Filter USB power?

[Noozer]

I have a set of LTB headphones. These headphones are designed for use with a
PC. They have an amplifier that is powered from the USB cable. Unfortunately
for me, my DFI mainboard seems to have really dirty power on the USB ports,
which causes buzzing and clicking. A fair number of folks have this problem
with the headphones and the official fix is to "use a powered USB hub"
instead of connecting directly to the PC's USB port.

I'm hoping that I can clean the power up by simply adding a couple
capacitors to the amplifier (or even USB ports on the mainboard).

Since this isn't noise at any specific frequency, I'm not sure how I can
choose the most effective values for the capacitors.

I'm also wondering if a ferrite ring would help here?

 

[Noway2]

I have a set of LTB headphones. These headphones are designed for use with a
PC. They have an amplifier that is powered from the USB cable. Unfortunately
for me, my DFI mainboard seems to have really dirty power on the USB ports,
which causes buzzing and clicking. A fair number of folks have this problem
with the headphones and the official fix is to "use a powered USB hub"
instead of connecting directly to the PC's USB port.

I'm hoping that I can clean the power up by simply adding a couple
capacitors to the amplifier (or even USB ports on the mainboard).

Since this isn't noise at any specific frequency, I'm not sure how I can
choose the most effective values for the capacitors.

I'm also wondering if a ferrite ring would help here?

Normally, the USB power lines already have inline ferrites and bypass
capaitors that are designed to filter the power lines. The USB
specification calls out for allowable voltage drops, transients, etc
but perhaps it allows more than your application can tollerate.

I am a little confused, though. You make it sound like the headphones
are receiving power from the USB, but that they don't get the data from
the USB port. If this is the case, this may be part of the problem.
If the data is coming from the USB port, it would be digitally
transmitted and this should go along way towards prevent noise
contamination.

You may be able to place a small valued capacitor across the +5 and GND
lines from the USB port. This would shunt AC components to ground. It
is hard to say what a good value would be as it depends on the rest of
the circuit that is hooked up and its input impedance. I would suggest
that you can experiment, trying different values ranging from 47uF to
..01uF and see what happens.

 

[Simon]

Normally, the USB power lines already have inline ferrites and bypass
capaitors that are designed to filter the power lines. The USB
specification calls out for allowable voltage drops, transients, etc
but perhaps it allows more than your application can tollerate.

I guess they try to save money wherever they can. I've got the USB cable
wrapped around a ferrite ring a few times and most of the noise is now gone.

I am a little confused, though. You make it sound like the headphones
are receiving power from the USB, but that they don't get the data from
the USB port.

That's correct. They just take the analogue audio from the soundcard and
pull power from the USB port to power the amplifier.

If this is the case, this may be part of the problem.
If the data is coming from the USB port, it would be digitally
transmitted and this should go along way towards prevent noise
contamination.

True... I considering going that way, but I wanted the headphones to be
portable - no drivers needed.

You may be able to place a small valued capacitor across the +5 and GND
lines from the USB port. This would shunt AC components to ground. It
is hard to say what a good value would be as it depends on the rest of
the circuit that is hooked up and its input impedance. I would suggest
that you can experiment, trying different values ranging from 47uF to
.01uF and see what happens.

The ferrite ring seems to have done the trick. If I notice any change I'll
pop the amp open and add some low value caps.