DS USB cable


#1

I don’t have any problems with my DS (well one… I don’t get to listen to it enough :) but have a question related to using a usb cable that I could not see an answer to after searching. My understanding is that after the dac does its initial handshake with the server, it doesn’t require the server’s 5V power. That seems to be the case with my server as I can swap to the un-powered output on the server and everything works fine. The thing I’d like to find out is does the DS have a logic board before the DAC process that (after handshaking with the DS) turns off the 5v via a relay etc?

The reason I ask is that there is a considerable price difference in dual headed usb cables v’s cables that are separated (signal and power) but share the same A and B plugs. I definitely want to get rid of the power side after handshake (at the USB B end in particular) by either 1) switching it off/unplugging it (5v side) or 2) using a dual headed cable [more expensive ] or 3) using a dual cable sharing the A USB plugs; IF the DS “disconnects” from the server’s power supply after handshake. ie no current flowing through the power side of the cable interfering with the signal side; and therefore making the use of a dual headed cable unnecessary?

If a tech person can tell us whether or not the DS does isolate (open circuit ) the 5v that would be great.beer_gif

Thanks, Steve


#2

It’s both simpler and more complicated than your guess. There’s a 1k resistor between the 5V line and the USB interface chip. That chip also switches off the 5V after handshaking. BUT to get rid of USB connections issues the Yale OS release for the DS now requires the 5V line be connected in the USB cable all of the time. With previous DS software releases it depends greatly on your OS, specific USB hubs, USB hardware, etc. whether you can run with no 5V cable connection, you need a 5V connection during initial handshaking, or you need the 5V all of the time.


#3

Your answer is not very clear.

Can you please clarify these points derived from your answers? Thanks.

1/ DS does not use that 5V from USB as power supply for any chip/circuit in DS; it only uses that 5V for connecting to the USB interface chip for handshaking/maintaining connection purpose? Is it true?

2/ “There’s a 1k resistor between the 5V line and the USB interface chip. That chip also switches off the 5V after handshaking. BUT to get rid of USB connections issues the Yale OS release for the DS now requires the 5V line be connected in the USB cable all of the time”

“The new Yale release requires that 5V line be connected all the time” --> Can you please clarify connecting to what? Do you mean 5V from USB cable connecting to that USB interface chip all the time then no more switches off after handshaking?

3/ “With previous DS software releases it depends greatly on your OS, specific USB hubs, USB hardware, etc. whether you can run with no 5V cable connection, you need a 5V connection during initial handshaking, or you need the 5V all of the time.”

Do you mean the fact that whether people can run with no 5V cable connection, need a 5V connection during initial handshaking, or need the 5V all of the time has to depend on the USB source as well, not just by the DS?

If that is true and if the source USB host controller needs 5V all of the time and the USB interface chip at DS switches off 5V after handshaking, USB connection will have problem in this situation. Is it correct?

4/ Now, the 5V is needed by DS all the time. How does DS handle 5V from USB if that 5V is dirty? A noisy 5V fed directly to chips will bring EMI/ESD to the internal circuits of DS. Does DS have any isolation/filter circuit to reduce/insert loss to the noise from this 5V?

I understand that when Steve asked the question above, he might want to check if that 5V is well isolated from DS after handshaking so he does not need turn that 5V outside DS after handshaking manually (if this case is OK for his system).

Or if noise in this 5V is filtered well by DS already, he does not need to switch it off.


#4
tamnguyennb said 1/ DS does not use that 5V from USB as power supply for any chip/circuit in DS; it only uses that 5V for connecting to the USB interface chip for handshaking/maintaining connection purpose? Is it true?
Yes, the DS's USB is powered by the DS. The connection to the 5V is only used for signaling.
"The new Yale release requires that 5V line be connected all the time" --> Can you please clarify connecting to what? Do you mean 5V from USB cable connecting to that USB interface chip all the time then no more switches off after handshaking?
Yes, starting with the Yale release the 5V must be connected all of the time. The interface chip follows the 5V signaling as specified by the USB spec, but thru it's own internal high impedance connection.
3/ "With previous DS software releases it depends greatly on your OS, specific USB hubs, USB hardware, etc. whether you can run with no 5V cable connection, you need a 5V connection during initial handshaking, or you need the 5V all of the time."

Do you mean the fact that whether people can run with no 5V cable connection, need a 5V connection during initial handshaking, or need the 5V all of the time has to depend on the USB source as well, not just by the DS?

Absolutely. When any item along the connection violates the USB spec the effects can be quite unpredictable. The 5V line is used to notice connections and disconnections of devices on the USB bus and whether those devices are USB hubs. But also the amount of current allowed to be drawn depends both on the source, every hub and the end device, it's negotiated over the 5V line.
If that is true and if the source USB host controller needs 5V all of the time and the USB interface chip at DS switches off 5V after handshaking, USB connection will have problem in this situation. Is it correct?
No, the USB interface chip keeps a high impedance connection to the 5V line, the phrase "switches off" is confusing, but there is no significant amount of current being drawn.
4/ Now, the 5V is needed by DS all the time. How does DS handle 5V from USB if that 5V is dirty? A noisy 5V fed directly to chips will bring EMI/ESD to the internal circuits of DS. Does DS have any isolation/filter circuit to reduce/insert loss to the noise from this 5V?
Noise and EMI conduction is much more complicated than this. In an ideal universe the 1k resistor gives 60dB of isolation, but that also depends on the current being drawn by the USB interface chip, etc. But higher frequencies take simpler inductive/capacitive paths that can be the same even if there was a relay on the 5V line. Further the USB interface chip takes pains internally to isolate the noise on the 5V line from the USB bus.
I understand that when Steve asked the question above, he might want to check if that 5V is well isolated from DS after handshaking so he does not need turn that 5V outside DS after handshaking manually (if this case is OK for his system).

Or if noise in this 5V is filtered well by DS already, he does not need to switch it off.

You can achieve good sound quality on the DS without disconnecting the 5V. Much of the noise "caused" by the USB cable/connection has nothing to do with whether the 5V line is connected. Ground loops and antennae effects (both sending and receiving) mater quite a bit and probably have more to do with the changes in sound quality differences of different cables than the details of whether the 5V is connected or isolated. I use a cheap USB hub with it's own clean power connection just before the DS as my quick and dirty USB isolation device.

Note that the OS and it’s USB drivers also matter and especially as devices strive to get more energy efficiency they put various peripherals into low energy use states more and more often. In the case of USB this involves negotiation over the 5V lines much more often than just when the system is powered up. In particular Windows 8.0 (as opposed to earlier and later versions) was very persnickety about USB connections since Microsoft had just introduced a bunch more levels of lower energy use states in the system. (8.1 cleaned it up somewhat and does a better job of dealing with USB connections.)


#5

Thanks, Ted for the answers. Your answers should get a thumb-up but I clicked the wrong one (up and down are next to one another) and seems like it does not allow me to fix that :).

Agreed that many factors - emi, esd, ground loops, etc. - impact sound quality for USB. IMO, coupling, if not handled well, can be a problem as well where noise is spread into nearby circuits as well. Filters are needed for both signals and power.

Shame that not many DACs do this very well. Maybe cost issues?

USB 5V in DS is used for signalling only so it may not be a big concern. In some DACs this USB 5V is used to feed USB part and further generator/re-clocking part after the USB part.

A DAC I am using will this 5V to feed USB part.

I was not happy with the sound and my solution for USB cable (http://www.stereo.net.au/forums/index.php?/topic/85328-a-diy-high-quality-hifi-music-server/) improves sound quality extremely well for that DAC. IMO, it benefits other DACS as well; the benefit level depends on the DAC design.

I run 2 short standard USB cables (10 inches each) out from my DIY music server which uses SOTM USB PCI card. One cable carries V+ and V- only from a SOTM USB connector; one carries 5V and GND from my DIY USB connector in the music server. This 5V is from a DIY linear PSU with Texas Instrument LP38798 regulator. The 5V from SOTM is not very clean.

The SOTM PCI does not get 5V from PCI bus. It takes external 12V and runs it thru several regulators to produce 5V output which is fed directly to USB connectors. From your answer, I don’t go deep into this but I think it fools OS/Driver/hardware that there is a device connecting to PCI 5V permanently when powered on.

The 2 cables are terminated by a small board and are filtered (both signal and power) via a WE-EPLE USB Connector on board.

(datasheet: http://katalog.we-online.de/pbs/datasheet/8492121.pdf). Wurth Electronics claims this one is the market’s first USB 2.0 connector with integrated magnetics for ESD protection and EMI filtering

The 5V is filtered further on board by a Murata BNX002-1 EMI filter (datasheet: http://www.farnell.com/datasheets/1481605.pdf; the insertion loss is quite impressive). Out from the board is type B connector to connect to DAC.

I experience a huge improvement in sound quality with this DIY USB cable. This may seem to be over-engineered to others (USB 5V and data are separated, filtered on signal and power, short cables, clean 5V - 5 μVRMS at 10 Hz to 100 kHz; PSRR: 90 dB at 10 kHz, 60 dB at 100 kHz, etc.) but reasonable to me where I have no good measurement equipment to tell what impacts the most.


#6
tamnguyennb said Your answers should get a thumb-up but I clicked the wrong one (up and down are next to one another) and seems like it does not allow me to fix that :).
Very easy to do and the software does not allow one to go back and change it. There are a couple of others that you cannot change as well once chosen. frown

I gave Ted’s post a thumbs up so at least it is now at neutral.


#7

I know this is old. Just want to say MR. Smith you are a Genius. Among everything else you have achieved you figured out that all these USB purifiers are actually based on hubs! Just throw on a LPS. It is awesome to be able to talk to someone that knows this kind of stuff. I mean I figured that out but I doubt I know 1/100th of what you do. Keep up the good work please! I am not being cynical I really mean you seriously know your stuff! No one will admit half of what you divulge which is awesome too.