USB cable comparisons

[sandyk]

If your DAC doesn't need to see +5V from the USB port, try fitting a small piece of insulation in the USB socket to prevent pin 1 from sending out +5V to the DAC.

Alex

[sandyk]

Mayhem13

Perhaps THD + Noise may be more appropriate ?

Regards

Alex

[sandyk]

Each cable was given a points score. Interestingly, they found that longer USB cables, sound noticeably worse than shorter versions of same cable.

So does John Kenny,and many members of his forum. Several members of another forum did a group buy of long expensive USB cables and cut them into shorter lengths with new plugs fitted. I use the shortest possible USB cables with Vbus disconnected at the PC end of the cable. Unfortunately you are unable to do that for some gear.IIRC, Martin also said something about coiled up cables ???

Regards

Alex

[sandyk]

Different types of USB cables can only do so much. It takes attention to detail in the USB power area to gain marked improvements in the LF area as well as right acrossd the board. There are quite a few threads about different products and methods to gain improvements there, including the SOtM USB power solutions.See also:

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/ifi-iusb-power-supply-neat-tweak-and-upgrade-usb-dac-13832/

[sandyk]

Sorry to be a cynic here, but if a 2M length of power cable makes such a huge difference after many KM of plain vanilla copper transmission lines,

and distribution transformers , then it suggests that the power supply area of the amplifier or whatever,could do with a lot of further improvement.

Even care with separation between power leads and interconnects can make a world of difference, as can ensuring that all mains plugs and sockets

are properly maintained.

Alex

[sandyk]

sb6

Agreed. However the best USB cable is no USB cable, or a very short USB cable. Although usually not practical ,a right angle adaptor from the USB device to the USB port has been reported to sound noticeably better.

[sandyk]

Lappy27

You will most likely need a shielded 0 volts wire as well.(black wire in normal USB cable)

This could be in a separate cable, but connected to both USB plugs.

For best results, you may need to ensure that the screen of the cable is not connected to the

Audiophilleo though, in order to remove an earth loop.

 

Regards

Alex

[sandyk]

The problem with comparisons like this, is that many people often choose the file with the most Jitter as sounding better.

It tends to sound louder too.

Barrows did a good post about this some time back, and explained why.

 

Alex

[sandyk]

IMHO this one is a giant killer:

Amazon.com: Belkin Hi-Speed USB 2.0 Cable (10 Feet): Electronics

 

This looks to me like a Computer Scientist poking fun at the 100s of members in this forum , posting in numerous threads,who not only hear differences between various USB cables, but quite often seem to have a marked preference for specific models.

Have a look at the +5V red wire of a typical USB cable at the device end with a C.R.O. while the device is operating, and you may see low level packets of Data among the SMPS rubbish ! Differential Receivers are supposed to be completely immune to things like that, at least in theory. It's also interesting to see that the most respected cables often claim to have improved isolation between D+ and D- and the SMPS +5V power wires.

[sandyk]

Yes, I am poking fun at people trying to fix their problems by cables instead of focusing on the components and an acceptable power supply

 

It's not just due to a suitable PSU. E.E. John Swenson has already been able to demonstrate that RF/EMI goes along for the ride with the data via the USB cable. You also need isolation from this using devices such as SOtM PCI USB cards , iFi USB etc.

Another problem is that when the computer is connected to I.E.C. mains earth, due to many USB devices having an internal connection between Shield and 0 volts leads, there is an earth loop.

[sandyk]

Hi Alex C

Although often desirable as demonstrated by results, it's not always a good idea to do that with some gear, and especially on 230V AC mains systems. It may also be illegal in some countries where the equipment doesn't meet double insulated specifications.

I would be very wary of doing things which may expose members of the family to danger in the event of certain fault situations. In Au. our Neutral is also bonded to earth back at the panel. Earlier installations often used the metal water pipes for this.

My advice would be, that unless you are very experienced in this area, don't even think about going there.

Kind Regards

Alex

 

P.S.

How are things progressing on your J.S. designed Mac Mini Linear PSUs ?

[sandyk]

Thanks for asking. Heading into production shortly. Lots of parts on order already. Not just for Mac minis though. Two outputs, individually selectable voltages, up to 5A+.

 

Alex C

Any further thoughts from John as to why the little Kelvin Sense PCB gave quite a noticeable improvement when the Linear supply would already have been well above average ? Will you be recommending it's use with most installations ?

 

Regards

Alex

[sandyk]

Although improved USB cable design can certainly result in improved SQ, it does not address the issues of isolation etc. that devices like iFi and others attempt to do. Improved cable design is only part of the complete answer.

[sandyk]

Alex, this is interesting.

 

Can you describe how you isolated the 5 volt lead?.

 

Are there USB cables available that already do this in their construction?

 

How do I find out if my Mytec DSD DAC needs this first?

 

I actually chop off the USB-A plug at the PC end and fit a solder type plug without terminating the red +5V wire.

The black (0 volts) wire still needs to be connected though.

I also use a 1.8M cable cut in half, as I am able to use my +5V Linear PSU close to the PC or a USB port of my Oppo 103 which also benefits from the external +5V linear PSU as it uses a SMPS .

There are commercial dual conduit cables available.

If your Mytec DSD DAC has it's own power supply you could try the isolation trick in the plug itself. Other members have done this without the need to modify the cable. Unfortunately, some DACs need to see +5V from the PC initially.

Regards

Alex

[sandyk]

Hi Sandyk,

Have you tried sticking a scope at the end? Did you just play a single frequency tone?Do you have a picture?

Cheers

 

When I looked at the incoming +5V red wire with a scope, I was playing a .wav file stored on the plugged in USB memory stick. No, I don't have a picture.

[sandyk]

At what V/div and t/div were you able to see the noise, do you remember?

I can't remember the settings as that would have been more than 4 years ago.

The CRO was an old Dick Smith 6MHZ one, and the level was fairly low down, and probably on one of the highest ranges.

[sandyk]

Another thing that can help is to feed a preamp from the secondary windings of a transformer located in a separate earthed metal case. I do this with both my Class A preamp and Class A headphone amplifier.

[sandyk]

Yes that is what I am using with my Mytek. The power side is not connected.

 

In which case it must be using the shield of the cable which is internally connected to the 0 volts (black wire) in most USB devices to function.If the PC is using a 3 pin mains plug then RF/EMI will still get into the USB device via the shield, although the situation will likely be improved over using a generic USB cable. A better option may be to use only the black 0 volts wire in the additional cable with the screens of both cables not connected to the actual USB device ?

So far I have been unable to find readily available and affordable lengths of correct impedance cables with 2 separately screened twisted pairs.The only cable that I can readily find has only a single screened twisted pair of the right impedance, but at 7mm thick it is too cumbersome to use a pair of them side by side and connect to single USB plugs at each end.

[sandyk]

It must have been awfully fiddly to extract the red and black wires without damaging the existing shielding and reducing it's effectiveness. I may try using 2 lengths of the single pair cable I mentioned, but remove it's outer jacket and put it all inside a length of heat shrink. The wire gauge will almost certainly be much thicker too.

[sandyk]

I have concluded that there is a hell of a lot going on which simple rationalisation does

not answer.

 

You are not wrong !

I also found, just as another member reported a while back, that even unplugging the cat5e lead from the PC to the Broadband Modem results in a further small improvement.

In fact, I am wondering how much rubbish gets back into the mains via a Broadband Modem's SMPS wallwart.

[sandyk]

Hi Jud

It appears to me to be a ridiculous situation when many people spend more on their USB cables than their DAC is worth.

Due to already using an external +5V Linear JLH PSU with around 4uV noise, modified cables and earthing arrangements etc. my system is pretty well sorted in that area, and any further gains will be minimal. I just like trying to push the envelope !

Even the Phasure NOS DACs are now doing fairly similar with the USB earth isolation area. As for the possibility of heavier gauge wire, it never hurts to have an as low as possible resistance earth/0 volts reference. Where the red +5V wire is not used, it may even be an advantage to double it up with the black 0 volts wire at both plugs ?

My USB leads are no longer than absolutely necessary, so probably not much further advantage for me. When we use USB cables with the +5V wire disconnected at the USB -A plug at the PC end, there may be further minor gains from snipping the + 5V wire right at the internal cable's USB plug where a front mounted USB socket is utilised ? If you have 2 front panel mounted sockets you would of course only do this for one feed.

Regards

Alex

[sandyk]

Alex C

A USB Memory stick will work with just a 220 ohm resistor across the 5V and 0 volt pins of the USB-A plug at the PC end, but you still need the shield to complete the connection, and this isn't ideal either. You may be better off using the black wire and breaking the shield connection at the device end where a separate linear PSU is used ?

If you are using another USB cable from the Linear PSU to the USB DAC you may be able to replace the disconnected shield wire from the incoming cable with a connection to a cleaner IEC earth of the Linear PSU ? ( assuming it is using a 3 pin power plug)

Regards

Alex K

[sandyk]

- There are two senses here: transparent, as you note, and complementing the other parts of the system the listener is using the best.

 

I fail to see how a USB cable can aspire to being more than just "Transparent."

There will always be a degree of colouration, often in the case of remaining RF/EMI that may sometimes appear to further enhance HF detail, and perhaps even give the appearance of an increase in the soundstage . A better than average USB cable on it's own, is simply incapable of doing more than just minimising, but not eliminating, the rubbish coming from the computer that affects the DAC.

That's where USB isolation in the form of iFi USB etc. come in.

 

Alex

[sandyk]

There's a lot of rubbish not coming from the computer too.

 

I presume that you weren't suggesting that what I posted was a lot of rubbish ?

The better quality USB leads have improved isolation between D+ and D-, and the +5V power wiring.

There is a lot of rubbish coming in via the mains supply too, especially from SMPS powered devices, mains powered LED lighting supplies etc.

[sandyk]

I was referring to any device together with RFI and EMI. Saying that the better USB cable are effective only because they reduce the issues from the computer supply is a bit simplistic.

 

O.K. The better cables also normally have a better shielding coverage and will also reduce the ingress of external to the PC RF and EMI .Some varieties of USB cables may in some instances, unless they are relatively short be problematical due to lack of proper screening.