Anybody look at the Supra USB cables? 23 gauge, individually shielded twisted pairs, 90ohm impedance, and around $50:
Supra cables - HDMI cables for 4k, Digital Video and Audio, USB, Cat7+ Ethernet, Fibre Optic, Quadrax, LoRad power cables, Sword Analogue and Speaker cables
Thinking of either those or the Pangea or Belden Gold usb.
Supra cables - HDMI cables for 4k, Digital Video and Audio, USB, Cat7+ Ethernet, Fibre Optic, Quadrax, LoRad power cables, Sword Analogue and Speaker cables
Thinking of either those or the Pangea or Belden Gold usb.
It is not a simple figure (like 75R video) coax requiring Source and Load impedances of that value to avoid reflections.
USB data uses differential drivers / sensors and usually has 45R to ground on both lines at both ends.
90 ohm differential when you are laying it out on a PCB and down the cable.
section 3.2 page 6
http://www.usb.org/developers/docs/hs_usb_pdg_r1_0.pdf
FYI
http://www.ti.com/lit/an/sprabt8a/sprabt8a.pdf
Just remember any trace widths and gaps are dependant on the PCB stack up... So impedance calcs are your friend... The first one is excellent and used widely by PCB designers and engineers...
Saturn PCB Design - PCB Via Current | PCB Trace Width | Differential Pair Calculator | PCB Impedance
PCB Impedance and Capacitance Calculator
section 3.2 page 6
http://www.usb.org/developers/docs/hs_usb_pdg_r1_0.pdf
FYI
http://www.ti.com/lit/an/sprabt8a/sprabt8a.pdf
Just remember any trace widths and gaps are dependant on the PCB stack up... So impedance calcs are your friend... The first one is excellent and used widely by PCB designers and engineers...
Saturn PCB Design - PCB Via Current | PCB Trace Width | Differential Pair Calculator | PCB Impedance
PCB Impedance and Capacitance Calculator
You'll get a laugh at this one!
I got a relatively cheap 10 metre usb cable and connected it between the Laptop (jRivers/Wasabi program) and the Ayre Acoustic usb dac and it worked okay, even with 24/96k signal!
Possibly it has something to do with that Gordon Rankin system they use in the AA dacs or something
The sound wasn't all that good with a loss of detail and such, but who would think it'd actually function at all ...
I got a relatively cheap 10 metre usb cable and connected it between the Laptop (jRivers/Wasabi program) and the Ayre Acoustic usb dac and it worked okay, even with 24/96k signal!
Possibly it has something to do with that Gordon Rankin system they use in the AA dacs or something
The sound wasn't all that good with a loss of detail and such, but who would think it'd actually function at all ...
As I posted in #464, it is not such a tightly controlled environment as coaxial cable / connectors that are specified to GHz.
Flat ribbon cable (as used in computers) with alternate ground wires is about 105R. Ethernt and twisted telephone cable is 100R to 120R.
It isn't that critical for the low speed / datarate of USB.
Add "controlled impedance" to you search criteria ....
Use USB connectors...
For the cable use a twisted pair designed for USB, the twisted pair is a requirement, with diff impedance of around 90...
Look at this page:
Electronic Components and Parts Search | DigiKey Electronics
I'd just buy a USB cable to be honest....
For the cable use a twisted pair designed for USB, the twisted pair is a requirement, with diff impedance of around 90...
Look at this page:
Electronic Components and Parts Search | DigiKey Electronics
I'd just buy a USB cable to be honest....
The data will get through or it wont, ascribing loss of detail is applicable to analogue cables not digital, you would get lost packets and retransmission and if it was serious you would get signal drop out... not loss of detail...
Yeah Marc, that's what they keep telling me - digits are immune to any analogue degradation and so on, but the result in this situation is a quite marked loss of sound quality - it was pretty obvious - I just now reconnected again and played a couple of tracks, one a 24/96 track and checked it on the headphones to be sure - sounds a bit like mp3.
Apparently, as the engineers here tell me, it's impossible for a usb signal to work over such a long distance regardless of cable quality and I'm just imagining this too!
Sigh, such ridgid attitudes and closed minds ...
Apparently, as the engineers here tell me, it's impossible for a usb signal to work over such a long distance regardless of cable quality and I'm just imagining this too!
Sigh, such ridgid attitudes and closed minds ...
How do you power your DAC? I can imagine the 10m long thin wire USB power cable feeding you DAC provides e.g. lower voltage, leading to lower level, higher distortion, noise in the supply voltage etc. But definitely the problem is not in the digital domain, a cable is not capable of DSPing.
Its nothing to do with ridged attitudes and closed minds... it all to do with determining the causes. Data is sent in packets so you cannot just correlate directly from digital to analogue, there is the packet data sent as well and this is likely to be corrupted just as much as the data embedded in each packet, leading to other problems....
In your case the system is probably not able to transmit the full bandwidth for a 24/96 track so may be down sizing the data or the DAC or receiver has some type of data recover or error correction...
Whatever the cause though ascribing analogue problems to digital data transmission cannot be done directly...
Of course you have only given your perception of this change, empirical data would help confirm or deny the problem.........
Somewhere recently someone using a silver USB cable claimed it made the sound brighter......
Oh and before anyone starts mooing, digital signals do suffer degradation, hence we use S parameters in SIV, but the effects on the resultant information the data is carrying is not the same as it would be on an analogue wave....
Oh and before anyone starts mooing, digital signals do suffer degradation, hence we use S parameters in SIV, but the effects on the resultant information the data is carrying is not the same as it would be on an analogue wave....
Most cable length limits for digital communications are not hard but soft. Exceed the limit and it may still work fine, or it may work most of the time or it may work with occasional errors. So someone saying that over a certain length is "impossible" is either mistaken or being misquoted. Given that it appears to be working, but poorly, with an overspec length I would say that it has done exactly what might be expected.
I have no problem telling the difference between a USB cable, and a USB cable wrapped around a ferrite choke, in my system. Also a difference using a y-cable with a separate USB power supply and lifting the power pin from the PC on the USB cable.
I'm of the opinion that in most cases it's an issue of noise through the USB power supply lines from the computer, and the cable picking up EMI/RFI noise. A properly designed and shielded USB cable with a 90ohm impedance should fix this. Noise from the computer on the USB power can be addressed in multiple ways - power supply, shielding, filters, isolation.
As far as data corruption............in optimal conditions, test shows there is usually almost none. But I'd be willing to bet there are many ways you can corrupt USB data a little or a lot, and still have it come out as music. For example read this:
http://www.usb.org/developers/docs/devclass_docs/frmts10.pdf
I'm of the opinion that in most cases it's an issue of noise through the USB power supply lines from the computer, and the cable picking up EMI/RFI noise. A properly designed and shielded USB cable with a 90ohm impedance should fix this. Noise from the computer on the USB power can be addressed in multiple ways - power supply, shielding, filters, isolation.
As far as data corruption............in optimal conditions, test shows there is usually almost none. But I'd be willing to bet there are many ways you can corrupt USB data a little or a lot, and still have it come out as music. For example read this:
http://www.usb.org/developers/docs/devclass_docs/frmts10.pdf
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.