Simple (I hope) cable question

[mansr]

3 hours ago, gmgraves said:

Technically, this is correct. Keep the USB interface as short as practical. Long analog interconnects will do less harm. A 20 foot interconnect is only like 1 dB down at 20 KHz and less in shorter runs. You can't hear that. OTOH, long runs of USB cable invite errors which definitely will make your music sound worse!

A 20-foot USB cable is non-trivial. However, if it works, there is no degradation of the music signal. Also, two 10-foot cables with a hub in between (even a bus powered one) should work reliably.

[mansr]

1 hour ago, gmgraves said:

Yes, a 20 ft USB cable is non-trivial. Suspect that it would work reliably, but I hesitate to say that there is no degradation of the music signal. I suspect that the error rate would go through the roof, and error correction would be working overtime. And if the errors force the DAC into interpolation, one can certainly hear it! 

Either the signal at the far end is within specified parameters, or it is not. If it does meet the spec, there will be no problems regardless of cable length (although actually achieving this for cables longer than about 10 feet is tricky). Errors only start showing up when the received signal is out of spec. At that point, the degradation quickly becomes catastrophic.

[mansr]

5 hours ago, esldude said:

The USB active extenders have a circuit of some sort in the far end of the cable.  It is of course USB powered.  The signal leaving that is fractionally lower than 5 volts (I think for some reason I can't remember they are 4.5 volts).  If you are feeding a USB powered device that might cause trouble (though not always I have found).  If you are simply passing data it has never been an issue in my experience with them.

High-speed USB signalling is only 400 mV, so if the circuitry at the far end doesn't require a full 5 V, a slight drop will be of no consequence.

 

5 hours ago, esldude said:

I have never tried feeding an active hub at the end of one and then feeding another extender from that.  It might well work for lengths of 150 feet that way.  Yes ethernet was made for that, but sometimes your gear only has USB. 

In such cases, an Ethernet based USB extender might be a better option.

[mansr]

2 hours ago, Em2016 said:

John mentioned he saw errors with cheap cables over 5ft.

 

But there are High Speed Certified USB cables that are cheap and up to 10ft (Supra not included in this because those aren't cheap).

 

So my question is asking if errors showing up with cables >5ft even included cheap High Speed Certified Cables.

 

Or does High Speed Certified mean error free, regardless of length. 

 

For those that don't know, the Supra USB cables are High Speed Certified for up to 15ft but I'm not asking about those.

The USB spec is quite clear. If the signal at the receiving end meets the eye pattern requirements, there should be no errors. Obviously, a longer cable is more sensitive to external interference, even if the signal degradation is otherwise within limits.

[mansr]

23 minutes ago, Em2016 said:

Understood. My question is to John regarding his actual findings though, i.e. were there any cheap High Speed Certified cables over 5ft that had errors showing up, or were the only ones that failed over 5ft NOT High Speed Certified.

Certified by whom?