Then, for not a lot of money......

[Jimbob54]

https://positive-feedback.com/audio-discourse/audioquest-dragonfly-usb-dac/

And in the grand scheme, $300 for a DAC isn't that much. But $1200!!! for an interconnect to run it to his preamp??

Also, what's the max feasible length of USB from computer to dac for audio?

 

[maverickronin]

Also, what's the max feasible length of USB from computer to dac for audio?

Just the USB spec. Without active repeaters 3 meters for 1(.1) and 5 for 2.0.

 

[Arvind]

Just the USB spec. Without active repeaters 3 meters for 1(.1) and 5 for 2.0.

Is there any study to show active repeaters degrade the signal?

 

[maverickronin]

Is there any study to show active repeaters degrade the signal?

In terms of USB, an active repeater is just a one port hub that serves to reclock the signal and reset the cable length limit so it doesn't cause any more problems than a normal hub.

 

[mansr]

Just the USB spec. Without active repeaters 3 meters for 1(.1) and 5 for 2.0.

The USB spec does not explicitly limit the cable length. It specifies a maximum signal propagation time of 18 ns for low-speed cables (must be captive) and 26 ns for full-speed and high-speed cables (detachable or captive). With normal cable construction, these limits translate to roughly 3.6 m and 5.2 m. Additionally, the length may be limited by signal attenuation.

In practice, delays longer than the permitted 26 ns will work provided the end-to-end round-trip delay remains less than 1.5 μs. An active signal conditioner/repeater might overcome the attenuation of a cable that would otherwise be too long. I wanted to test this, but I don't seem to have enough widgets to string together a cable run long enough to compromise the signal.

 

[maverickronin]

The USB spec does not explicitly limit the cable length. It specifies a maximum signal propagation time of 18 ns for low-speed cables (must be captive) and 26 ns for full-speed and high-speed cables (detachable or captive). With normal cable construction, these limits translate to roughly 3.6 m and 5.2 m.

Yeah, but that's basically the the same thing unless you've got an active cable that converts the signals to tachyons or something...

 

[mansr]

Yeah, but that's basically the the same thing unless you've got an active cable that converts the signals to tachyons or something...

If the signal velocity is 0.9 times light-speed, you can have a 7 m cable without violating the 26 ns rule. Such a cable is possible to construct, though it will be expensive.

 

[Jimbob54]

If the signal velocity is 0.9 times light-speed, you can have a 7 m cable without violating the 26 ns rule. Such a cable is possible to construct, though it will be expensive.

Probably cheaper than that £1200 aux cable though

 

[maverickronin]

If the signal velocity is 0.9 times light-speed, you can have a 7 m cable without violating the 26 ns rule. Such a cable is possible to construct, though it will be expensive.

Yeah, I was intentionally over simplifying it since you don't see that kind of construction in USB cables. Even going all the way to c is still only a ~50%-ish increase in length though.

Probably cheaper than that £1200 aux cable though

Probably not very ergonomic either.

I found this air-gapped coax which claims 93% the speed of light. Only problem is it's 3 inches wide. You probably only need two of them for the D+ and D- though (mansr probably knows for sure...) so the cable would "only" be 3"x6". You can also use your extra cable stands to keep it propped up so it doesn't tear the socket out of your PC or DAC.

It would even put those crazy power cables to shame.

 

[mansr]

I found this air-gapped coax which claims 93% the speed of light. Only problem is it's 3 inches wide. You probably only need two of them for the D+ and D- though (mansr probably knows for sure...) so the cable would "only" be 3"x6". You can also use your extra cable stands to keep it propped up so it doesn't tear the socket out of your PC or DAC.

One of those for each of D+ and D- ought to work. In such a setup, the impedance should properly be 45 Ω, but the 50 Ω of that cable is within the 15% allowed tolerance.

 

[maverickronin]

One of those for each of D+ and D- ought to work. In such a setup, the impedance should properly be 45 Ω, but the 50 Ω of that cable is within the 15% allowed tolerance.

We've got a new audiophool product on our hands.

 

[Jimbob54]

Yeah, I was intentionally over simplifying it since you don't see that kind of construction in USB cables. Even going all the way to c is still only a ~50%-ish increase in length though.



Probably not very ergonomic either.

I found this air-gapped coax which claims 93% the speed of light. Only problem is it's 3 inches wide. You probably only need two of them for the D+ and D- though (mansr probably knows for sure...) so the cable would "only" be 3"x6". You can also use your extra cable stands to keep it propped up so it doesn't tear the socket out of your PC or DAC.

It would even put those crazy power cables to shame.