USB cable comparisons

[Superdad]

Any failures in the sysem will result in necessary retransmissions, affecting the data flow.

 

Amithlon: I have no problem with everything else you wrote, but this one line above is incorrect. There is no error correction in the USB protocol, hence there are never retransmissions.

 

 

- is Supra a TE connectivity cable? This low impedance (90 instead of 150) might well explain my problems.

- English is not my native language: what do you mean by mushrooms, other than the ones I can eat :-)

 

Hi Johan:

--TE Connectivity (a corporate behemoth with US$13 billion in sales and 90,000 employees) is not at all related to Jenving Technology, the 40-year old Swedish firm that manufactures a variety of very high quality cables under the Supra brand.

 

--90 ohms is the desired nominal impedance for USB cables, though many brands are actually at about 100 ohms. 150 ohms would be well outside the USB 2.0 spec. (By the way, I use and recommend the Supra USB cable and find it to be excellent--and amazing for $50. I have compared it to a couple of $400-650 USB cables and preferred the Supra. Very rich and detailed; no congestion, no fatigue.)

 

--The mushrooms jabbr was referring to have nothing to do with audio. He was trying to be funny and was referring to psilocybin "magic" mushrooms--the hallucinogenic kind.

Well, those mushrooms are not entirely unrelated to audio: When I was 16 in 1978, I tried some magic mushrooms for the first time (safely at home with the hi-fi in my bedroom). I had this beautiful gatefold of a double-LP Santana album pinned to my wall:

 

At some point during my "trip," I looked up and those clouds were very clearly moving among the mountains. Can't recall what music I was listening to that day, but I sure do remember that "living" album cover!

[Superdad]

I spent 5 minutes looking for a description and a link to buy. I see none, except for a pic .

 

Don't see why I should bother if the sellers aren't.

 

Hi fmak (Fred isn't it? From over at AA?):

 

Sorry that info on our forthcoming, John Swenson designed, USB REGEN device has been coming out piecemeal. CA is not a place for promotion and I dislike putting out too much in advance of a product being ready to ship, but some "fans" and friends here have done a great job of racing ahead of me.

Anyway, production boards and enclosures arrive at the end of March and shipments will begin by mid-April. A web site for ordering should be up by then.

 

Here is a very brief description of the UpTone USB REGEN:

 

The REGEN's primary function is to generate a completely new USB data signal from a carefully chosen USB hub chip running from an ultra low-noise regulator and low-jitter clock--which it does with ideal impedance matching and right at the input of your DAC.

 

Its secondary function is that it interrupts the 5VBUS of the USB cable coming into it, and provides clean 5VBUS on its output--for DACs that need it--via a second ultra low-noise regulator.

 

The REGEN is about improving the "signal integrity" of what the DAC or converter is fed so that its USB PHY chip does not have to work as hard to decode the data from the voltage/noise. That results in less generation of packet noise voltage (in the data steam and often around the 8kHz USB packet rate but also broadband), less perturbation of the PS from spikes, and less ground-plane noise.

If you can hear a difference between USB cables and upstream computer stuff, then the REGEN will definitely have an effect.

 

Here is the link to a recent post listing dimensions and describing everything that will come in the box for $175: http://www.computeraudiophile.com/f10-music-servers/uptone-audio-regen-22803/index6.html#post403638

 

To those who are familiar with the Schiit Wyred, the USB REGEN serves the same functions, but we believe in a more focused way (no AC PS inside the box, no output USB cable needed, really good clock and expensive regulators, 4-layer board and careful impedance matching).

It is pure coincidence (or that "great minds think alike") that the Wyrd and REGEN are similar. John started working on the REGEN months before the Schiit Wyrd was even announced. I recently discovered that both devices even use the same USB hub chip.

Schiit products are always a tremendous value--I admire them and often wonder how they do it--and my tiny firm can not compete with their volume or pricing. It will be up to the market to decide if the sonic differences between our two devices are worthy of the price difference.

I actually have a Wyrd here on my desk. It arrived last week but I have been too busy to give it a listen. Even after I do, I doubt I will say anything about it:

a) it is bad form for a manufacturer to make subjective claims about another's product; b) I have a lot invested in this project so talk about "expectation bias" going in!

 

Thanks and regards,

Alex C.

[Superdad]

Then why source boxes from expensive Japan?

 

Hi Fred:

 

The answers to your questions belong over at the REGEN thread, but I'll address them here since you seem to be following on this one.

 

While it is true that the big chassis with heat sinks for our JS-2 LPS comes to my cost of almost $160 (landed at my door, after complete custom milling and silkscreening), it is a far nicer, precise and flawlessly finished piece than anything I could source from China. And USA costs for a comparable custom chassis in aluminum would be through the roof for our quantities. (There is a reason Schiit goes with painted, bent steel chassis, and their high volume plays into that).

 

As for the little extruded aluminum case for the REGEN, my Japanese supplier is entirely competitive with China--and again with a nicer part and complete milling and silkscreening with zero defects. And the fast service and fair prices at 100-500 pieces is impressive. Trust me, I literally spent days contacting vendors on three continents (at least those who had something decent to offer).

 

''Direct usb connection and smps PS''

 

I personally find direct usb connectors a liability if there is any weight or size. Is this a SQ issue?

 

A key part of John's design for the USB REGEN is the careful impedance control of the output. Some of that is lost by adding a long USB cable. He originally wanted to have me solder a short USB 'B' plug pigtail to output of the board, but we decided against that for several reasons, not the least of which was the labor involved and the challenge of obtaining decent pre-made cables with plugs. Plus, it is more flexible to have a USB 'A' output so people can try the REGEN in various applications. (If someone wants to modify their REGEN output with a custom short cable/plug, they are welcome to do so--at their own risk. They could probably stick a rubber grommet in the hole where the 'B' jack was to give the cable some strain relief.)

 

As for strain due to weight or size, the REGEN, with the solid A>B adaptor, weighs about 1.4 ounces (40 grams). So any strain will not come from weight. Rather it will be from the input USB cable and the DC plug cable. Unfortunately, DACs have many different USB input-jack heights, so there is no one-size-fits-all little block/cradle to be offered. Users can get creative--one fellow said he may 3D print a cradle/block for his, though perhaps he will share the model file for others to adjust height and print their own. Me, I use a little foam block that does not slide around. Maybe the REGEN packing box will have a 1"x 1.5" x 2" foam block that might work on one side or another for some users. I have not had time to look for such a thing. Am always open to suggestions.

The REGEN will also come with a 6" USB A>B cable for those who have space issues behind their DAC or due to other jacks/cables.

 

 

I also don't like the smps. Some designers will say otherwise, but to a fair no of pairs of ears, a good linear sounds better with such USB devices.

No doubt you would have done listening before you chose the PS and perhaps you didn't hear anything untowards.

 

I won't deny that a really good linear PS will slightly improve the SQ from the REGEN. I use the second output of my JS-2, but any decent unit in the 6-9V range with reasonable output capacitance will do nicely as well. So I suspect that many REGEN buyers will experiment.

 

However, when speaking about small supplies (tabletop bricks or wall-warts) that can be included with a product, the choices narrow considerably. Cheap linear PS units of small size are mostly crap for audio system applications.

 

First off, virtually all the reasonable tabletop non-SMPS bricks are not regulated units--just transformer/diode affairs with a small filter cap. They are rather nasty. Then, pretty much all the small, factory regulated linears are wall warts (and the interchangeable plugs for different countries are not available for them). Their current ratings are really low (at least for any I could find in the 6-8V range), and while the REGEN does not need much, the low current linear warts sounded bad--which we suspect is partly due to having just a 1000uF or less filter cap at the output.

 

So after trying a small pile of cheap linears and SMPS of various brands and ratings, I found that the 22W Mean Well (a fairly advanced company producing solid and reliable units meeting stringent specs) was the winner and at a fair price. Their bigger ones, are well, big--and several times the price, and would move the whole kit price up will little gained--better to have people save for a real LPS.

BTW, very low noise SMPS units do exist, but they are really expensive (and not packaged externals)--see the best of them Daitron Ultra Low Noise Switching Power Supply

 

Anyway, in the end we find the REGEN sounds great with the supply I have chosen to include (the Mean Well GS25A07). A big part of that is the power network John designed for it. We use two of the really good TI TPS7A4700 ultra-low noise regulators (those fellas are $4.05 each even at 100+ qty.), plus carefully chosen capacitor sizes.

 

I hope the above answers most of your questions Fred. I know you are a technical guy, and I just wanted you to understand that this is actually a very carefully thought out product.

 

Cheers,

 

--Alex C.

[Superdad]

Will the REGEN work with 5 volt power? I have a high quality 5 volt line running to the DAC powering the jcreate hub. I am hoping to re-use it for the REGEN.

 

Well I was about to say no, because at least 6 volts is needed to allow one of the regulators to make clean 5V VBUS power (regs need to have some drop to do their thing). But the other regulator--the one for the hub chip and clock serving the REGEN's primary purpose is making just 3.3V and 1.8V. So that side runs fine on 5V.

 

Only thing I don't know is if whatever reduced voltage the REGEN will be putting out on USB pin 1 will be enough for the "all good" handshake with DACs that look for it. The only DACs I have here either don't require ANY USB bus power at all (not even for handshake) or actually run on bus power (one entirely, another just to for the USB input stage). So I can't really test if those DACs that don't draw bus current but do want handshake voltage will work if you only give the REGEN 5V.

 

Maybe John or I will play around with that when we have time, but that is one thing we are both rather short on right now.

 

Ciao,

AJC

[Superdad]

 

Agreed, the Corning cable is not quite dead yet. It has many properties that are attractive and make it worthwhile.

......

Lastly, I am very curious to hear a galvanically isolated Corning cable, and will continue to pursue this goal.

 

Hi Larry:

I agree with both you and EuroDriver. And John is not done with his testing/disceting of the Corning. However, it looks like the only way to power the Corning separately--at the receiving end, thereby achieving galvanic isolation from the computer--is going to be by cutting open the end module and finding the right solder points. I just did not see how the fine wires on the fiber cable could be accessed without damaging the optical cable itself. Maybe John will find a way to do it under his microscope.

 

All that said, it is important to understand what a complicated, dual-protocol, very active device the Corning really is. I think John wrote some about how it works, but he also bent my ear extensively about it (most of which I have forgotten). This was during the week when I pushing him to think about and sketch out how we might be able to offer an UpTone optical device pair. He explained the challenges and shortcomings to doing so, and since it would have all been in the name of achieving galvanic isolation, we went back to our discussion and reasearch on accomplishing that with an "über-Regen," one that uses an FPGA with a licensed hub core, two PHYs, two clocks (I think), 4 of the expensive ultra-low noise regulators we like, a power isolator chip, and one DC-DC regulator, plus support caps and resistors. Without even knowing the licensing fees for one of the very few available USB 2.0 hub cores, it was clear that such a product would be much more expensive than the present REGEN. Then we did the resistor thing to make the "amber" REGEN, and "über-Regen" suddenly seemed rather unimportant to pursue. John and I are much more interested in proceeding with a couple of larger, vastly more innovative and ambitious projects that have been gestating on his bench for a while. I appears that the success of the USB REGEN is going to make that possible. But it will take lots of time.

[Superdad]

12 minutes ago, Ralf11 said:

true but not really relevant to noise - you do know about noise and signal extraction, right?

 

If one spends time looking at variations in signal integrity eye-patterns  (caused by cables and hub chips)—and simultaneously measuring activity and ground-plane noise around the USB PHY/processor in the DAC—then it starts to make sense. 

[Superdad]

1 hour ago, xyzzy1 said:

from what I observe I believe digital cables inject noise into the DAC. No cable will improve sound, better cables will only degrade it less. The perfect cable material, I think, would be super conducting. 

 

Noise in cables comes from impurities in the metals causing reflections. From junction points where the cable connects to the jack metal also causing reflections and ringing among many other causes of noise being generated. 

 

There is is no such thing as perfect isolation between components. Noise generated in the cable by poor contacts, reflections, etc is going to get into the dac and into the analog amp in the dac. Better cables have better (purer) materials and better junction points causing less reflection noise that is the cause of electron turbulence.

 

Sadly, for USB cables what you wrote is not at all how it works.  It is all about variations in  USB signal integrity (as seen in an eye pattern—which quantifies jitter, slew, amplitude, and a form of noise), and also shielding. These are things that have an effect on the DAC’s input stage and affect how much ground-plane and packet-data noise the USB PHY and processor generate inside the DAC—which in turn can effect sensitive power supply and ground circuits, resulting in small variations at the DAC chip’s master clock input pin.

This has been written about extensively.  It can be measured at very low levels, though it is not hard to hear in a good system with acoustic music.

[Superdad]

2 hours ago, seeteeyou said:

Maybe signal integrity ain't snake oil after all?

 

Ya think?! 

[Superdad]

10 hours ago, Teresa said:

I guess we misunderstood each other, I'm glad in the end we agree.

 

Sal & Teresa agreeing?  Not since the friendship between Justices Ginsberg and Scalia has there been a more unlikely pairing.

A Christmas miracle if I ever saw one! 

[Superdad]

1 hour ago, OldBigEars said:

I've never been able to move the tiny, little switch that apparently turns the VBUS on/off.  I can't even tell for sure what position it's in.  Does that matter and why would anyone want to turn it on/off?

 

Regarding the VBUS switch on UpTone USPCB A>B Adapters:

It should stay in the ON position (that is the downward position of the tiny white switch—where it then forms a line with the two white painted dots on either side).  

 

It is a bit of a joke—and really only useful for use in determining if you DAC needs VBUS.  If it does, then you leave the switch on; If it does not use VBUS then well, what’s the point of tuning it off?  It is not like the 5V goes anywhere.  You are not even keeping it off a long USB cable (some people like that to reduce capacitive coupling to the data lines). 

Useful as horns on a bull (at least for us)! 

 

That said, if you are able to hear differences between USB cables, then our USPCB Adapter should sure make a nice difference for you. It is a 4-layer impedance-controlled circuit board--designed to preserve the exact signal integrity of whatever you are feeding through it. In other words, unlike all USB cables (which modify SI--that's why you can hear character differences!), a high-speed USB eye-pattern put through a USPCB Adapter will look virtually identical.

There is no cable like no cable; and the USPCB Adapter is not a cable... 

[Superdad]

2 hours ago, OldBigEars said:

Alex - do you happen to know if a Chord Qutest needs VBUS?

 

I think it does. But you you can use the VBuS switch of the USPCB Adapter to find out for sure. 

[Superdad]

16 minutes ago, MarkusBarkus said:

This is a $3,500.00 USB cable (plus 5K retainer for the divorce attorney), so stop reading now or grab your air-distress bag if you get queasy during take-off and landing. ....  

Time to rewatch Ocean's 11 for tips on planning a casino heist. 

 

LOL!