Greetings. Posted a question earlier about a sound card for my budget basement sound system. I have since bought an ASUS Xonar U7, which has good specs and measured well at the audiosciencereview site. Hooked it up to an old desktop computer and power amp and was disappointed in the amount of computing noise I heard. It was down about 60dB, but extremely noticeable due to the noise characteristics. After playing around with ground loop elimination I eventually realized that the USB ports themselves seem to be noisy. If I plug my headphones directly into the sound card (no amplifier/speakers), I still hear the noise. Tried different USB ports but they all have the same issue. If I use my laptop, the card is dead quiet. Looked at some USB filters, but the commonly available ones only support 12Mbps, which is not fast enough. I have a free PCIe slot and wonder if a $15 USB card would resolve my issue. Any thoughts or other ideas? Thanks.
Thanks for the responses. Audiosciencereview measured SINAD at around 102dB, and when I connect it to my laptop I hear no noise, so it is not the soundcard itself - it sounds great!
I don't see how it can be a ground loop when there is only one path - AC plug --> computer --> USB sound card --> headphones. Nothing other than the computer is plugged into an outlet. Once I hook up the amp there can of course be ground-loop issues, but right now I'm hearing noise through the headphones connected directly to the soundcard with no amp connected.
Haven't checked what happens when I unplug the monitor - I'll check later today.
A little more background - previously I had the computer connected to a TV via HDMI, with the TV return audio (toslink) driving a receiver. This setup produced no computer noise, however, the receiver has a number of functional and performance issues, so I plan to simplify things by just having computer, sound card, power amps. The setup works well with the laptop, so am hoping to resolve the desktop noise issue.
I don't see how it can be a ground loop when there is only one path - AC plug --> computer --> USB sound card --> headphones. Nothing other than the computer is plugged into an outlet. Once I hook up the amp there can of course be ground-loop issues, but right now I'm hearing noise through the headphones connected directly to the soundcard with no amp connected.
Haven't checked what happens when I unplug the monitor - I'll check later today.
A little more background - previously I had the computer connected to a TV via HDMI, with the TV return audio (toslink) driving a receiver. This setup produced no computer noise, however, the receiver has a number of functional and performance issues, so I plan to simplify things by just having computer, sound card, power amps. The setup works well with the laptop, so am hoping to resolve the desktop noise issue.
Sorry I missed that part in your description.
What happens if you hook the soundcard through a powered hub?
Didn't seem to make a difference whether or not the laptop (Dell Latitude E6410) was plugged in, which surprised me. It uses a grounded plug. When I use the laptop with a Behringer UAC222 to make amplifier measurements I unplug it to reduce measurement noise; the Asus Xonar U7 does not seem to have that issue - perhaps better filtering/isolation?
Don't have a powered hub, but maybe that would be a better option than the USB PCIe card I was contemplating. Think I'll do some loopback measurements using the Asus card on my laptop and desktop and see how they compare.
Don't have a powered hub, but maybe that would be a better option than the USB PCIe card I was contemplating. Think I'll do some loopback measurements using the Asus card on my laptop and desktop and see how they compare.
Put a few uF across the USB power supply at the DAC end.
I run USB scopes off USB and have 100uF on the 3v3 regulator.
I run USB scopes off USB and have 100uF on the 3v3 regulator.
It depends on the internal conception of your computers beccause it is a galvanic issue, i've tryed to put an AC mains isolation transformer on my desktop that was touched by this problem and the probelm was stil there, but it has disappeared only with a line transformer (Jensen isomax).
I never encountered this issue with a professional USB soundcard.
OK, made some more measurements and observations. Bottom line - it's a ground loop. The noise I was hearing when the amplifier (Class I, all grounds tied to chassis) was connected was computer noise, most noticeable when I had Task Manager open and I'd clearly hear the one second refresh burst of noise. When I went to headphones I also heard noise, but on more careful listening it is not computer generated hash, just typical low level noise. I hooked up a cheap Class II bluetooth amplifier (aux input) and no computer noise was heard, so I'll have to figure out how to resolve my vintage amplifier ground loop issue.
You could put a switch between vintage amp zero volts and ground.
That will break the ground loop.
That will break the ground loop.
So why tie the amp ground to the chassis? It would seem easy enough to lift it off (along with RCA input ground and speaker ground). Maybe hum would be worse? I've also seen a diode bridge used between safety ground and the chassis which eliminates the low level ground loop noise but gives protection in a wiring fault.
The audio should only be tied to one ground to stop ground loops.
If the PC is grounded it connects via USB bus ground to DAC to amplifier.
If amplifier zero volts is grounded then you have a ground loop.
If you disconnect amplifier zero volts from ground it should remove ground loop. However sometimes the amp is connected to ground through RCA sockets etc which messes up the exercise.
Yeah, have to figure out the easiest way to break the ground loop. Maybe time for a new amplifier....
Interestingly all consumer-level amps (i.e. with no balanced inputs) used to be class II. Even the very powerful ones. AV receivers (with single-ended inputs) are class II too, even with large 600W transformers.
Making a power device class II is expensive though. No using powerful low-cost switched power supplies (which would not comply with EMI/RFI certification), transformers tested for higher isolation, etc... I understand why amp manufacturers opt for class I instead, much cheaper for them.
Making a power device class II is expensive though. No using powerful low-cost switched power supplies (which would not comply with EMI/RFI certification), transformers tested for higher isolation, etc... I understand why amp manufacturers opt for class I instead, much cheaper for them.
My mid-70's stereo amp with a 1.5kVA transformer ties safety ground, power supply ground (split supply), and speaker output grounds to the chassis, along with the RCA input grounds through 1.2 ohm resistors. It would not be particularly hard to raise the supply, speaker, and RCA grounds off of the chassis. I'm pretty familiar with the wiring as I just rebuilt it, including mostly new wiring. Is that advisable, or is there an issue I'm not seeing? My backup plan is to learn to enjoy computer sounds in the background. 🙂