USB power headphone Amp has noise

Lord Banshee

Golden Member
Sep 8, 2004
1,495
0
0
Ok i built a headphone amp based off the:
http://tangentsoft.net/audio/cmoy-tutorial/

But i wanted to power the device with USB power as it is for my laptop and i hate buying batteries. Well it seems the USB is causing lots of noise (no music player). If i hook the same circuit but instead of USB power i use a DC adapter or 9V battery i get almost no noise so this is why i know it is the USB.

So do you guys know of of any good way to stop this USB noise issue or am i out of luck as it is the computer which has the issue?

Thanks
 

herm0016

Diamond Member
Feb 26, 2005
8,512
1,128
126
usb is only 5 volts and is non bias. depending on your op amp you will saturate it very quickly with just 5 volts, creating a lot of noise. try using the voltage from 2 usb ports, 10 volts and using a large capacitor between your voltage input leads. also match the resistor on the output to the impedience of your headphones to make the amp run the most efficiently.
 

Born2bwire

Diamond Member
Oct 28, 2005
9,840
6
71
Originally posted by: herm0016
usb is only 5 volts and is non bias. depending on your op amp you will saturate it very quickly with just 5 volts, creating a lot of noise. try using the voltage from 2 usb ports, 10 volts and using a large capacitor between your voltage input leads. also match the resistor on the output to the impedience of your headphones to make the amp run the most efficiently.

I was just going to say this too. I would be very concerned about clipping and dirty power using the USB 5 V. One thing you may also want to do is replace the virtual ground with a buffered ground using a BUF634 or TLE2426. Probably want to use the TI rail splitter as the BUF634 is kinda expensive.
 

Lord Banshee

Golden Member
Sep 8, 2004
1,495
0
0
ok well let me explain my circuit:

The input 5V is used as an input to a TI BB dcp020515 which creates +15V and -15V output. The capacitors are set up as in spec sheet. Both the +15V and -15V are connecting with 100uF capacitors, this +15 and -15 is the +/- source for the opamp.

But again i have used a 5V DC plug and the noise isn't there.

So my ground is a real ground and my opamp is getting +/-15V

Does this help to help me try to solve the issue?
*edit* 20515 not 21515
 

Born2bwire

Diamond Member
Oct 28, 2005
9,840
6
71
Originally posted by: Lord Banshee
ok well let me explain my circuit:

The input 5V is used as an input to a TI BB dcp021515 which creates +15V and -15V output. The capacitors are set up as in spec sheet. Both the +15V and -15V are connecting with 100uF capacitors, this +15 and -15 is the +/- source for the opamp.

But again i have used a 5V DC plug and the noise isn't there.

So my ground is a real ground and my opamp is getting +/-15V

Does this help to help me try to solve the issue?

Correct me if I'm wrong, but it looks like TI says that the minimum input voltage for the 021515 is 13.5 V.

Text

EDIT: Do you mean the 020515?
 

Lord Banshee

Golden Member
Sep 8, 2004
1,495
0
0
yes sorry 20515.

my DMM shows the output underload to be +/- 15 volts.

I tried a ferrite beads usb cable from a USB mouse still can not touch the battery cleanness.
 

Born2bwire

Diamond Member
Oct 28, 2005
9,840
6
71
Originally posted by: Lord Banshee
yes sorry 20515.

my DMM shows the output underload to be +/- 15 volts.

I tried a ferrite beads usb cable from a USB mouse still can not touch the battery cleanness.

Maybe the USB does not have the necessary currents needed by your regulator. I looked at the wiki for USB (very authorative I know) and they mentioned that the USB standard give a current from 100-500 mA, but in 100 mA increments. Maybe that could be your problem, feeding steady currents. What you probably should do is sit down with your voltage regulator and monitor it on a scope. See what your ripple is from it with different sources under varying loads. I wonder if placing a unity gain buffer in between the USB and voltage regulator will help. The buffer will help source a steady current and isolate the load characteristics of the regulator from the USB.
 

Mark R

Diamond Member
Oct 9, 1999
8,513
16
81
So the ferrite beads helped? Did you put them on the supply, or just use a cable with beads?

What happens if you have the USB connected (i.e. provide USB ground but disconnect USB 5V) but use battery supply for the 5V? If you get bad noise then it could indicate some form of coupling between the audio and USB grounds, and you'll need to use some form of isolating transformer in the signal path. Not ideal.

However, what is more likely is that you're just getting filthy power from the USB, which is cross modulating with the switching regulator. If that's the case then you need to filter the incoming supply.

You need to build a filter for the incoming 5V line with series ferrite beads and parallel capacitors. So the 5V line would pass through a bead, then into a 10 uF cap to GND, then another bead, etc. You'll probably need at least 3 stages, but more can be added with larger caps, as required.
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
Why is everyone recommending ferrite beads?
Don't get me wrong, beads are good if you want to get rid of high-frequency noise (MHz); but that is rarely a problem in amps.
Since the problem IS likely to be "dirty" power from the USB port, you DO probably need inductors, but in this case this case means chokes (a few uH); not beads. Perhaps a pi-filter would help?

As far as I know, drawing "high-quality power" from an USB port is pretty tricky due to the VERY limited amount of current you can draw, as far as I remember there is also a limit to how much parallell capacitance you can use (i.e. you can't simply add more caps). You need a pretty sophisticated PSU.


edit: The "search" function over at www.diyaudio.com came up with this link

http://www.ecp.cc/images/HPDAC_0.1b_sch.GIF

also, look in some of the threads over there for more information about the power regulation problem.











 

Lord Banshee

Golden Member
Sep 8, 2004
1,495
0
0
Thanks guys

For the time being i am going to drop the idea of the USB power and leave it with the 9V battery as i got two lab reports to do this week, i hate lab reports lol.

But to answer the questions from above:
With the battery i measure the current with my DMM and it was around 130mA, for some reason when i tried to measure it when using USB it did not read(could be due to my sucky DMM). I don't have an oscilloscope (wish i did) so my testing is pretty limited to my ears and my DMM.

The ferrite bead i only used a USB cable that already had one installed (thanks microsoft). and it help a little and i mean a little and it most likely didn't at all and it was in my head thinking i wanted it to help.

i am interested in the inductor idea as stated above, how exactly is this going to help? The have really used inductors in any of my classes and the most i know of them is that they don't allow instantaneous current change. What other feature do they do?

I wish my college Analog classes were like the Digital ones. i think i could figure out and design many different digital designs but when it comes to analog components i feel lost sometimes. I just wanted to let this steam out lol :)

Anyway thanks for the link f95toli, when i have some time i'll check the forum out.

If anyone else has any other ideas i am willing to listen.

Thanks,
Chris
 

Mark R

Diamond Member
Oct 9, 1999
8,513
16
81
Originally posted by: Lord Banshee
i am interested in the inductor idea as stated above, how exactly is this going to help? The have really used inductors in any of my classes and the most i know of them is that they don't allow instantaneous current change. What other feature do they do?

They work by resisting current change. The usual situation in a filter is to feed a capacitor with an inductor.

If there is a change in voltage, the cap will attempt to charge to that new voltage - however, the sudden change in current is resisted by the inductor - as a result the voltage on the cap changes slowly. So repetitive fluctuations on the input get damped by an inductor/capacitor filter.

The 'cutoff frequency' above which damping occurs depends on the values of the inductor and capacitor. The higher the frequency above cutoff, the more it is damped. Multiple stages increases damping without changing frequency.

Ferrite beads are inductors - however, f95toli is right - they are way too low in value for this use.

search for LC filter for more info.
 

f95toli

Golden Member
Nov 21, 2002
1,547
0
0
The inductors is one the "basic" components: Resistors, capcaitors and inductors (RCL).
The impedance of an inductor is directly proportional to the frequency: Z=2*pi*f*L. Hence, inductors tend to "block" high frequency noise.

Essentially, ALL passive filters are implemented using combinations of R, C and L in various topologies.
A pi-filter is special type of low-pass filter which looks like this

in ------L------ out
_____|___|
_____C__C
_____|___|
____GND_GND

It is quite common in PSUs.


Edit: If you want to play around with various filters you can download PSU designer II for free from www.duncanamps.com

edit again: Used underscore instead of blankspace in my "drawing" above...
 

Lord Banshee

Golden Member
Sep 8, 2004
1,495
0
0
thanks Mark and f95toli i have a couple inductors laying around so any idea what a good value to start with would be?

Also thanks for the link f95toli
 

MrDudeMan

Lifer
Jan 15, 2001
15,069
94
91
i would try anywhere between 1 and 20uH then work from there. you would really need to do some calculations if you wanted to be accurate, not just throw in some random values, but you may find some that work.
 

cheesehead

Lifer
Aug 11, 2000
10,079
0
0
PSU inductors can do some bad things to op-amps. They're finnecky that way.

From experience, I can tell you that USB has a vast amount of "crud" on it. It shares a rail with both a few system components, and everything you have plugged in through USB, so it's going to get lots and lots of fluctuation.

There's a reason that people use batteries, you know.
 

Lord Banshee

Golden Member
Sep 8, 2004
1,495
0
0
lol yeah i gave up the idea of using the USB and just use a 9V battery. It came in handle the other day watching a movie from my laptop :)