USB isolation between Computer and DAC

[mowgli80]

Oh ! I thought that Digital Guaranteed Perfection, always .....

Since cables cant influence Digital Transmission / sound, how can a ferrite bead, sitting on a cable.

Its all Purrrrrr fect ... No ?

It should be perfect ... scroll down to error handling. Good cable only reduces chances of errors but does not mean errors from regular cable will be used. In other words, sound wont change .

 

[ravi_d]

It should be perfect ... scroll down to error handling. Good cable only reduces chances of errors but does not mean errors from regular cable will be used. In other words, sound wont change .

That is fine for copying data which is not time critical. In case of audio application the DAC can't wait for the cyclic redundancy checks to happen when there is error.
When the DAC working as a master requests the computer to send the next few packets it has to make sure the arriving data is correct or the playback won't be seamless.

 

[Thad E Ginathom]

And, although I'm not using it at the moment ... there is a ferrite thingie on my USB--->ODAC cable. As supplied by JDS Labs*.

Oh well, nothing nor nobody is purrrfect! :lol:

But, although the theory fascinates, I am often more interested in the fact that things work than I am in the reasons why they shouldn't.




*When I do get back to ODAC use, I do mean to change this, just to see.

 

[ravi_d]

@Thad
That tiny ferrite thingie is on the cable which you can chuck it or buy a cable without one instaled. What you will do if its at a place where you can't remove it?

This $1800+ BAD (Berkeley Audio Design) usb to spdif converter seemed to have a massive ferrite gate permanently installed around the isolator chip carrying critical music data.

That product seemed to have proven the best in a shootout of 15 usb to spdif converters. Here is the link of shootout and link to product review.

So I think the ferrite bead doesn't cause any harm at least.

 

[mowgli80]

That is fine for copying data which is not time critical. In case of audio application the DAC can't wait for the cyclic redundancy checks to happen when there is error.
When the DAC working as a master requests the computer to send the next few packets it has to make sure the arriving data is correct or the playback won't be seamless.

Interesting viewpoint. IMO, an audiophile product should care for both correctness and timing. The least I would expect is for the DAC to identify errors and drop the packets and indicate bad cable through gaps in music play. I have no knowhow of how USB works but CRC is not compute intensive and is generally done at hardware layer so it should be feasible. With some buffer on receiver side, even error recovery can be implemented and would be an icing on the cake. However, I still believe that with decent cables built to specifications which can reliably transmit data, sound will not vary with cables.

This is my last post on this subject as it is becoming tiring now. We have fanatics on either side and none will change their point of view :indifferent14:. My belief based on everyday experience is digital communication is pretty robust and reliable within reasonable limits and exotic components should not be required. Others will keep believing otherwise. To each their own.

 

[Thad E Ginathom]

@Thad
That tiny ferrite thingie is on the cable which you can chuck it or buy a cable without one instaled. What you will do if its at a place where you can't remove it? ... So I think the ferrite bead doesn't cause any harm at least.

Likewise, I cannot say that the bead on my cable does any harm. I really don't know. Anyway, even the better USB cables are cheap (or should be hyeah: ) so we can easily chop and change.

I don't know much about USB protocol either. I just thought I'd throw in the information, because it does come from an authoritative source. It's up to others whether they are interested or not.

Also others know far more about shielding and isolation within audio devices. Is it done properly or not? No idea. Don't know if engineers even agree.