%(* stupid case window BS. Cases with windows are the dumbest idea ever and should be illegal.
I've been wondering why my studio monitors pick up so much noise when I have my PC on. Thank you so much for this. I need to find a metal panel immediately.
 
Let us know how that works out. Thanks for sharing the comparison. I was about to order the iDefender+ and iSilencer+ from Amazon to see which one works better but I'm going to wait to see how the iDefender+ works for you.

I got the iDefender+ and I think it works! Without the 5v adapter it reduced the noise but it didn't elliminate it completely. Fortunately it worked with a simple 5v PC adapter connected to it! I tried a different 5v adapter first, that didn't work, but my old phone adapter does work! No noise now but I gotta do some more testing.
 
Back
Top Bottom