USB-Cable - can it make a difference?

[orson]

Don´t be afraid - this should not be a theoretical discussion with flamethrowing and popcorn-eating-amusement.
Instead I want to encourage you to make this inexpensive test and let your ears decide.
As I am not sure whether wires make a difference in digital transport cable I am quite sure that dirt and EMF in the wire can lead to non-optimal signal transmission.
That said I wanted to see if it´s worth to take away the power wire in the USB cable.
Of course that makes only sense if your DAC does´t need the power over USB. That depends on the controller. Having a DAC with its own power supply is no guarantee that it works without the USB power. My Audinst for instance does not, all Audio-GDs do.
So grab yourself an USB cable (should be 2.0 certified) that you don´t need anymore. Remove the insulation near the source plug. Carefully loose the braided shielding and separate the red wire that carries the +5 Volts. Cut it and put a switch in.
That´s it!
Now you are in the comfortable situation to make a real A/ B comparison by listening and switching back and forth.
The result is not predictable as it depends mainly on your source computer and how much dirt it infuses into the power wire.
In my case I noticed a wider soundstage without the power wire. I did concentrate on a trumpet of a big band that was placed on the left side. Without the power wire it´s place was about 2m further away and a bit more reflections of the surrounding room were audible. Note: the effect took about 1-2 seconds to show up when switching off but nearly instant when switching on again.
It´s an easy project that costs nearly nothing and could give you a slight improvement or the certainty that there is no effect.

 

[butchbass]

Orson, I like the way you think. Thanks for posting this.

 

[Grainger49]

First, thank you for the thread.  I'm eagerly and patiently waiting for my Bottlehead DAC.  I'm not much on computers but know from previous experience that the wire can make a difference.  It doesn't always so it is a crap shoot unless you try it in your system or get a recommendation from someone who you trust.

This looks like a great test I can do with a really cheap cable I have coming.

What I do remember from my Waveguides & Transmission Lines course oh so many years ago is that the termination is critical.  Both the plug  and jack must be correct.  Otherwise there are reflections happen back toward the other end of the cable. 

This goes for the source and destination terminations.  The data for high res audio is high enough to have an effect on it.  And, again, since I'm not much of a computer guy I haven't seen any discussion of this.

 

[mcandmar]

I found the 5v rail on my computers USB bus to be incredibly noisy so i built a small box with an ADUM4160 isolator powered by a low noise TI TPS7A4700 regulator for my USB powered DACs.  The difference it makes varies from one DAC to another but there is a small audible and measurable difference in my opinion.

The RMAA trace below shows a PupDAC (USB Powered) being fed direct from the PC, vs an isolator with a LM317 power supply (Beezar Doodlebug).

For the BH DAC the isolator has to go as its only USB 1.1, but i am going to measure it with the direct USB supply, vs my own USB power supply and see if it makes any difference at all.  I suspect it wont from what John has said so far, he seems to have gone to great lengths to isolate the input side of things.

 

[Grainger49]

From what I understand the Bottlehead DAC is USB 2.0.  And it doesn't use the USB supplied power supply.

 

[Doc B]

I just want to remind everyone that this product is not a kit and we can only warrantee use with the stock power supply and a standard USB 2.0 cable. We have no idea if other power supplies and/or powered cables are safe to use with the DAC. This isolator might be OK, but I can't say it is safe to use because I haven't tried it. 

I emphasize this only because the repair price of a Bottlehead DAC that gets zapped from someone experimenting with non-stock powered connections is going to be $1550, if you get my drift. You can't recover from an experiment gone bad by just replacing a blown LED or diode like you might in our kits. We have built in what we can to protect the circuit from blunders, but there's only so much that can be done in that respect. No user serviceable parts inside is for real on this one.

We have found the Supra USB 2.0 cable to be very good, and at this point in time that is our recommendation for an upgrade over a generic USB cable.

 

[Grainger49]

A warning well taken.

 

[mcandmar]

Agree 100%.  I should point out all my experimentation has been with $100 DACs so the cost of any human error would be a mere annoyance.  The BH DAC on the other hand, that will be treated with up most care.  Cue gollems "my precious!" sound clip.

 

[orson]

True. But I think my initial mod could do no harm as it does not add any external power. As long as you keep the other 3 wires intact it should be safe.
The question for the Bottlehead DAC would be: does it need the +5V wire to say hello to the computer as some DACs even with a separate PSU do or not?

 

[Natural Sound]

True. But I think my initial mod could do no harm as it does not add any external power. As long as you keep the other 3 wires intact it should be safe.
The question for the Bottlehead DAC would be: does it need the +5V wire to say hello to the computer as some DACs even with a separate PSU do or not?

Quote from John Swenson in another thread.

Nothing is powered off the VBUS. The VBUS is used, but it just tells the DAC it is connected to a computer. So in normal operation there are just a few nano amps of current coming from the VBUS wire.

John S.

http://bottlehead.com/smf/index.php?topic=6908.msg73242#new

 

[Doc B]

If someone mods a Bottlehead DAC or its power related connections, the warranty on that DAC is void. John Swenson is the designer of the Bottlehead DAC circuit. However I am the owner of the company and I set the warranty policy.

 

[orson]

Alright, so don't try that with the Bottlehead DAC. If I interprete John Swensons statement correct, it wouldn't work anyway as the computer doesn't see the DAC if the +5V wire is disconnected.

 

[Natural Sound]

If someone mods a Bottlehead DAC or its power related connections, the warranty on that DAC is void. John Swenson is the designer of the Bottlehead DAC circuit. However I am the owner of the company and I set the warranty policy.

Not sure if that was aimed at me but I was simply trying to help answer the technical question and was by no means challenging your authority.

 

[Doc B]

I wasn't aiming at anyone. We simply feel it is very important that our customers understand that we can't do anything if they break a DAC by trying to mod it or its associated components like power supply and cables. This is a different beast than kits. We offer help all the time when customers attempt to mod our kits with their own mod ideas and have problems with the mods, because it's usually easy and inexpensive to repair them on a component level.

The DAC on the other hand will cost someone big bucks to replace if they hook up some overvolted homebrew power supply or cut open and rewire a cable and it kills the DAC. There is no component level repair, at the very least an entire board would have to be replaced, and the motherboard would cost around $1000 before including installation labor. We feel it is of utmost importance to be clear about this, and thus I will take the responsibility of mentioning that the warranty will be void whenever someone discusses trying to hack the way the DAC is put together or connected.

 

[Grainger49]

Right.  But only a few of us will have Bottlehead DACs for the foreseeable future.

But this "tweak" has worked on other DACs as the OP observed.

 

[orson]

Lets get back on track with the initial idea after all warnings.
If you feel an improvement after testing like me I suggest to make a nice cable out of the experiment.
I added an additional braided shielding - I think nowadays with all the WiFi and cellular networks you can´t take care enough of shielding your equipment.
As this shielding should only be connected to the source plugs ground we can happily leave the other plug as it is.
Furthermore I added a textil sleeve.
Don´t forget the shrinking tube before assembling the plug. After soldering the 3 wires (black, white, green) to the plug I recommend to add a layer of hot glue to avoid shortening the pins in any case.
For the length of the cable I read several opinions that suggest 1,5 metres to avoid collisions by reflections in the data-wire. Just google that and decide for yourself. I needed that length anyway.

And last but not least: if anybody tries that mod I would be very interested in your findings - either positive or negative.

 

[Grainger49]

You made me laugh with your comment about the shrink tube.  I'm a master of putting RCA plugs on cable, why?  Because I almost always forget to put the cap on before soldering.

Would it be possible to defeat the pin for the 5V supply at the computer end and not have to cut into the cable?

 

[orson]

Would it be possible to defeat the pin for the 5V supply at the computer end and not have to cut into the cable?

Yep, thats how it is meant to be.
The initial "experimental cut" should be as close as possible at the source (=computers") end of the cable.
For final assembling cut off the source plug at the experimental cut (like shown in the first picture) and put a fresh new USB-plug on.
Or do you mean modding the female plug within the computer? If so: I could imagine applying a very small strip of insulation tape to the pin. But that would not last very long.

 

[Grainger49]

I was thinking of taking the pin from the cable end, at the computer end.  The other end is a mess to work with.

 

[orson]

Ah, got it.
But I would not try to remove the pin as you cannot see whats going on inside the plug and you could risk shortening other pins.
So: fast and dirty is the tape method (but not reliable nor durable), best is a fresh new plug.