• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Cleaning USB for Bus Powered Audio Devices: Discuss

OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
Thanks for still following this.
It's my conclusion that I will have to isolate the USB somehow, and safely.
ADuM4160 based boards can be had cheap <£10, so I might give one of those a try, even though they are going tot op out at 24/96 or less.
If one works then I suppose the ifi Defender may be in order, or the HiFiMe 480mpbs offering - both are pricey relative to the D10.

I'll post back my findings when I try more things.
 

Jimmy

Active Member
Joined
Jun 20, 2018
Messages
255
Likes
167
I have very similar problems with two desktop PCs, all sort of noises can be heard when moving the mouse, varying cpu loads, in my case just plugging the DACs to a hub (non powered) fixes the problem (at least makes it inaudible). Most USB hubs do some kind of power cleaning, and expensive industrial ones usually are galvanically isolated.

Cheap galvanic isolators are only USB 1.0/1.1 compatible, and 2.0/3.0 ones are much more expensive.

If you have some experience with electronics you can just try putting a capacitor between VBUS and GND, but I you have a hub at hand I would try it before doing/buying anything else.

Thanks for still following this.
It's my conclusion that I will have to isolate the USB somehow, and safely.
ADuM4160 based boards can be had cheap <£10, so I might give one of those a try, even though they are going tot op out at 24/96 or less.
If one works then I suppose the ifi Defender may be in order, or the HiFiMe 480mpbs offering - both are pricey relative to the D10.

I'll post back my findings when I try more things.
 

mcgo

Member
Joined
Nov 29, 2018
Messages
19
Likes
21
Threads like these remind me why I ditched USB and now use galvanically isolated spdif coax exclusively. Much cleaner, no noise troubles, DSD via DoP compatible and no jitter. Simplify your life and lose the USB port!
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
I have a hub in my monitor - made no difference.
I don't have any other USB hub unfortunately.

Unfortunately, as for audio interfaces there are so few choices currently. Out of the choices out there, if one is seeking low latency (live FX in the box) and stable drivers on Windows, there isn't much that has a pretty solid reputation except RME.
USB seems to be a thing that is fine for battery powered laptops, but more of a nuisance in desktops. Really, I place the blame at motherboard manufacturers being cheap or lazy on their designs. A lot of huff has been made about isolated traces on the motherboard for the audio chip and it's components, but what about USB - oh no, just chuck that together folks only plug just about every peripheral you can think of in it - most of which are not susceptible to these noises.
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
@graz_lag So what you are saying is to take +5v and GND from a battery (or +5v PSU), then Data + and Data - from the computer player source?
At the moment I have a 6 inch Male-Female USB A cord, and I've cut the +5v, and wired in another USB Male A cable to that, along with cleaning the insulation from the GND and T-ing the GND.

Are you explaining this:
USB_split.PNG


I am thinking a split-power USB cable should be more like this:
USB_split2.PNG


I am not sure if it's such a great idea to run Data without GND or Shield.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
The data of a USB signal needs the ground it is designed this way.
It's not designed like Ethernet or RS485 connections.

A 47 Ohm resistor (lower drawing) may lower the common mode currents which pass through the USB ground and the audio RCA outs of the DAC. Maybe low enough that you aren't bothered by it, but maybe not enough.
It will only cost you 2 cheap USB cables and some soldering.
The lower circuit is worth a try for that reason.
 
Last edited:
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
I have made the lower setup (wrapping cables, insulating with sticky tape, no soldering), without resistor or cap, and it makes no difference. Using the (clean) RPi as a power source.
I should probably grab me a micro USB socket when I get a few resistors (and/or pot) and maybe caps. It'd mean I could just use the RPi PSU, rather than having to go through the RPi - I am such a bodger!
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
You have a balanced input on your monitors that is completely going to waste right now. Why not make use of it?

You have multiple options:
1. Ready to go but not the cheapest - 2x RCA to TS cable (length uncritical), 1x Behringer HD400, 2x TRS to XLR cable (keep <3 m or so). Works a treat. Cheap cables may need some inspection and resoldering (avoid molded-on connectors should you encounter those), but it sounds like you wouldn't be afraid of this. (Distortion performance of the HD400 around 1 Vrms with a 22R source impedance looked uncritical to me. I rarely run more than 30 dB below fullscale (2 Vrms) average into my monitors these days, so that's probably under half a volt peak. Used to have input gain turned pretty far down with little headroom, now it's arguably more than I need.)
2. Solder up your own RCA to XLR adapter cables. Required per cable: 1x male XLR connector, shielded TP cabling, 1x RCA connector, 1x 100 ohm resistor, heatshrink tubing or similar insulating material. Connect as follows:
RCA signal - hot (usually red cable) - XLR pin 2
RCA ground - 100R - cold (usually black) - XLR pin 3
RCA ground - screen - XLR pin 1 (may need to be left open at XLR end)
Note that the 100 ohm resistor is chosen to match the D10's nominal output impedance. If the output is AC-coupled, further refinement may be possible.
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
Update:
My scenario testing gets more interesting!

(1) [Mains Earthed] Desktop PC > USB DAC > Netbook ADC [No Mains Earth]
No noise.

(2) [No Mains Earth] Netbook / RPi > USB DAC > Desktop PC ADC [Mains Earthed]
No noise.

(3) Any Computer Device > USB DAC > Desktop Speakers [transformer with 3 pins, no Earth?)
No noise that I can hear, even at maximum volume.

(4) [Mains Earthed] Desktop PC > USB DAC > Desktop PC ADC [Mains Earthed]
Noise!

(5) [No Mains Earth] Netbook > USB DAC > Netbook ADC [No Mains Earth]
Noise!
Sounds similar but a little cleaner, fewer harmonics and less amplitude.

(6) [Mains Earthed] Desktop PC > USB DAC > Active Near-field Monitor(s) [Mains Earthed]
Noise!

So, what conclusion I am making from all this, is that if there is no ground loop then all is well, but if there is, then we have a problem. The ground loop can come from two devices being connected to the same Mains Earth (PC and Near-field Monitors) or via sharing the same linked internal shield (USB / ADC).

Thanks for all your help, ideas and encouragement folks.
I now know that I am not trying to "clean up USB bus power" as the topic title suggests, rather, I am trying to cure a ground loop, as some of sharp minded people have actually suggested.

Galvanic isolation will do this, so an ADuM4160 device supporting 24/96 max. will solve this, but may or may not introduce a pinch of extra noise.
The ifi Defender 3.0 might work too, using an external PSU. This looks to do hand-shaking first then cut the ground to the USB host, but I am not sure.
HiFiMeDIY offer two products, one being 24/96 max. the other more expensive and handling the higher sample rates.
I'll leave off listing more products as there is another thread doing exactly that.
Of course, isolating the analogue connection would also serve the same purpose, so a Behringer HD400 should work just the same.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
So, what conclusion I am making from all this, is that if there is no ground loop then all is well, but if there is, then we have a problem. The ground loop can come from two devices being connected to the same Mains Earth (PC and Near-field Monitors) or via sharing the same linked internal shield (USB / ADC).

Thanks for all your help, ideas and encouragement folks.
I now know that I am not trying to "clean up USB bus power" as the topic title suggests, rather, I am trying to cure a ground loop, as some of sharp minded people have actually suggested.

Galvanic isolation will do this, so an ADuM4160 device supporting 24/96 max. will solve this, but may or may not introduce a pinch of extra noise.
The ifi Defender 3.0 might work too, using an external PSU. This looks to do hand-shaking first then cut the ground to the USB host, but I am not sure.
HiFiMeDIY offer two products, one being 24/96 max. the other more expensive and handling the higher sample rates.
I'll leave off listing more products as there is another thread doing exactly that.
Of course, isolating the analogue connection would also serve the same purpose, so a Behringer HD400 should work just the same.

Yes, would be my conclusion as well.

The Behringer might work well here AND you could use the balanced in from the speakers.
I suspect there are 2 line transformers in there.
But .. every upside has a downside.
These are bandwidth limited and are not specified anywhere. Both on the low and high frequencies.
There are no specs other than 500 Ohm load impedance and they should probably be loaded with 600 Ohm as well.
Chances are the D10 sees a load that may be a bit too low and chances are when loaded with a 10k load there may be a bit of an upper treble boost and maybe in the lows as well.
But it is cheap and an interesting device... may buy one myself ...

When we are talking maximum fidelity I would go for a proper USB isolator but ... these too may have some leakage via its own power supply.

decisions...
 

graz_lag

Major Contributor
Joined
Oct 13, 2018
Messages
1,296
Likes
1,583
Location
Le Mans, France
I have been listening to hi-res FLACs for many years now, thru different setups and headphones as well.

I can pick the difference between a 16 and a 24-bit, but incapable of distinguishing a 96 over a 192 file, so the limiting factor to 24/96 for your USB isolator would not be a limit at all for me ...
It looks to be the best compromised solution, not least considering the cost ...
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
Any DSD wouldn't work over a 12mbps connection, would it?

Sure, I have lived with 24/96 for ages, just re-sampled any material that was higher, on-the-fly. Never felt I had issues, was quite happy using a PCI audio interface for that.

I still want to solve my setup, because I don't think it's necessarily a good idea to run it like this all of the time. That said, I just shoved my ear right next to the speaker driver and I have to turn the gain up more than where I have it set to actually hear the noise from the DAC over the amps own low hum and hiss. So really, I will never hear the noise in my normal use, and is probably why I didn't even notice right away!

I wonder how the ifi iDefender works?
It allows a fast data exchange, so it's not using the usual isolator chips. It has sensing to 'cut' the ground and power from the host, if "other" ground(s) are found. I am thinking it keeps the Data lines intact, but shoved a couple of diodes or something on the GND and VBus from the host, so it effectively turns the connection into Data +/- only. It must allow the enumeration to occur firstly, because a straight up green/white connection only (power form external PSU) doesn't do anything at all - OS doesn't 'see' the device, device is still in standby mode, then turns off.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
index.php


This is basically the ifi defender.
The ground is 'broken' here, which is what USB should have, but reckon when there is a ground loop the 'common' of the PC is still there but running through the mains.
Problem that might be there is if there is a voltage present between the 2 'grounds' (which makes the current loop flow) that exceeds that of the USB input of the DAC and is not 'protected' in the DAC it might destroy the DAC.

Perhaps measure the DC resistance (while everything is connected but the USB is disconnected) between the USB screen of the PC and the USB screen of the D10 input. If that reads close to 0 (or perhaps 100 Ohm or so) it may be safe to use the circuit above and save money on buying an ifi ?
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
I'll get the multimeter out in a while.
I already tried the cable above splitting D+/D- and +5v/GND to two different locations, and it did not work for me.
The D10 turned on, but showed the dashes on display, eventually going into standby mode (single dot?), there was no recognition in the PC.

What about making that same cable, but shorting +5v and GND on the PC (data source) with a resistor? Would this 'trick'the PC into thinking a device is present, or would it just result in potential damage?

Ah, now this looks interesting.
https://www.bamlog.com/usbisolator.htm
In particular, the author states that the device may still need to be left un-isolated on connection, so that it can enumerate/initialise. Once that is done, the toggle switch can then isolate the power to use external.

I have demonstrated that the red +5v line to the Host device is not required when the DAC is powered with external PSU. So, really, all I would need to do is to just break the USB ground to Host _after_ the DAC is initialised, basically when I see a sample rate on the D10, then the ground loop (and buzzings) should just go away.

USB_split3.PNG

Not sure what values to use, and whether or not the cap. is in the right place, or required at all.

Another version, same thing more or less.
http://www.madscientist-audio.com/gbu.html

However, is that a good idea? USB is designed with different lengths on the tracks of the connector. The Data lines are the last things to connect, after ground and then power.

I am now heading more towards the original topic title, even if I am a little skewed (not clean power, rather clean ground/brake ground loop).
http://zone.ni.com/reference/en-XX/help/370592AB-01/digitizers/513x_ground_loop_noise/

...and for those who missed it:
Silanna ICE08USB is the USB2.0 (up to 480mpbs) alternative to the ADuM4160 USB1.1 (up to 12mbps) galvanic icolator.
It's seen in the UpTone ISO Regen (which did nothing in Amir's test - but he didn't have a ground loop) and also CoolGear's Isolator.
 
Last edited:
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
@graz_lag I don't know.
I was just saying that it's the only other chip solution around, that I could find.
It might be rather expensive, as a chip, and so hasn't filtered into cheaper implementations as of yet.
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
I just tried the above cable layout, without the extra parts.
That works.
Once the D10 screen shows a sample rate, cut the GND to the PC and it'll happily play music and switch sample rates. No buzz.
At least I know the Data lines aren't affected by anything funky.

So, for the cost of a USB A-B cable, a USB A cable, a DC power input, a switch and maybe a resistor and capacitor and a nice little box - we're talking less than £10 inc. shipping really. I can certainly accept that cost.

However, I don't know if it's particularly safe long-term to do this. What about voltage bumps or other strange things I have no idea about?
I wonder how tough it might be to have a chip auto-cut the ground or both VBus and ground when enumeration is complete?
 

graz_lag

Major Contributor
Joined
Oct 13, 2018
Messages
1,296
Likes
1,583
Location
Le Mans, France
@L5730 nice, I am going to DIY this implementation on my basic layout, which in fact equals the iFi defender as pointed out by @solderdude ...
I like a lot this thread, you go very depth into a very specific issue, which many folks try to solve with some fancy and expensive-considering-the-application devices ...
Thank you so much ! ;)
 
OP
L5730

L5730

Addicted to Fun and Learning
Joined
Oct 6, 2018
Messages
669
Likes
435
Location
East of England
Please be clearly warned to the fact that I have very, very limited experience in electronics - which could also be easily be read as zero experience.
I am spending many hours searching the 'net for terms that might get me something useful, and am trying to put together something from that.

I don't know if this diagram makes it clearer for people.
Close the switch and connect to PC - device initialises - then open the switch and kill the ground loop.
1546878688425.png
 
Top Bottom