USB isolation between Computer and DAC

[ravi_d]

I'm using the SB Touch connected to the an Arcam rDAC for sometime as an alternative to CD playback. I was thinking of eliminating the SB Touch from the chain as the computer needs to be running all the time whether I use the SB touch or the DAC alone.

I bought a cheap A-B type USB cable on ebay and the moment I connected the USB cable between the computer and the DAC I have been greeted with a humming noise from the speakers :sad:
First I suspected the noise issue is with the DAC. But on quick inspection the humming noise from the speakers have been tracked to my laptop's power supply adapter which I replaced sometime ago. When I remove the PSU adapter the humming noise is going away and the moment its connected back the humming noise is promptly appearing again and my 3 year old laptop battery doesn't last for more than 15 mins to leave the computer running on the battery

I read somewhere that a cheap/decent USB cable with USB isolator works better than a costly/boutique USB cable alone. I'm sure fixing the laptop psu will solve the issue but I have no clue where to start. And buying a new battery may not be a feasible option as my 17" laptop when new was lasting for less than 2 hours and my laptop is due for replacement sometime soon and don't want to invest anything on it. So the other alternative is to use some kind of isolation between the computer and the DAC.
I ordered the Olimex USB isolator on ebay UK hoping it would fix the humming noise from the laptop PSU adapter.
The Olimex isolator has a small footprint and requires two A-B USB cables to connect between the computer and DAC. It has a DC power supply input but in most cases it doesn't require an external power supply. On connecting the USB isolator, to my surprise the USB isolatior worked like a charm in eliminating the humming noise from the computer psu adapter :yahoo::clapping: The ground loops between equipment being the enemy in Hi-Fi set-up's, the USB isolator works on breaking the ground loops between the computer and Hi-Fi providing galvanic isolation.

Computer SMPS/ laptop psu adapter may not be designed to the audio/ hi-fi standards. I think its a good practice to have some kind of isolation to get the best performance from the USB DAC. In fact I would say a decent USB cable with an isolation works better than a costly USB cable. At least it worked in my case. You don't need such an isolator if your USB DAC or USB to SPDIF converter implements some kind of isolation on its USB input.

I paid about 37 including shipping with in UK and I'm not sure if it would be available for same price again due to gaining popularity in hi-fi applications.. I remember this item when introduced the initial versions were sold for about 20. This is a generic USB isolator which can be used in many applications and hi-fi use between the DAC and computer is one of them.
Here is a link to Element 14 India website for the same product.

Here are some pics of my USB cable with isolator.

Mods: Feel free to move the thread to relevant section if necessary

 

[Thad E Ginathom]

Ground loops and other sources of mains hum: here's an interesting article on Audioholics

 

[kaushik]

This tiny guy is well made and many audio people have used and found excellent results in USB dac sound Q:USB Isolator, Galvanic Isolated USB 2.0 - Electronics-shop.dk

 

[indialogue]

I paid about 37 including shipping with in UK and I'm not sure if it would be available for same price again due to gaining popularity in hi-fi applications.. I remember this item when introduced the initial versions were sold for about 20. This is a generic USB isolator which can be used in many applications and hi-fi use between the DAC and computer is one of them.
Here is a link to Element 14 India website for the same product.

Thanks for sharing details of this product - looks interesting. However, the Element14 guys must be off their rocker - 6000INR+!!! Talk about daylight robbery

 

[ravi_d]

Thanks for sharing details of this product - looks interesting. However, the Element14 guys must be off their rocker - 6000INR+!!! Talk about daylight robbery

Well technology has advanced and many are designing USB modules with isolator built into it to separate the noisy computer environment from the clean hi-fi environment. I just ordered such USB module from diyaudio group buy with features such as isolation between computer and dac, asynchronous usb transfer up to 32 bit 384 kHz and DSD over pcm (dop), option to power the noisy computer part of module and clean signal part after isolation with separate power supplies etc. all for that much hyeah:
Then why buy just the isolator for 6k then

 

[indialogue]

Well technology has advanced and many are designing USB modules with isolator built into it to separate the noisy computer environment from the clean hi-fi environment. I just ordered such USB module from diyaudio group buy with features such as isolation between computer and dac, asynchronous usb transfer up to 32 bit 384 kHz and DSD over pcm (dop), option to power the noisy computer part of module and clean signal part after isolation with separate power supplies etc. all for that much hyeah:
Then why buy just the isolator for 6k then

Very nice, Can you share details of the diyaudio group buy please?

 

[ravi_d]

Very nice, Can you share details of the diyaudio group buy please?

PM sent with details.

 

[anandkrishnantc]

I wonder how noise will get transferred from PC to a dac thru USB?
I understand its only digital signal passing thru this cable to DAC. Since PC bipasses the onboard sound card and the sound digital information (only high and low voltage patterns) to be sent to the USB DAC which the machine detects.

Where is the noise signal comes into picture here? I think there should be a signal verification at the receiving /sending end in the case of digital signal transmission.
Any decent quality USB cable should be enough to have USB dac to work.

 

[IndianEars]

kaushik

A USB Opto-isolator is ideal, and what you have ordered seems like an opto isolator.

The Isolated USB on some of the better DACs is usually a Transformer based isolator.

If your DAC does not draw power from the Computer, then you could try the following DIY :

1. Carefully remove the outer sleeve for about 1 inch of the USB Cable, anywhere on the cable.

2. Cut-off the red wire. This is the + Power Supply on the 4 wire USB Cable.

If this still does not cure your problem, cut off the Ground wire ( Usually Balck) in the 4 wire USB Bumch.

The other 2 wiresd carry the USB Data.... Dont Cut Them !

My Revelation USB Cable ( Yup, an Audiophile USB Cable ) does this trick, and says that it will not carry USB power so cannot be used where the DAC must be powered by the Computer USB Port.

Let me know the results if you do try this. ( At Very Worst, you will have destroyed a USB A to B cable )

 

[Thad E Ginathom]

I wonder how noise will get transferred from PC to a dac thru USB? ...

I am a great believer in the audiophile potential to find problems where none exist but this would not be at all impossible, and I'd suggest...

--- yes, it is a digital signal? Is it not possible that something other than the digital signals could be passed on the signal lines? People talk of noise being transmitted even on coax S/PDIF cables.

--- there are the power transmission lines, as discussed in the previous post.

These may or may not be of any consequence. Do the designers of DAC USB interfaces not take this into consideration? Whilst I don't know the answer to that, it might be like internal sound cards, where audiophiles claim they cannot be used because of the noise inside a PC cabinet, whereas any decent sound card has had all that taken into account in its design. And with a good power supply, there is probably a lot less noise than people like to imagine.

All that said, ravi_d had an undeniable hum problem (could it be an earth loop thing?) and it was undeniably fixed by the simple expedient of the USB isolator. Fascinating ...and one for all of us to remember should it ever happen to us. Thanks ravi_d

 

[mowgli80]

@anandkrishnantc - This is my understanding and may be wrong. If the DAC is synchronous and depends on computer clock and has no clocking infrastructure on the receiver side, the noise can lead to retransmissions and affect the timing. As such. when the PSU is not there, the cable works just fine without the noise and retransmissions and the isolator.

@IndianEars - If the DAC is not drawing power, the red power wire will not carry any power and is just a bare wire. Am not sure how just cutting it will help. Can you point me to the source where you found this information.

 

[reignofchaos]

One thing to be careful about in these usb isolators is that those based on the ADUM4160 will only support USB full speed (aka 12mbps) and not high speed (aka 480mbps). Many custom 24/192 usb receivers such as the m2tech hiface can't work properly in usb full speed mode.

 

[ravi_d]

@anandkrishnantc - This is my understanding and may be wrong. If the DAC is synchronous and depends on computer clock and has no clocking infrastructure on the receiver side, the noise can lead to retransmissions and affect the timing. As such. when the PSU is not there, the cable works just fine without the noise and retransmissions and the isolator.

It was indeed a ground loop issue I faced as explained in the article posted by Thad in post #2. My dac was having an asynchronous USB input too. Here is the explanation from that article

...a system grounded at two different points, with a potential difference between the two grounds can cause unwanted noise voltage in the circuit paths. Currents flow through these multiple paths and develop voltages which can cause damage, noise or 50Hz/60Hz hum in audio or video equipment.
...

@IndianEars - If the DAC is not drawing power, the red power wire will not carry any power and is just a bare wire. Am not sure how just cutting it will help. Can you point me to the source where you found this information.

Wouldn't the return ground wire running between the device and the computer joining both the ground planes which may be at different potential as mentioned in that article.

One thing to be careful about in these usb isolators is that those based on the ADUM4160 will only support USB full speed (aka 12mbps) and not high speed (aka 480mbps). Many custom 24/192 usb receivers such as the m2tech hiface can't work properly in usb full speed mode.

I'm playing a 24bit 96 kHz flac tracks which were running at some 27xx kbps bit rate. A 192 kHz track might require less than 6 Mbps to play.
Wouldn't ADUM4160 full speed 12mbps enough for our requirement?

 

[reignofchaos]

It is not a question of rate, it is a question of compatibility of the receiver chips. I tried one of these and my m2tech hiface wouldn't get detected. However my audinst mx1 which is based on an adaptive TE7022 chip got detected just fine.

I believe many xmos based receivers will have the same issue.

 

[mowgli80]

It was indeed a ground loop issue I faced as explained in the article posted by Thad in post #2. My dac was having an asynchronous USB input too. Here is the explanation from that article

Agreed. Ground loop is the cause. I was trying to guess its (the noise) effect on a digital(usb) signal based on anandkrishnantc's query. Another guess would be that the analog section of the DAC may be sharing the ground plane with digital section and may be the one that is actually getting affected. My electrical/electronics knowledge is very rudimentary so if somebody with actual knowledge can explain it, it would be nice.

Wouldn't the return ground wire running between the device and the computer joining both the ground planes which may be at different potential as mentioned in that article.

Agreed. But that does not explain why cut the power wire. Also, will a setup not having ground loop issues benefit from cutting those wires? Just want to know the theory behind it before implementing in my DIY effort .

 

[doors666]

have u tried to plugin the psu to a different socket. anything else connected to the laptop? try the usb cables with a ferrite bead.

 

[ravi_d]

Agreed. Ground loop is the cause. I was trying to guess its (the noise) effect on a digital(usb) signal based on anandkrishnantc's query. Another guess would be that the analog section of the DAC may be sharing the ground plane with digital section and may be the one that is actually getting affected. My electrical/electronics knowledge is very rudimentary so if somebody with actual knowledge can explain it, it would be nice.

Even I'm not an expert here but I did some googling long ago before you
From what I understood there is no harm in connecting the digital and analogue ground planes at a single point depending on the grounding topology used as the return paths always follow the the path having least inductance/ impedance (usually the path on the ground plane just below the signal trace) and the analog and digital return paths don't cross each other.

Agreed. But that does not explain why cut the power wire. Also, will a setup not having ground loop issues benefit from cutting those wires? Just want to know the theory behind it before implementing in my DIY effort .

Don't cut any wires. If its working well leave it as it is. The main idea of an isolator is to separate the two devices physically yet maintain the connection electrically. So cutting the voltage wire and ground wire doesn't physically isolate the connection.

I had that nasty ground loop issue when using my old laptop with aftermarket power adapter. I never faced that issue later with the rest of the gear being same as I moved out to a new place and changed the laptop at same time.
Now just using that device as a joint to extend the USB cable when needed :lol:

 

[Thad E Ginathom]

try the usb cables with a ferrite bead.

I have no idea about this, and am repeating it in case of interest, but on the USB site (will put link if I find it) there is some mention that ferrite beads may actually be bad for USB sound.

Link: USB FAQ. search on that page for 'ferrite'

As I say, I have not a clue why. If you can read and see if you have any ideas...

PS... it actually says something like "degrades the signal." I think it is not audio-specific, but if the digital signal is degraded in any way, it can't be much good for audio.



.

 

[IndianEars]

PS... it actually says something like "degrades the signal." I think it is not audio-specific, but if the digital signal is degraded in any way, it can't be much good for audio.

Oh ! I thought that Digital Guaranteed Perfection, always .....

Since cables cant influence Digital Transmission / sound, how can a ferrite bead, sitting on a cable.

Its all Purrrrrr fect ... No ?

 

[Thad E Ginathom]

Oh ! I thought that Digital Guaranteed Perfection,

hee hee... I don't know. Certainly I never said that!