Oh Look, A New USB Cable Thread

[Robert Hutton]

Hi, John. Looks like folks jumped in before I got a chance. :-)

 

OK, here's as far as I can think, not being a tech guy, about a couple of things that may be happening:

 

Notice that the design of the Clearlink does not hold the power and signal legs in such close proximity as most USB cables. I suppose this might conceivably result in less low-level electrical noise making it into the analog side of the system through power and ground. And it might result in lower electrical noise into the DAC's clock, where electrical noise can result in worse jitter performance.

 

Two other things that may be going on (but I confess I lack the understanding to reason how it might affect the signal coming through my DAC's async USB input) are the effects of reduced dielectric "pumping" and Mapleshade's provisions to minimize "skin effect" in the conductors. Dielectric "pumping" results in a faint "echo" of the signal being released at a slight delay from the original. Skin effect causes differences in the speed with which different frequencies propagate through the cable. Minimizing these problems means the analog signal through the USB cable that represents the digits in the stream maintains better integrity in terms of timing of the zero-crossing points, helping to minimize jitter. (If I didn't word that last sentence clearly, a close analog to what I'm talking about is nicely represented pictorially in Figure 1 of Damien's paper on integer mode: http://www.amr-audio.co.uk/large_image/MAC%20OSX%20audio%20players%20&%20Integer%20Mode.pdf .) However, as I said above, I'm not sure how or whether any timing effects in the cable might make it through to the conversion process in an async USB DAC, where the DAC's clock is controlling the timing of the digits out of the bitstream.

 

So, to sum up: Maybe there could be reduced impacts from electrical noise in terms of noise getting through to the analog side and jitter performance of the DAC's clock. Beyond that, I'm not tech-savvy enough to have anything solid to say.

 

Problem with this theory is - there is no "analog signal through the USB cable", only a digital one.

 

Still, it is absolutely true that capacitance of a USB cable and the degree to which it deposits other noise and junks through to the DAC - which ultimately gets to the DAC's analog stage - may have a degree of sonic impact. And different DACs have different re-clocking algorithms which will come into play also, acting differently depending on the quality of the digital stream being sent to it.

 

Many other things come into play with USB sound - the size of the computer's RAM buffer, how the media player is set up, whether an external hard drive is used or an internal drive (and which one), and so on.

 

So IMO a USB cable can sound a bit different, but having tried a few, not a whole lot. I suspect that getting the power leg out of the equation may be a good move, I will surely try that at some point.

 

I also suspect, without direct experience, that a lot of 'garage makers' and 'cottage brands' of USB cables sound different because they don't operate at the correct capacitance - I believe a UK audio magazine did a story on this last year and showed through measurement that a large percent of their sample did not measure correctly electrically. Defeats the purpose for me, but the sonic impact might be preferable to some.

[Jud]

Problem with this theory is - there is no "analog signal through the USB cable", only a digital one.

 

Hi, Robert. What many people fail to take into account is that the "digital" bitstream in a USB cable is an electrical representation of 1s and 0s. This signal cannot, as a matter of physics, have infinitely fast rise time to whatever amplitude is designated as representing 1, or infinitely fast fall time to whatever amplitude is designated as representing 0. So the electrical signal is in fact not digital (i.e., either 1 or 0 with no transitional values), but analog (waves with finite rise and fall time - in other words, wavelength - and amplitude). To minimize the chance of dropouts (neither 1 nor 0), if the signal is above a designated amplitude it is evaluated as a 1, below as a 0. So anything that shifts the time at which the signal rises above or falls below this so-called"zero crossing point" is a potential cause of jitter. However, as I noted in my prior comment, where the DAC's clock controls timing of the bitstream into the conversion process, as with an asynchronous USB input, I don't know if such timing effects in the cable could have any impact.

[Robert Hutton]

Hi, Robert. What many people fail to take into account is that the "digital" bitstream in a USB cable is an electrical representation of 1s and 0s. This signal cannot, as a matter of physics, have infinitely fast rise time to whatever amplitude is designated as representing 1, or infinitely fast fall time to whatever amplitude is designated as representing 0. So the electrical signal is in fact not digital (i.e., either 1 or 0 with no transitional values), but analog (waves with finite rise and fall time - in other words, wavelength - and amplitude). To minimize the chance of dropouts (neither 1 nor 0), if the signal is above a designated amplitude it is evaluated as a 1, below as a 0. So anything that shifts the time at which the signal rises above or falls below this so-called"zero crossing point" is a potential cause of jitter. However, as I noted in my prior comment, where the DAC's clock controls timing of the bitstream into the conversion process, as with an asynchronous USB input, I don't know if such timing effects in the cable could have any impact.

 

Fair enough, good point. Although that electrical 'representation' is still simply a series of pulses, and as you say, the whole point of asynchronous USB is to reclock that series coming in and correct it.

 

But I can see where USB cables can have sonic signatures due to the other electrical 'noise' passing through to the DAC. Or maybe there are things happening with different cables that have yet to be understood/explained.

 

Along those lines, I have been using Neotech CUG-1 fluid treatment on cable jackets - including the USB cables. On my video setup, it had a massive impact on detail and color, as well as sound. On my main USB, it did also have a notable sonic impact. There is no logical explanation I can think of why.

[mayhem13]

Let's try and not confuse the issue. Yes, it's a waveform, but a digital waveform where the if stream fails, the bit is lost or misrepresented.

 

As for piggybacked EMI, discerning it's presence and identifying it's impact, simply 'listen' to the ambient noise floor of each cable seperately or if you don't trust your ears, measure the ambient noise floor. You'd find if you did this in an anechoic environemental, the change would be a fraction of real world noise, including inhale and exhale. For one to make the case that such small changes would impact complex music signals at normal listening levels is far from probable.

 

Now since we're talking the better designed DACs here, and the case of the OP being the BitFrost which I've recenty had the chance to hear, we'd have to make the assumption that Schiit implemented a pretty bad circuit by design that couldn't reject or filter a little EMI that a cable could.....counter the support the product has received from it's users. To that, and the need for high end or specialized USB cables......let's ask Schiit themselves? Would a consumer find it reasonable that a mfgr of a $500 stand alone DAC would suggest a premium USB cable to reduce inter modular jitter and EMI?.......as we neglected to include suitable protection in the unit itself. Not likely but hey, maybe there's some new world business model out there for this new disposable economy.

[jhwalker]

Let's try and not confuse the issue. Yes, it's a waveform, but a digital waveform where the if stream fails, the bit is lost or misrepresented.

 

As for piggybacked EMI, discerning it's presence and identifying it's impact, simply 'listen' to the ambient noise floor of each cable seperately or if you don't trust your ears, measure the ambient noise floor. You'd find if you did this in an anechoic environemental, the change would be a fraction of real world noise, including inhale and exhale. For one to make the case that such small changes would impact complex music signals at normal listening levels is far from probable.

 

Now since we're talking the better designed DACs here, and the case of the OP being the BitFrost which I've recenty had the chance to hear, we'd have to make the assumption that Schiit implemented a pretty bad circuit by design that couldn't reject or filter a little EMI that a cable could.....counter the support the product has received from it's users. To that, and the need for high end or specialized USB cables......let's ask Schiit themselves? Would a consumer find it reasonable that a mfgr of a $500 stand alone DAC would suggest a premium USB cable to reduce inter modular jitter and EMI?.......as we neglected to include suitable protection in the unit itself. Not likely but hey, maybe there's some new world business model out there for this new disposable economy.

 

To answer your question, Schiit definitely does *not* recommend "high end" USB cables. In several exchanges with Jason, he adamantly insisted we should be using only a good quality, certified USB 2.0 cables; in fact, he stated that such "high end / audiophile" cables were frequently the *cause* of problems users were having. FWIW.

[mayhem13]

.......interesting. I like the people over at Schiit more and more.....

[Jud]

Let's try and not confuse the issue. Yes, it's a waveform, but a digital waveform where the if stream fails, the bit is lost or misrepresented.

 

Agreed, as long as we include timing issues (those that are potential causes of jitter).

 

As for piggybacked EMI, discerning it's presence and identifying it's impact, simply 'listen' to the ambient noise floor of each cable seperately or if you don't trust your ears, measure the ambient noise floor. You'd find if you did this in an anechoic environemental, the change would be a fraction of real world noise, including inhale and exhale. For one to make the case that such small changes would impact complex music signals at normal listening levels is far from probable.

 

I do have to get back to my "Noise and Objective Testing" thread with that second thematic comment I keep threatening to post, and actually will one of these days soon.

 

Now since we're talking the better designed DACs here, and the case of the OP being the BitFrost which I've recenty had the chance to hear, we'd have to make the assumption that Schiit implemented a pretty bad circuit by design that couldn't reject or filter a little EMI that a cable could.....counter the support the product has received from it's users. To that, and the need for high end or specialized USB cables......let's ask Schiit themselves? Would a consumer find it reasonable that a mfgr of a $500 stand alone DAC would suggest a premium USB cable to reduce inter modular jitter and EMI?.......as we neglected to include suitable protection in the unit itself. Not likely but hey, maybe there's some new world business model out there for this new disposable economy.

 

A few comments on this:

 

- Yes, I like the Schiit folks, too. :-)

 

- Notice that the cable that's the subject of this thread is about half the price of the one it replaced. So I'm definitely not up for telling people to buy accessories that are comparably expensive to the main items of equipment they've got. (E.g., USB-to-SPDIF converters that are more costly than the async USB DACs they're attached to.) To me, if you need high-priced accessories to make something sound good, that's Nature's way of telling you to buy a better piece of equipment. In fact, I bought the Coffee that the Clearlink replaced only because I got what I considered to be a screaming deal on it (which involved *two* substantial discounts, one from the seller and one from a finance company as an introductory tease).

 

- Really, really great EMI rejection from ground and the DAC's clock is not just a consequence of the DAC's design, but involves all the other system components, as well as environmental factors (house electrical system, RF environment). With regard to just the DAC's design, near-absolute isolation involves things like opto-coupling that aren't currently going to be available in a $500 DAC.

 

- But let's not let the perfect be the enemy of the good. Just because near-perfect isolation is not practically achievable at the relative bargain price range of my DAC, that doesn't mean improved associated equipment can't conceivably improve the sound. I'm subjectively hearing improved sound quality with a USB cable half the price of the one it replaced.

 

Whether one of the potential explanations I was able to think of as a layperson is the cause, or whether there's a different explanation, or whether I'm just making it all up in my head, I can't tell you for certain. But hey, I'm happy. Looking forward to reading the experiences of others who've ordered the Clearlink, and interested to see whether they agree with my impressions or are very different.

[AlainGr]

Though I am not technical, I hope to post some thoughts about the Clearlink when I receive them (got myself a iUSB a few weeks ago and after reading Jud's posts, I decided to order some Clearlinks).

 

I also want to add that Pierre (of Mapleshade) is a very ressourceful person. The prices are reasonable and the documentation is very interesting.

 

Alain

[mayhem13]

Didn't mention I did enjoy the Bitfrost very much.....for the owner who's I auditioned my only comment was money well spent.

 

Me as well Jud, i try not to allow the cost of an item predetermine it's performance. I've build systems with $20 tweeters that measured and performed better than $100 counterparts.

[Jud]

A request for iUSB owners listening to Clearlinks - I'd love to know what your impressions are with two Clearlinks with the iUSB vs. one Clearlink without the iUSB.

 

OK, I'll owe you beers when I meet you. (Not entirely unlikely - Blake, my brother lives in SLC, and Alain, I've always wanted to visit the Montreal area.)

 

[AlainGr]

Will do Today would not be a good day for visiting, as we are under another snow storm, but no problem for the beer

[Blake]

A request for iUSB owners listening to Clearlinks - I'd love to know what your impressions are with two Clearlinks with the iUSB vs. one Clearlink without the iUSB.

 

OK, I'll owe you beers when I meet you. (Not entirely unlikely - Blake, my brother lives in SLC, and Alain, I've always wanted to visit the Montreal area.)

 

 

 

mmmmmmm...... beer. Consider it done!

 

I do have my work cut out for me in the demo arena. A care package from Sablon Audio is set to arrive tomorrow (this a box full of Sablon Audio demo power cables roaming around the USA to allow us Yanks to try these British cables). Clearlinks should arrive shortly, and I just swapped in some new tubes in my amp. I will definitely report my thoughts but it may be a week or two.

[Blake]

A request for iUSB owners listening to Clearlinks - I'd love to know what your impressions are with two Clearlinks with the iUSB vs. one Clearlink without the iUSB.

 

A very late follow-up. Several months ago, I acquired a second Clearlink, to use with my iUSB and my other Clearlink. Unfortunately, this setup resulted in noticeable, constant popping noises occurring every second or two, similar to vinyl with static or dirty grooves. I called Pierre Sprey of Mapleshade and he mentioned another iUSB with the same problem.

 

As a result, I was not able to compare 2 Clearlinks + iUSB, to 1 Clearlink and no iUSB.

 

The other item worth mentioning is that while the iUSB noticeably improved the sound of my system when using a s/pdif converter powered by the computer's usb hub, when I upgraded to the BADA converter, which has its own robust power supply, I could no longer detect any improvement or difference in the sound with the iUSB in place.

 

Finally, I want to mention how pleased I am with the Clearlink Plus usb cable. I compared the Clearlink head-to-head against many different usb cables in my system, most of them quite a bit more expensive, and the Clearlink beat them all. Off the top of my head, the Clearlink beat: iFi Gemini dual head, KingRex uArt dual head, Nordost Blue Heaven, AQ Coffee, AQ Cinnamon, Kimber Silver, Kimber copper and some others I am not remembering at the moment.

 

The Clearlink gets everything right in terms of sonics. The aspect where the Clearlink is clearly (pun intended) superior to the other cables is transparency. It is extremely open, airy and quick sounding. No veiling. The result is a very realistic presentation, that helps with creating the illusion of having the band in the room with you. No more experimenting with usb cables for me. The Clearlink is here to stay.

[barrows]

Hmm: Jud thanks for being bold enough to post this. Since this cable is unshielded, it does not meet the USB cable specification, and cannot be guaranteed to work. I certainly like the idea of attempting to make a high speed cable for USB transmission, but I am very skeptical about unshielded cables carrying high speed signals in close proximity to analog circuitry. Clearly, the cable relies solely on twisted pair configuration to control RF emission. Perhaps twisted pair is enough? At the speed of USB signals, these signals are as likely to travel through the air as through the cable; without shielding, this is troubling to me. It is very possible that this cable is broadcasting a lot of RF, which may be getting into the audio system. I would recommend very careful cable dressing, and keeping this cable as far away from analog circuitry and cabling would be advisable. This is making think I should build a USB cable myself using thin Cardas litz in a twisted pair, and just a ground wire (I do not need USB power) as an experiment.

 

Blake: Note that the Berkeley Alpha USB does use USB power to power the "dirty" side: the USB receiver and XMOS processor, it's onboard power supply is devoted to powering the oscillators, and SPDIF output circuitry. This approach allows for complete galvanic isolation of the computer, and (XMOS) processor noise form the output.

[Jud]

Hmm: Jud thanks for being bold enough to post this. Since this cable is unshielded, it does not meet the USB cable specification, and cannot be guaranteed to work. I certainly like the idea of attempting to make a high speed cable for USB transmission, but I am very skeptical about unshielded cables carrying high speed signals in close proximity to analog circuitry. Clearly, the cable relies solely on twisted pair configuration to control RF emission. Perhaps twisted pair is enough? At the speed of USB signals, these signals are as likely to travel through the air as through the cable; without shielding, this is troubling to me. It is very possible that this cable is broadcasting a lot of RF, which may be getting into the audio system. I would recommend very careful cable dressing, and keeping this cable as far away from analog circuitry and cabling would be advisable. This is making think I should build a USB cable myself using thin Cardas litz in a twisted pair, and just a ground wire (I do not need USB power) as an experiment.

 

 

Well, I see three cables/wires, and only one looks twisted to me. I can tell you I don't hear any noise; especially since adding the Spectral amp last year to the preamp I've had for 20 years, the music seems to come from an absolutely silent background. We are talking about just my ears, though. I don't have the type of test equipment you'd want to make really sure of that.

 

I think if you could get hold of Pierre by email or phone, that might be a very worthwhile and interesting correspondence or conversation. For an aerospace engineer who used to ride herd on Pentagon contractors very effectively (he's still an acknowledged expert on defense spending and force strategy in the area of air power) and a record producer who builds pretty much everything he uses to make his very excellent recordings (which I've near-pleaded with him to offer in hi res, but just CDs so far), many of his ideas seem way out there, right up until you hear them yourself. I really like Audioquest USB cables, and this thing at half the price absolutely beat the pants off a Coffee, not close. Then for good measure Pierre's cheap power strip improved on the MIT Z-Strip it replaced. (I'd said to him the USB cable sounded a little light in the bass right out of the box and felt it would improve with break-in; Pierre said it shouldn't be light in the bass even at the start and asked me to listen to it with his power strip, and he was right. I wrote him only half-jokingly, "Huh, bass actually plays notes instead of just rumbling along underneath everything. Who knew?"

 

I'd guess he'd be a very worthwhile sounding board re a DIY USB cable as well.

[barrows]

Hey Jud, I believe for the most part in Pierre's cable design approach: minimal dialectrics, thin conductors, and focusing on cable speed. This is similar theoretical approach to what Nordost aims for, as we have discussed before. Not so sure about the Mapleshade recordings though, maybe I have not heard enough of them. We had two CDs at PS Audio, and to me while they exhibited a remarkable sense of clarity from the upper midrange to the high frequencies, they also seemed to miss out on the body of the music, as if the warmth and body was bleached out in favor of upper frequency air and extension.

Anyway, on the USB cable, I am sure the twisted pair are the two halves of the USB signal (which is a psuedo balanced signal) and the other two wires are the ground and +5V. Once I get my hands on some 28 AWG Cardas litz, I'll try a DIY version of something similar for an experiment.

As for the RF concern, I would not expect RF transmission from the cable to produce audible changes in the noise floor, RF sound degradation like this generally manifests by modulating the audio signal, and degrading it in subtle ways. Sometimes (and without measurements I am not saying this is happening here) is can even create a (false) impression of increased HF extension/air. If you have a portable old school transistor radio around with a pair of headphones, you can sniff around the cable for RF emissions by tuning the radio low in the AM band to a position where there are no stations and listening for noise: it is surprising how much you will find with such a simple test!

[Blake]

Blake: Note that the Berkeley Alpha USB does use USB power to power the "dirty" side: the USB receiver and XMOS processor, it's onboard power supply is devoted to powering the oscillators, and SPDIF output circuitry. This approach allows for complete galvanic isolation of the computer, and (XMOS) processor noise form the output.

 

Thanks for the clarification Barrows- you are right. What I intended to convey was that one of my prior s/pdif converters was completely usb powered (Bel Canto uLink) and the iUSB did improve things with that converter in the system.

 

I am not sure if usb cable experimentation is high on your "to do" list, but in case it is, Mapleshade does have a 30-day money back guaranty. I obviously love this cable, but others may prefer different sonic qualities. If for nothing else, to me, the Clearlink is proof that not all usb cables sound alike. Many of the usb cables I purchased and demo'd in my system for extended listening sounded quite similar. The Clearlink had the most "unique" sound that was, for me, easily and reliably different sounding than the others- it is reliably more transparent sounding than any other usb cable I tried.

[Jud]

Anyway, on the USB cable, I am sure the twisted pair are the two halves of the USB signal (which is a psuedo balanced signal) and the other two wires are the ground and +5V. Once I get my hands on some 28 AWG Cardas litz, I'll try a DIY version of something similar for an experiment.

 

Yep, that sounds right. I do think you could pick up some tips on the DIY (remember, as a wise man once said, "everything matters," and Pierre has been there, done that, very, very thoroughly) and/or have a good, interesting conversation with Pierre if you wanted to contact him.

 

As for the RF concern, I would not expect RF transmission from the cable to produce audible changes in the noise floor, RF sound degradation like this generally manifests by modulating the audio signal, and degrading it in subtle ways. Sometimes (and without measurements I am not saying this is happening here) is can even create a (false) impression of increased HF extension/air. If you have a portable old school transistor radio around with a pair of headphones, you can sniff around the cable for RF emissions by tuning the radio low in the AM band to a position where there are no stations and listening for noise: it is surprising how much you will find with such a simple test!

 

Going down the old mine

With a transistor radio

Standing in the sunlight laughing....

 

Memories! :-)

 

Don't think I have an old transistor (wish I did, just for those memories), but maybe there's an old battery powered boom box with an AM band around here somewhere.

[T1000]

I have a Cardas Clear USB that works great with some dacs (and other dispositives) that i have used as M1DAC, Yulong Da8...

BUT when i connect it to interface Audio-Gd DI-V3, there are several problems! music stops, restart pc, pops/cliks sounds... like a hell.

First i thought: the interface is wrong (is new). But i connect with stock usb cable (chinese) and works perfect!

This have not explanation!! Interface and usb cable works well but this combination do not work!

I investigate more and saw that conector type "B" in cardas are different! (the design of cardas conector have two notch up and down in the interior plastic of conector)

and this produces wrong contact.

 

I want to buy mapleshade cable but i would like to see frontal of type B conector.

I will put photo of chinese vs cardas type B conector to you see the differences.

 

Regards.

[T1000]

 

look the differences

 

left: stock

right: cardas

 

Regards

[Allan F]

Even if the chip doesn't reject the noise, this noise still doesn't alter the actual content of the data......and if the EMI was so severe as to affect the differential receivers ability to receive the stream, the result would be clear dropouts of content, not more 'air' between guitars or tighter bass. Such claims as I asserted above need a much more complex, rational answer as to why. If such an answer doesn't exist, I'm OK to let it go as just one of those things that defies explanation.

 

But, if the USB cable carries noise from the PC to the DAC, that noise can also affect the analog portion of the DAC which produces the ultimate sound you hear.

[drez]

Probably unnecessary but I thought I would offer few thoughts from my limited understanding on USB cables. Basically if we are talking about things like capacitance and skin effect this means that we are worried about how well the cable will transmit a square wave. The aim with digital cables is to transmit as square a square wave as possible within certain limits. This video does a better job of explaining basic concepts of digital cables than I can:

 

 

When it comes to D+/D- pair, the purther these are spaced, the greater the inductance - when a pair of conductors which transmit differential signal are paired together, inductance is lowered, noise pickup is reduced, and noise radiating from the cable is reduced. Additionally, when D+/D- conductors are at varying distance, the impedance between the conductors changes, and this results in signal reflection, adding even more noise to the signal. When the D+/D- pair are in close proximity, possibly with a certain twist rate, this is more efficient for transmitting the signal. There is the slightly downside that capacitance is higher, however all the other advantages mentioned above greatly outweigh the slight disadvantage of some capacitance. Almost all cable manufacturers use closely spaced conductors in digital cables - they do it for a good reason.

 

Basically capacitance and inductance a similar effect on the signal, they reduce the size of the "eye" of the USB signal, which increase jitter. By spacing the wires, capacitance may be lowered, but inductance is increased, so the net effect on the signal waveform is detrimental especially considering the other problems like impedance mismatches, noise pickup, noise emission, etc.

 

I apologise if this seems to have nothing to do with the cable discussed in this thread, however I thought it might be useful if I tried my best to explain my understanding of digital cables.

[freddie40]

Got my new Mapleshae Clearlink Plus USB cable today (thanks for the recommendation Jud). I do not need the 5v (the red wire) so I can easily cut the wire at both ends and remove it. Would you? I wouldn’t do it for a month at least.

 

Dave

[sandyk]

Dave

There is a very good chance that you will either damage the cable, or reduce it's performance by doing that.

If you have soldering experience,it would be far better to fit a new USB-A solder type plug at the PC end without terminating the red (+5V) wire. You can then fit some heatshrink over the new plug.

Alex

[Jud]

Got my new Mapleshae Clearlink Plus USB cable today (thanks for the recommendation Jud). I do not need the 5v (the red wire) so I can easily cut the wire at both ends and remove it. Would you? I wouldn’t do it for a month at least.

 

Dave

 

I'd recommend asking Pierre before doing this or trying out sandyk's suggestion.